World Bank Support to Reducing Child Undernutrition An Independent Evaluation © 2021 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org ATTRIBUTION Please cite the report as: World Bank. 2021. World Bank Support to Reducing Child Undernutrition. Independent Evaluation Group. Washington, DC: World Bank. COVER PHOTO Shutterstock/ Riccardo Mayer EDITING AND PRODUCTION Amanda O’Brien GRAPHIC DESIGN Luísa Ulhoa 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 bound- aries, 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 dissem- ination 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. World Bank Support to Reducing Child Undernutrition An Independent Evaluation October 26, 2021 Contents Abbreviations v Acknowledgments vi Overview viii Management Response xvi Chairperson’s Summary: Committee on Development Effectiveness xxi 1. Introduction�������������������������������������������������������������������������������������������������������������� 1 Challenge 2 Evolution of the Global Nutrition Agenda 4 Evaluation Objectives and Scope 8 Methodology 9 Approach to Multidimensionality: Evidence-Based Interventions, Engagement of 2.  Global Practices, and Alignment with Country Needs������������������������������������������14 Portfolio of Nutrition-Related Interventions 16 Evidence-Based Interventions: Is the World Bank Doing the Right Thing? 21 Multisectorality of the Nutrition Portfolio: Engagement across Global Practices 29 Multidimensionality in Country Programs 33 Are Interventions Supported by the World Bank Based on Country Needs? 39 3. World Bank Contribution to Nutrition Results�������������������������������������������������������44 Nutrition Results: Project Performance and World Bank Contributions 45 Measuring Nutrition Results 54 Explaining Nutrition Results: Successes and Failures behind Project Performance 58 4. Conclusions and Way Forward�������������������������������������������������������������������������������63 Lessons 64 Recommendations 68 Glossary���������������������������������������������������������������������������������������������������������������������� 69 Bibliography���������������������������������������������������������������������������������������������������������������� 71 ii Boxes Explaining the Logic of the Conceptual Framework Box 1.1.  of Child Undernutrition 11 Box 2.1. Examples of Interventions by Behavior Change Area 20 Box 2.2. Factors That Facilitate Multisectoral Coordination Efforts 37 Box 3.1. Contributions to Institutional Strengthening in 12 Countries 48 Figures Figure 1.1. Nutrition in the Life Cycle of Mother and Child 4 Figure 1.2. Evolution of the Global Nutrition Agenda 5 Figure 1.3. Conceptual Framework of Child Undernutrition 10 Figure 2.1. Nutrition Interventions in the Portfolio 18 Figure 2.2. Behavior Change Interventions in the Portfolio 21 Figure 2.3. Effective Interventions to Improve Stunted and Linear Growth 24 Alignment of Nutrition Interventions with Evidence on What Works, by Figure 2.4.  Intervention Area 25 Figure 2.5. Projects by Approval Period and Global Practice 30 Figure 2.6. Interventions in Projects by Global Practice 31 Figure 2.7. Projects by Global Practice and Degree of Multidimensionality 32 Figure 2.8. Ethiopia Project Time for World Bank Nutrition Support 35 Figure 2.9. Undernutrition Determinants and Nutrition Outcomes 41 Figure 2.10. Alignment of Portfolio Interventions with Country Needs 42 Figure 3.1. Portfolio Performance 46 Distribution of Interventions, Intended Outcomes, Figure 3.2.  and Project Indicators 55 Figure 3.3. Measurement of Nutrition Results at the Project Level 56 Figure 3.4. Tracing Evidence of Behavior Change Levels in Actors 58 Table Table 2.1. Systematic Review Map Interventions with a Broad Positive Impact 28 iii Appendixes Appendix A. Overall Methodology 78  ystematic Review Map and Relevance of the Appendix B. S World Bank Nutrition Portfolio 90 Appendix C. Behavior Change Process Map 167 Appendix D. Nutrition Portfolio 189 Appendix E. Behavior Change Portfolio Analysis 224 Appendix F. Heat Map of Country Needs 231 Appendix G. Case Studies 244 Appendix H. Stocktaking of Multidimensional Approaches 298 Appendix I. Multivariate Regression Analysis 316 iv Abbreviations ASA advisory services and analytics CBN community-based nutrition CHW community health worker ECD early childhood development FY fiscal year GP Global Practice HNP Health, Nutrition, and Population IEG Independent Evaluation Group LBW low birthweight M&E monitoring and evaluation RETF recipient-executed trust fund SBCC social and behavior change communication SDG Sustainable Development Goal World Bank Group Independent Evaluation Group    v SPJ Social Protection and Jobs SR systematic review SRM systematic review map UNICEF United Nations Children’s Fund WASH water, sanitation, and hygiene All dollar amounts are US dollars unless otherwise indicated. Acknowledgments This evaluation was prepared by an Independent Evaluation Group team led by Jenny Gold, senior evaluation officer, and Mercedes Vellez, evaluation officer, under the overall direction of Alison Evans, Director-General, Eval- uation, and with the guidance and supervision of Galina Sotirova, manager, Corporate and Human Development, and Oscar Calvo-Gonzalez, director, Human Development and Economic Management. Santiago Ramirez Rodriguez, a core evaluation team member, was respon- sible for the portfolio review and the econometric analysis (appendixes D and I). Michael Premson supported the use of machine learning to identify the portfolio, and the Endeavour Programme team conducted artificial intel- ligence topic modeling. Rocio Garabito supported coding of the portfolio. Case studies were conducted by Richard Anson, April Connelly, Judith Gaubatz, Jenny Gold, Aliza Inbal, Victor Malca, Denise Vaillancourt, and World Bank Support to Reducing Child Undernutrition  Acknowledgments Mercedes Vellez, with support from Jean-Jacques Ahouansou, Chefou Balla, Mulusew Gerbaba, Umi Hanik, Hamani Harouna, Zione Kalumikiza, Ventura Mufume, Felix Muramutsa, Evelyne Ndipondjou, Huguette Noromiadana, Epimaque Nsanzabaganwa, Denise Van Wissen, and Helder Zavale. Andrea Spray prepared the systematic review map (appendix B); the team is grateful to Ann Flanagan for her comments on an early draft. Dawn Roberts prepared the behavior change analysis (appendixes C and E) and the stock- taking on multidimensional approaches (appendix H); Brian Allen, Salim Habayeb, and Bjorn Ljungqvist supported the analysis. Ryoko Sato supported the heat map analysis (appendix F). Estelle Raimondo provided overall methods support, and Qihui Chen provided advice on the heat map and econometric analysis. Jean-Jacques Ahouansou provided administrative support to the team. William Hurlbut edited the re- port, and Sharon Fisher provided design and editorial support. External reviewers for this evaluation were Shawn Baker, chief nutritionist at the United States Agency for International Development, and former director of nutrition at the Bill & Melinda Gates Foundation and vice president for vi Africa at Helen Keller International; Professor James Levinson, former pro- fessor at Boston University, Friedman School of Nutrition at Tufts University, Harvard University, and University of Massachusetts Amherst, and former director of nutrition at Tufts University International Food and Nutrition Center, Massachusetts Institute of Technology, and United States Agency for International Development; Dr. Bruno Marchal, evaluation methods expert at the Institute of Tropical Medicine, Antwerp; and Dr. Olivia Yambi, cur- rently cochair of the International Panel of Experts on Sustainable Food Sys- tems, and former United Nations Children’s Fund regional nutrition adviser for Eastern and Southern Africa and representative in India, Kenya, and Lao People’s Democratic Republic. The team is grateful to all the staff who generously shared documents, insights, and experiences and engaged with us throughout the evaluation. Thanks are due to the country offices of Ethiopia, Indonesia, Madagascar, Malawi, Mozambique, Nicaragua, Niger, and Rwanda for their support during the case studies. World Bank Group Independent Evaluation Group    vii Overview Insufficient intake or absorption of nutrients results in undernutrition in children and negatively affects their health, physical growth, and cognitive development. These and other nutrition outcomes are affected by immediate determinants that include caregiving practices, dietary intake or diversity, and the health status of the mother and child. These immediate determi- nants are all difficult to realize when communities lack adequate underlying determinants of nutrition, such as access to nutrient-rich food, caregiving resources, health care, and water, sanitation, and hygiene (WASH) services. Successfully addressing both the immediate and underlying determinants of nutrition requires changing behaviors related to feeding, caregiving, health, and WASH practices throughout the life cycle of the mother and child and social norms related to early marriage, early pregnancy, birth spacing, and women’s empowerment. This evaluation assesses the contributions of the World Bank to improving nutrition determinants and outcomes for children through its interventions during fiscal years (FY) 2008–19. The evaluation uses a variety of evidence at World Bank Support to Reducing Child Undernutrition  Overview the global, country, and portfolio levels. Its findings are intended to inform the design of future nutrition support. Main Findings In line with the conceptual framework of child undernutrition, the World Bank’s approach to nutrition has evolved from a narrow focus on food secu- rity to a portfolio of multidimensional and multisectoral support. The multi- dimensional support combines nutrition-specific, nutrition-sensitive, social norms, behavior change, and institutional strengthening support. Institutional strengthening accounted for the largest share of the rapidly growing portfolio over FY08–19. Nutrition-sensitive interventions that aim to improve access to nutritious food, maternal resources, health care, and WASH services increased during the evaluation period. Meanwhile, nutrition-specific interventions that aim to address the immediate determinants of nutrition have not seen the viii same increase. Behavior change interventions are cross-cutting in the portfolio, especially in support to communities. Social norms interventions, which can support an understanding of gender roles in decision-making that may in- fluence nutrition status among children and pregnant and lactating women, remain relatively limited in the nutrition portfolio. The portfolio supports interventions known to be effective in improving nu- trition determinants and thus contributing to the reduction of child under- nutrition. The increasing focus on nutrition-sensitive interventions in recent years is consistent with growing global evidence of the need to support both nutrition-specific and nutrition-sensitive interventions in countries where there is a need. Nonetheless, there is an opportunity for World Bank projects to better emphasize nutrition-sensitive and nutrition-specific interventions that work. Nutrition-specific interventions that work to address immediate nutrition determinants in countries can be balanced with support to inter- ventions that work across sectors to address underlying nutrition deter- minants; support to institutional strengthening of stakeholder, policy, and services; and knowledge work to facilitate evidence, learning, and leadership. The evaluation confirms that the World Bank’s approach to nutrition—ad- dressing dimensions of underlying and immediate nutrition determinants, social norms, behaviors, and institutional strengthening—provides a plausi- ble pathway to improve nutrition outcomes. A combination of results across these dimensions is critical to support needs in countries. The associations among access to health services and social norms and a country’s nutrition World Bank Group Independent Evaluation Group    ix outcomes are the strongest, followed by access to WASH and food and care. Although World Bank interventions generally address country needs at the national level, significant gaps remain in addressing social norms and WASH. Gaps in country needs relate to areas where there are low levels of nutrition determinants and a lack of support for improvement. The alignment of the nutrition portfolio with country needs is particularly high in access to health care that has the strongest association with country nutrition outcomes, but synergistic support in social norms and WASH is often lacking in countries where these determinants are disadvantaged. Case studies revealed that at the subnational level within-country alignment and targeting is challenging. Support to nutrition is led by various Global Practices (GPs), and in most countries, interventions are fragmented across projects and time, and coordination to ensure support to all relevant nutri- tion determinants is limited. Country experiences also suggest a need for strengthening multisectoral arrangements for nutrition. The key for multisectoral response is having consistent support to develop leadership, services, systems, policies, and ev- idence to help countries sustain support to nutrition that involves multiple actors and sectors. Most institutional strengthening efforts in the case study countries are in one sector, with increasing examples of projects that con- tribute to strengthening multisectoral approaches for nutrition. Moreover, case studies revealed that core nutrition projects are important because of their intentional design to address nutrition determinants. Noncore projects that integrate nutrition interventions do not have explicit nutrition objectives, are often not designed to improve nutrition determinants, and do not have a heavy nutrition focus. Health, Nutrition, and Population projects, for example, focus on health and family planning interventions, Water proj- ects on WASH interventions, and Agriculture projects on agriculture and food approaches, and these interventions may integrate support to nutrition. Core projects, in contrast, are intentionally designed to support nutrition interven- tions that target immediate and underlying nutrition determinants. World Bank Support to Reducing Child Undernutrition  Overview The World Bank is also increasingly successful in achieving results relat- ed to underlying nutrition determinants and institutional strengthening, although the achievement of immediate nutrition determinants is more challenging given that they are higher on the results chain. The performance of World Bank projects in achieving underlying determinants improved over the evaluation period, with the most successful area being agriculture and food, although evidence also shows that the targeting of projects to address underlying nutrition determinants could be improved. In addition, successful institutional strengthening of national and subnational systems is helping in some countries to institutionalize policies, effective services, and stake- holder engagement to enhance the achievement of nutrition determinants and outcomes and to ensure sustained programs for continued outcomes improvement. Project achievements in immediate determinants of nutrition resulting from nutrition-specific interventions have declined in recent years and require greater emphasis and more consistent longer-term support. x Although the overall measurement of results has improved, persistent measurement gaps highlight areas to strengthen the portfolio results. Mea- surement of expected results, especially those related to immediate nutrition determinants and to behavior change and social norms, must increase to foster learning and improve the results monitoring when these interventions are implemented in projects. The evaluation highlights encouraging bright spots, including an increasing nutrition portfolio in countries burdened by undernutrition and improved nutrition outcomes in some countries. In countries burdened by undernu- trition, the World Bank invested an estimated $22 billion in nutrition across multiple sectors from FY08 to FY19 (including about $5.8 billion in recipient- executed trust funds), with the number of projects tripling in recent years. This financing has supported interventions with broad positive evidence of effectiveness that can influence multiple nutrition outcomes and deter- minants. Some countries, Madagascar and Senegal among them, now have more than a decade of experience using a combination of financing and knowledge work to improve nutrition outcomes through multidimensional nutrition programs, from which other countries can learn. At the same time, the nutrition portfolio is young, with many countries recently developing their support, and there are opportunities to further improve the evidence base of interventions, knowledge work, the addressing of nutrition in the country programs, and results achievement and measurement. World Bank Group Independent Evaluation Group    xi Lessons Five lessons follow from the findings: 1. More intentional planning of nutrition support (financing and adviso- ry services and analytics) is needed in the country portfolio to improve nutrition determinants, social norms, behavior change, and institutional strengthening. The multidimensionality of the country portfolio matters for results. » Interventions can be supported by multidimensional projects that imple- ment a range of interventions to address nutrition determinants or by trust funds and partnership, and better GP coordination. Interventions can also be integrated in noncore projects in GPs if they are accompanied by learning to design and target nutrition interventions and internal efforts to coor- dinate implementation. Trust funds and partnerships have been especially catalytic to designing new support in countries, which can be expanded with government ownership to develop comprehensive nutrition services. » Institutional strengthening can be done through support to stakeholder en- gagement, the development of nutrition services, and the coordination of plans, financing, and policies. At the national level, institutional strengthen- ing can help develop multisectoral nutrition approaches and arrangements to coordinate, finance, plan, and communicate nutrition. At the local level, institutional strengthening has been important to engage stakeholders for the planning, monitoring, and delivery of nutrition programs. Links among these levels are also important for accountability and capacity building. » Addressing social norms is important to improve nutritional outcomes in countries. Only 6 percent of World Bank nutrition interventions address so- cial norms. In particular, supporting the empowerment of key change agents can influence other behaviors and facilitate changes toward nutrition deter- minants. World Bank Support to Reducing Child Undernutrition  Overview 2. The targeting and continuity of support in countries matter to successfully influence nutrition determinants. The evaluation finds that the targeting, continuity, and sustainability of nutrition interventions are important for achieving expected results from multisectoral nutrition approaches. » The quality and extent of subnational targeting of multisectoral interven- tions matter for the ability to address (disaggregated) needs within coun- tries. Interventions must come together in the same community to syn- ergistically address identified needs. Multidimensional projects are one option to coordinate interventions to meet needs in the same community, but they have not performed better or worse overall. An alternative is im- proved coordination across GPs and with other development partners in the implementation of multisectoral interventions. » Continuity of support, particularly at the community level, is important for successfully influencing nutrition determinants for results. Community in- terventions involve building the capacity of a wide range of actors and pro- xii moting behavior change, which need to be sustained. Strong community- based implementation is shown to be a success factor for improving project performance. 3. Improving the measurement of results for interventions addressing nu- trition determinants and behavior change will support improvements in nutrition outcomes in countries. » Although the World Bank has improved its results measurement in the past 10 years, some areas still are not well measured. Projects measure only about 60 percent of the achievements of supported interventions toward nutrition determinants. The evaluation consistently identifies monitoring and evalu- ation of nutrition indicators as a pathway to improve project performance. » The World Bank’s nutrition-sensitive interventions increasingly have achieved results in underlying determinants of nutrition in countries. Yet, nutrition-specific interventions, mainly implemented by Health, Nutrition, and Population, have not seen the same improvements in immediate deter- minants of nutrition, and these results are more challenging to achieve and require consistent support in countries. Areas where projects had limited success include diet diversity, child feeding, and micronutrient outcomes in women and children. » Most projects do not track behavior change results along the results chain (engage-learn-apply-sustain). The World Bank’s contributions to behavior World Bank Group Independent Evaluation Group    xiii change focus mostly on lower-level indicators related to the engagement of actors. There is a need for learning in countries to better track behavior change, including on routine and periodic data sources to support results. Appendix C offers an example of a qualitative tool assessment used to track behavior change. 4. Refocusing the portfolio to have greater emphasis on a mix of nutrition- specific interventions balanced with nutrition-sensitive interventions across GPs can improve nutrition programs in countries. Although nutrition-sensitive interventions have increased in the portfolio, a simi- lar proportional increase in nutrition-specific investments supported by health and other sectors is seen in only some countries (such as Rwanda), despite the critical importance of supporting these interventions in coun- tries. The evaluation’s systematic review map shows that effective inter- ventions can be delivered by health, social protection, agriculture, and WASH sectors. Investing in improvements to nutrition-specific interven- tions and nutrition-sensitive support in countries is needed. 5. Learning—the systematic generation and use of knowledge work—is important to help countries design and expand effective nutrition policy and programming. Some case study countries have used a combination of knowledge work to help develop nutrition interventions and policies. Key examples are Ethiopia, Indonesia, Madagascar, Rwanda, and Senegal. » Country-level learning requires a stream of analytical work (evaluations, di- agnostics, and so on) to improve interventions and expand their targeted delivery in national programs. For example, Madagascar had over a decade of advisory services and analytics to develop its community-based program, which is being expanded. » Because nutrition is often not the objective of GP projects (such as those in Agriculture and Water), interventions do not target improving nutrition determinants and in some cases might even negatively affect child under- nutrition (as in the example of cash cropping). Attention to this issue and World Bank Support to Reducing Child Undernutrition  Overview learning has already started at the global level, for example, through re- search on nutrition-sensitive agriculture. » Combining analytical work (such as evaluations and diagnostics) with knowledge sharing (within and across countries) and leadership-building activities in countries helps generate political commitment and the use of evidence to inform policies and programs and leverage resources. Recommendations The preceding lessons support two recommendations for the World Bank: 1. Adjust nutrition programming in country portfolios to (i) give more priori- ty to institutional strengthening of stakeholder engagement, coordination, and services for nutrition and (ii) increase focus on subnational targeting of interventions to reflect areas of greatest disadvantage and persistency of need. xiv 2. Strengthen nutrition support in GPs to (i) rebalance investments to have greater emphasis on nutrition-specific interventions and (ii) increase focus on social norms interventions and behavior changes, with more attention to tracking expected achievements to improve nutrition deter- minants. World Bank Group Independent Evaluation Group    xv Management Response Management of the World Bank thanks the Independent Evaluation Group (IEG) for the opportunity to respond to the report, World Bank Support to Reducing Child Undernutrition. The World Bank appreciates the close consul- tations IEG maintained with the operations teams during the evaluation. Overall Management welcomes this timely evaluation, given that the coronavirus pandemic (COVID-19) is undermining global progress toward Sustainable Development Goal 2.2. One projection is that over and above the current 149 million stunted children, an additional 9.3–13.6 million children will suffer from acute malnutrition, and 2.6–3.6 million more children will be stunted by 2022, rolling back years of progress. There is also a grave risk that these World Bank Support to Reducing Child Undernutrition  Management Response malnourished children will learn less in school and grow up to be less eco- nomically productive as adults.1 The Human Capital Index Update for 2020 warns that a decade of human capital gains could be reversed by COVID- 19.2 The decision to elevate human capital as an special theme for the 20th Replenishment of the International Development Association reaffirms the World Bank’s commitment to enhancing the focus on the nutritional status of children as part of the World Bank’s COVID-19 response. Management is pleased with the report’s conclusion that the World Bank’s ap- proach to nutrition is sound and has evolved in a positive direction. The report states that “the World Bank’s approach to nutrition has evolved from a narrow focus on food security to a portfolio of multidimensional and multisectoral support” (viii) and it “provides a plausible pathway to improve nutrition out- comes” (ix). Further, management welcomes the finding that nutrition is being mainstreamed into sectors beyond Health, Nutrition, and Population and that non–Health, Nutrition, and Population projects accounted for 63 percent in FY14–19, with Agriculture being the largest at 29 percent. Management is also pleased to note the report’s finding that the World Bank is selective in its country engagements and has targeted its nutrition operations especially xvi in those countries that had significant child undernutrition levels. The report notes that “in countries burdened by undernutrition, the World Bank invest- ed an estimated $22 billion in nutrition across multiple sectors from FY08 to FY19 (including about $5.8 billion in [recipient executed trust funds] RETF), with the number of projects tripling in recent years” (63). Outcome Orientation Management agrees with the report’s findings regarding the World Bank’s effective support to indirect pathways to high-level outcomes. The report highlights the strategic role that the World Bank has played in convening and influencing the global nutrition agenda, so the World Bank’s impact goes well beyond the projects that it finances. Scaling-Up Nutrition has been cited “as an example of the Bank Group’s effective convening” (6). The growth in the nutrition portfolio over the review period is also reflective of the World Bank’s efforts at the regional and country levels in advocating with govern- ments to invest in nutrition. At the regional level, child nutrition interven- tions have been progressively integrated within the human capital regional plans and embedded in the human capital upstream support for policy and institutional reforms—including through development policy financing instruments, particularly in the South Asia and East Asia and Pacific regions. At the country level, many country programs have supported institutional reforms to support the nutrition agenda. The report also notes that “the World Bank is also increasingly successful in achieving results related to World Bank Group Independent Evaluation Group    xvii underlying nutrition determinants and institutional strengthening . . . Suc- cessful institutional strengthening of national and subnational systems is helping in some countries to institutionalize policies, effective services, and stakeholder engagement to enhance the achievement of nutrition determi- nants and outcomes, and to ensure sustained programs for continued out- comes improvement” (x).3 Management believes that the long-term support that convening, knowledge and operational engagement in institutional strengthening brings to countries—beyond the typical project implementa- tion time frame—is fundamental to improving child nutritional outcomes. Management supports the report’s quest for better measurement of child nu- trition results in projects, yet it notes that high-level outcomes materialize long after project closing. First, the report notes that “most projects do not measure sustained behavior change results... The World Bank’s contributions to behavior change focus mostly on lower-level indicators related to the engagement of actors.” (xiii). Although management supports the increased focus on monitoring and evaluation, including interventions to change social norms and behaviors, it also cautions that sustained behavior change is a long-term development impact that is not easily measured or captured with- in a typical Bank project cycle. Project development objectives are grounded in the realism of what can be measured during the project lifetime, and proj- ects are to be evaluated based on the impactful change that they can rea- sonably expect within the project and its lifetime. Also, the persistent gaps that exist when tracking achievements from nutrition-specific and social norms interventions requiring behavioral changes may be explained by the limited availability of information to differentiate adherence to social norms from the usual patterns of behavior reflecting food availability, affordability, convenience, and familiarity. It is important to note that very few studies in the global literature measure outcomes related to social norms; this reflects measurement challenges and the lack of globally validated indicators of World Bank Support to Reducing Child Undernutrition  Management Response social norms globally. As noted in the report, several projects are now pi- oneering more rigorous evaluation of nutrition programs, for example in Madagascar, Rwanda, and India, among other countries. In this context, it is worth noting that the Bank Group has recently issued guidance to strength- en measurement of high-level outcomes, such as improved child nutrition, at the country level and over multiple Country Partnership Frameworks, as part of a road map to strengthen its outcome orientation. Recommendations Management agrees to adjust nutrition programming in country portfolios (i) to give more priority to institutional strengthening of stakeholder en- gagement, coordination, and services for nutrition; and (ii) to increase the focus on subnational targeting of interventions to reflect areas of greatest disadvantage and persistence of need (recommendation 1). As mentioned above, the emphasis on institutional strengthening at the country level is a key aspect of World Bank engagement with clients, and the World Bank will continue to support institutional strengthening yet more decisively. As xviii highlighted in the evaluation, the World Bank has been effective at target- ing investments in countries with child undernutrition and will continue to emphasize subnational targeting to reduce child nutrition disparities within countries whenever needed. Although management agrees with the spirit of recommendation 2 ([i] to rebalance investments to have greater emphasis on nutrition-specific in- terventions and [ii] to increase the focus on social norms interventions and behavior changes, with more attention to tracking expected achievements to improve nutrition determinants), management will continue to be guided by global evidence. Global evidence, compiled in The Lancet Series (2008, 2013, 2021),4 suggests that nutrition-specific interventions may be more effective when complemented with nutrition-sensitive interventions, and vice versa, and this is the approach that the World Bank plans to continue pursuing, de- pending on specific country contexts. Management therefore finds the word rebalancing somewhat ambiguous, as it suggests that greater emphasis on nutrition-specific interventions is required. Management also believes that in the case of social norms and behavioral change interventions, the World Bank should also follow global best practices. Management agrees with the report’s finding that greater focus on changing social norms is needed but also notes that inducing impactful behavioral change requires a long-term multifaceted engagement informed by evidence. Although there is consen- sus on the need to do more in relation to social norms, the evidence base to support country-level changes is still evolving. Best practice evidence World Bank Group Independent Evaluation Group    xix suggests that increasing investments in evidence-based nutrition-specific interventions complemented with nutrition-sensitive sectors is paramount to improving nutrition outcomes, in addition to scaling-up interventions to address social norms. In addition, in many country contexts, other local or international partners that have a larger presence in the field may have a greater comparative advantage than the World Bank in changing social norms, and these changes may in fact be catalyzed by their complementary projects. Further, as stated above, it is important to note that social norms take a very long time to change and cannot realistically be measured with- in the time frame of World Bank projects, particularly considering the lack of globally validated indicators for measuring social norms. Several World Bank projects are now pioneering measurement of nutrition outcomes, and management will continue to support such efforts, in the quest for increased outcome orientation. https://blogs.worldbank.org/voices/financing-sdg2-hunger-and-malnutrition-what-will-it- 1  take. 2  Through simulations, the 2020 Human Capital Index report shows that without any remedi- ation, a decade of human capital gain could be reversed by the pandemic with a 0.44 percent drop in the index globally, and up to 0.73 percent loss in low income countries. (See World Bank. 2020. The Human Capital Index 2020 Update: Human Capital in the Time of Covid-19.) 3  “Almost 40 percent of World Bank support is institutional strengthening, especially aimed at improved nutrition service delivery, such as quality assurance approaches, capacity building, and performance-based systems” (14). 4  https://www.thelancet.com/pdfs/journals/lancet/PIIS0140-6736(21)00568-7.pdf World Bank Support to Reducing Child Undernutrition  Management Response xx Chairperson’s Summary: Committee on Development Effectiveness The Committee on Development Effectiveness met to consider the Indepen- dent Evaluation Group (IEG) evaluation entitled World Bank Support to Re- ducing Child Undernutrition and the draft World Bank management response. The committee noted the timeliness of the evaluation at a time when the coronavirus pandemic (COVID-19) has undermined global progress toward Sustainable Development Goal 2.2 and deepened many nutrition challenges. Members welcomed management’s broad agreement with IEG’s recommen- dations and its acknowledgment that, despite the progress made, there is room for improvement. They asked management to be more specific about the actions it will take to implement the recommendations and the evidence it will be tracking and collecting to monitor progress. Members were pleased to learn that the World Bank’s approach to nutrition is sound and that it has evolved in a positive direction, moving from a nar- row focus on food security to a portfolio of multidimensional and multisec- toral support; that nutrition is being mainstreamed to non-Health Nutrition World Bank Group Independent Evaluation Group    xxi Population sectors; and that the World Bank is selective in its engagement, and has steered its nutrition operations specifically toward countries that have significant child undernutrition issues. Members welcomed the World Bank’s enhanced focus on the nutritional status of children as part of the World Bank’s COVID-19 response and commitment to elevate human capital as a special theme in the 20th Replenishment of the International Develop- ment Association. They encouraged management to increase investments in nutrition-specific interventions; further strengthen monitoring and evalua- tion frameworks; enhance learning and knowledge sharing across the World Bank; translate lessons into operational guidance for Global Practices and country teams; and strengthen coordination across Global Practices and with other partners. Many attendees highlighted the broader aspect of malnutrition—including obesity—stressing the importance of tackling this challenge, which also affects some low-income countries, compromises human capital, and in- creases susceptibility to noncommunicable diseases. They appreciated man- agement’s explanation that it conducted a major analytic study on obesity through trust-fund support from the government of Japan, which they plan to use to inform its work. Nonetheless, alluding to the Food System Summit, the committee underscored the importance for the World Bank to adopt a holistic approach and play a convening role in supporting the implementa- tion of voluntary guidelines on food systems and nutrition, and in scaling up national food systems that prioritize access for all to healthy diets, ensure food security, and promote a nutrition-sensitive agenda. They called on the World Bank to demonstrate its comparative advantage as a knowledge and solutions bank that strives to help client countries meet their Sustainable Development Goals and prevent and mitigate adverse economic impacts and health consequences for its citizens. Management explained that Pres- ident Malpass had made a commitment on behalf of the World Bank Group at the United Nations Food Security Summit to focus on the food security World Bank Support to Reducing Child Undernutrition  Report to the Board and nutrition-sensitive agenda, noting that nutrition-specific issues will be the focus of the upcoming Nutrition for Growth Summit to be hosted by the government of Japan on December 7 and 8, 2021, in Tokyo. Members acknowledged that although nutrition-sensitive projects have generally achieved results, achieving results for nutrition-specific inter- ventions has been more challenging. They asked management to clarify its plan for enhancing nutrition-specific interventions. They also asked IEG and management to comment on World Bank’s value added on social norms and behavior change interventions, particularly as the World Bank seeks to ramp up efforts to empower women and strengthen cross-sectoral work across Global Practices to address coordination bottlenecks. Members encouraged the World Bank to continue seeking strategic opportunities to engage more frequently and in a more effective manner on these types of interventions. The committee recognized the World Bank’s convening role in mobilizing partnerships in support of the nutrition agenda and look forward to im- proved metrics that better reflect the measurement of outcomes. xxii 1 | Introduction Highlights Undernutrition negatively affects the health, physical growth, and cognitive development of children, with consequences that last through adulthood and reduce their potential to learn and contrib- ute to society, ultimately affecting human capital accumulation in countries. Stunted growth of children under five, anemia, and low birth- weight—all indicators of child undernutrition—still severely affect the Africa and South Asia regions. Globally, an estimated 150 mil- lion children (22 percent) had stunted growth in 2018, compared with 198 million (33 percent) in 2000. Coordinated effort throughout the life cycle of the mother and child is required to improve nutrition determinants, including diet diversity, child feeding, the health of mother and child, and access to food, caregiving, health services, water, sanitation, and hygiene. The evaluation assesses the contribution of the World Bank’s nutri- tion support in improving outcomes for reducing child undernutri- tion and in improving nutrition determinants through multidimen- sional and collaborative multisectoral interventions. The findings support lessons and recommendations to inform the design of future nutrition support. 1  Challenge Undernutrition negatively affects the health, physical growth, and cog- nitive development of children. It arises from the insufficient intake or absorption of nutrients, which starts with the nutrition and health of the future mother, affecting the growth and development of the child in utero and birth outcomes. Child undernutrition has irreversible effects in early childhood and beyond. The causes of child undernutrition are influenced by the mother and child’s access to and practice of behaviors related to nutrition determinants: caregiving practices, diet diversity, maternal and child health, and access to food, maternal resources, health services, and water, sanitation, and hygiene (WASH). Reducing child undernutrition is essential for enhancing human capital accumulation, boosting economic growth, and reducing poverty. The con- sequences of undernutrition for young children last through adulthood and reduce their potential to learn and contribute to society. These consequences are also often intergenerational, extending to future children. Galasso and Wagstaff (2018) estimate the average per person income penalty from stunt- ed growth is about 7 percent. World Bank Support to Reducing Child Undernutrition  Chapter 1 Global reports on indicators of undernutrition show mixed progress across regions in reducing the stunted growth of children under five, anemia, and low birthweight (LBW), with Africa and South Asia most severely affected. Moreover, although much of Latin America and the Caribbean and of East Asia and Pacific have low national prevalence of stunted growth of children, some countries and subnational areas have levels of stunted growth simi- lar to Africa and South Asia. Stunted growth, wasting, and underweight are the most-used anthropometric measures of child undernutrition. Globally, more than 150 million children (22 percent) were estimated to have stunted growth as of 2018, compared with 198 million (33 percent) in 2000. In Afri- ca, stunted growth rates have improved since 2000, yet total undernutrition is worsening as the population is growing; therefore, the total number of children with stunted growth is increasing. Reducing anemia and LBW have seen similarly mixed progress. The latest figures show that the prevalence of anemia in girls and women of reproductive age has stagnated at about 33 percent. Approximately 20 million babies are LBW globally, compared 2 with 22.9 million in 2000 (Development Initiatives 2020; UNICEF, WHO, and World Bank 2019). Improving child nutrition requires efforts at each stage of the life cycle of the mother and child. Malnourished pregnant women may deliver LBW new- borns, and mothers with low body weight or micronutrient deficiencies may struggle to sustain exclusive breastfeeding or to feed and care for their ba- bies (figure 1.1). Children with low or inadequate nutritional status are more prone to childhood infections, which further aggravate the child’s capacity to absorb nutrients, and have slower growth and impaired cognitive capacity (Maternal and Child Nutrition Study Group 2013). Evaluation of child undernutrition requires the assessment of outcomes at different points in the life cycle of mother and child, with a focus on the ear- ly years of life. Within the life cycle, mother and child are most sensitive to the consequences of undernutrition from preconception through pregnancy, until the child is about two years old. For this reason, nutrition interventions often target mothers, children, and future mothers during this period, in- cluding girls, adolescents, and women before conception and during preg- nancy, and households with mothers and young children. Given the many nutrition determinants that affect the life cycle of mother and child, the challenge of improving nutrition outcomes (anthropometric measurements, micronutrient status, and cognitive development) becomes multidimension- al, requiring interventions in health, agriculture, WASH, social protection, World Bank Group Independent Evaluation Group    3 education, and governance. Thus, improving outcomes in countries requires coordination to improve diet diversity, child feeding, the health of mother and child, and access to food, caregiving, health services, and WASH. It also involves engaging a range of actors, including government, communities, and households, to influence nutrition determinants. Figure 1.1. Nutrition in the Life Cycle of Mother and Child Reduced capacity to care for baby Household caregiver LBW baby Inadequate breastfeeding, malnourished weaning Frequent infections Inadequate Inadequate catch-up in fetal nutrition Inadequate food, care, growth, delayed health, WASH milestones Child stunted Mother malnourished, low energy Reduced cognitive Pregnancy capacity low weight gain Inadequate food, learning, health, WASH Higher neonatal Adolescent mortality, anemia stunted Inadequate food, care, World Bank Support to Reducing Child Undernutrition  Chapter 1 Reduced capacity to contribute health, WASH socially and economically Sources: Adapted from ACC/SCN 2000 and UNCNC21 2000. Note: LBW = low birthweight; WASH = water, sanitation, and hygiene. Evolution of the Global Nutrition Agenda Historically, the World Bank’s nutrition agenda has focused on access to food. In the 1970s, the World Bank approached nutrition by integrating it into poverty reduction through multisectoral rural development projects. Government commitments to implement these projects were often weak (MacNally 1983; World Bank 2014). Later projects shifted to focus mainly on the health sector (Berg 1987), where the challenge became how to meaning- fully integrate nutrition interventions into one component of the project or to expand interventions in health services or interventions that had been confined to small geographical areas (figure 1.2). 4 Figure 1.2. Evolution of the Global Nutrition Agenda Nutrition and rural Agriculture sector starts Rome Increasing role development tracking nutrition-sensitive Declaration of Governance integrated in approaches in projects, on Nutrition Human sector in Institutional developments poverty reduction; following World Bank Capital nutrition focus on food Agriculture Action Plan Early childhood UN Decade Project financing intake nutrition part of of Action on increases World Health Assembly Comprehen- World Bank Nutrition focus on Nutrition- Review of sive Implementation Plan on Nutrition response to (2016–25) nutrition sensitive WASH nutrition in first shared prosperity approaches SecureNutrition World Bank Launch Adoption of Early Years Knowledge Platform projects; first of MDGs SDG nutrition Initiative Nutrition-smart shares knowledge multisectoral N4G raises target and increases agriculture projects had Launch of the SUN commitment stunting focus on innovative limited success movement to nutrition indicator nutrition approach 1970s-90s 2000 2006 2008 2010 2012 2013 2014 2015 2016 2017 2018 2019 Malnutrition: Repositioning First Lancet Scaling Up Second Improving First Global An Investment Key publications What Can Be Nutrition as series, Nutrition: Lancet Nutrition Nutrition Framework Done? shifted Central to providing What Will It series through Report for Nutrition focus to health Development, evidence on Cost?, first Multisectoral All Hands sector galvanizing interventions estimate of Approaches, Incentivizing on Deck, leadership financing framed Nutrition multisectoral Combating needs multidimensional convergence Malnutrition: actions Future of Food, Time to Act, food system UNICEF and nutrition collaboration outcomes Scope of evaluation: fiscal years 2008–19 Sources: Adapted from Rokx 2006 and Shekar et al. 2017. Note: Black type indicates World Bank actions; blue type indicates multipartner actions. MDG = Millennium Development Goal; N4G = Nutrition for Growth; SDG = Sus- tainable Development Goal; SUN = Scaling Up Nutrition; UN = United Nations; UNICEF = United Nations Children’s Fund; WASH = water, sanitation, and hygiene. World Bank Group Independent Evaluation Group    5 Over the years, World Bank support to nutrition has evolved into a more multidimensional and collaborative multisectoral agenda. Countries and development partners have adopted the United Nations Children’s Fund (UNICEF) framework of child undernutrition (UNICEF 1990, 2015), which highlights the need to address multidimensional determinants, including ac- cess to food, caregiving, health services, and WASH throughout the life cycle of mother and child (figure 1.3). Among the milestones in the World Bank’s adoption of a multidimensional or collaborative multisectoral approach is a series of reports on combating nutrition (Gillespie, McLachlan, and Shrimp- ton 2003), strengthening country commitment (Heaver 2005), repositioning nutrition in the development agenda (World Bank 2006), scaling up nutrition (Horton et al. 2010), and improving nutrition through multisectoral ap- proaches (World Bank 2013a). The Scaling Up Nutrition Movement (2010) brought together countries, sectors, and development partners to act on nutrition and began to organize learning and operational efforts regarding the UNICEF framework of child undernutrition and addressing of nutrition determinants. In some countries, the movement initiated policy and institutional reforms to coordinate, plan, measure, and implement nutrition interventions and find solutions to over- World Bank Support to Reducing Child Undernutrition  Chapter 1 come previous challenges relating to the countries’ ownership and delivery of the agenda; that is, nutrition does not fall within the mandate of any one sector (SUN Movement 2019). Within the World Bank, the commitment to the movement renewed the engagement of sectors (agriculture, social pro- tection, health, water, and so on) to address nutrition in country programs (Alderman 2016; Hawkes and Ruel 2008; World Bank 2013a, 2014). The 2020 Independent Evaluation Group (IEG) Evaluation of the World Bank Group’s Global Convening cited the movement as an example of the Bank Group’s effective convening that transformed the execution of nutrition efforts by creating a multisectoral, multistakeholder platform and galvanized momen- tum in reducing malnutrition. In 2008, the first of several Lancet series on nutrition began consolidating the knowledge and evidence on interventions that were effective in improv- ing nutrition outcomes (Maternal and Child Undernutrition Study Group 2008). In 2010, the World Bank published the first estimates for financ- ing nutrition interventions in countries; these estimates have led to more 6 detailed country-level investment cases (Horton et al. 2010), work on the Optima Nutrition budget allocation decision tool (Pearson et al. 2018), and from 2013 became the basis for mobilizing financing for nutrition and polit- ical commitment through Nutrition for Growth, together with partners such as the Bill & Melinda Gates Foundation. Multisectoral knowledge sharing has also been supported through the SecureNutrition Knowledge Platform (World Bank 2017). Since 2016, the Sustainable Development Goals (SDGs) have been adopted to improve nutrition outcomes, and the United Nations has declared the De- cade of Action on Nutrition (2016–25). The Millennium Development Goals had focused on halving the prevalence of underweight children under five by 2015, which did not fully address the importance of nutrition to healthy growth and child development. The need for better nutrition is further recognized in SDG 2, which aims to end hunger, achieve food security and improved nutrition, and promote sustainable agriculture. SDG 2 emphasizes the transformational role nutrition can play in driving human capital devel- opment and the need to address multidimensional nutrition determinants and inequalities in the life cycle of mother and child. The SDG 2 focus on stunted growth was influenced by the World Bank’s strategy of reducing extreme poverty and promoting shared prosperity and by its emphasis on in- equalities in early childhood development (ECD) and nutrition (Denboba et al. 2014; World Bank 2013b). Since the creation of the SDGs, the World Bank has supported nutrition investments in countries and analyses on the eco- nomic costs of child undernutrition (Galasso and Wagstaff 2018; Laviolette World Bank Group Independent Evaluation Group    7 et al. 2016; Shekar et al. 2017; WHO 2014). Global nutrition targets set by the World Health Assembly for 2025 include a 40 percent reduction in stunted growth, a 50 percent reduction in anemia in women, a 30 percent reduction in LBW newborns, and an achievement of at least 50 percent for exclusive breastfeeding (WHO 2014). The launch of the World Bank’s Human Capital Project in 2018 further rein- forced the importance of reducing child undernutrition and of implementing a package of multidimensional interventions to achieve results. The percent- age of children under five who do not have stunted growth is now used as a proxy for healthy child growth based on its emphasis in the Human Capital Index (World Bank 2018). The human capital agenda has led to (i) efforts to improve data on nutrition indicators and (ii) analysis to understand as- pects of multidimensionality relating to how interventions from different sectors can be prioritized and integrated in a package to address disad- vantaged nutrition determinants in a country context, that is, inadequate access to nutrient-rich food, caregiving resources, health services, and WASH (UNICEF, WHO, and World Bank 2019; Skoufias, Vinha, and Sato 2019). Evaluation Objectives and Scope The objectives of this evaluation are (i) to assess the contribution of the World Bank in improving outcomes related to reducing child undernutrition and improving nutrition determinants and (ii) to inform the design of future nutrition support. The evaluation provides evidence on results across sec- tors and lessons from operational experience to feed into country strategies, multidimensional and collaborative multisectoral approaches, and project design, particularly in those countries where child undernutrition is an important factor inhibiting the healthy growth of children and the accumu- lation of human capital. The overarching evaluation question is, “What has been the contribution World Bank Support to Reducing Child Undernutrition  Chapter 1 of World Bank support to improve outcomes and intermediate outcomes in reducing child undernutrition and improving nutrition determinants in countries burdened by undernutrition?” Underlying this question are three main lines of inquiry: » To what extent is the World Bank supporting relevant interventions to im- prove outcomes and intermediate outcomes of child undernutrition and its determinants within the country context? » How is the World Bank implementing multidimensional approaches to sup- port outcomes and intermediate outcomes that reduce child undernutrition, improve its determinants, and strengthen countries’ institutional capacities? » To what extent have World Bank interventions contributed to achieve out- comes and intermediate outcomes of reducing child undernutrition and improving its determinants, and what were the factors of success and failure? 8 To answer these questions, the evaluation focuses on World Bank engage- ments in nutrition (investment operations, development policy lending, and recipient-executed trust funds [RETFs]) that were active during fiscal year (FY) 2008–19 in countries that have reported high levels of stunted growth. Methodology The evaluation design adopts a multilevel analysis at the global, portfolio, country, and intervention levels using quantitative and qualitative evalu- ative evidence and applying participatory, theory-based, and case-based principles. The conceptual framework underpinning this evaluation is adapted from the UNICEF framework of determinants of child undernutrition (figure 1.3; Maternal and Child Nutrition Study Group 2013; UNCNC21 2000; UNICEF 1990, 2015). The framework models interlinked dimensions to sustainably address child undernutrition in a country context. In doing this, the eval- uation takes a systems approach to look at the World Bank’s support and results across these dimensions. These dimensions are nutrition-specific and nutrition-sensitive interventions addressing the immediate and underly- ing determinants of nutrition, respectively, social norms interventions, and institutional strengthening support, considering factors within the country that are used to prioritize and target interventions (box 1.1). The evaluation methods look at each of these dimensions and confirm the interlinkages among the dimensions. This emphasizes the need for a mix support tailored World Bank Group Independent Evaluation Group    9 to needs in countries to achieve results across these dimensions to contrib- ute to nutrition outcomes (anthropometric measures and micronutrients deficiencies). Figure 1.3. Conceptual Framework of Child Undernutrition Human capital benefits during the life course Morbidity and mortality Cognitive, motor, socio� School performance Work capacity and Economic growth in childhood emotional development and learning capacity productivity Outcomes Outcomes for future mothers, mothers, children under 5 and child in the life cycle of mother Improved undernutrition (stunting, wasting, underweight, low birthweight, micronutrient deficiencies) Ag Nutrition-specific Nutrition-sensitive Improved Improved Improved health interventions intermediate outcomes and outputs nutrient intake/ feeding and of mother and interventions Preconception through early childhood (first 1,000 days) Determinants of undernutrition and (Examples) (Examples) SP diet diversity caregiving child » Adolescent nutrition » Agriculture and food Interventions throughout the life course systems, safety, security Edu Lending » Maternal nutrition Lending approaches Health » Breastfeeding, child » Social safety nets for feeding and stimulation Water Access to Maternal and Access to households with children Analytical nutrient-rich childcare health services Analytical work » Dietary support and work food resources and WASH » Complete early child micronutrient development (such as supplementation or Urban stimulation, learning, fortification for children Social norms care, nutritious food, safe » Treatment of Women’s empowerment, early marriage, early environment) malnutrition pregnancy, birth spacing » Women and girls’ Social » Childhood disease empowerment (literacy, prevention and etc.) management Coordination Environment to enable outcomes » WASH approaches Coordination Gov with partners (such as access to clean with partners » Nutrition policies, financing, and coordination water, hygiene » Arrangements to deliver interventions (such as promotion) strategies, targeting, capabilities, coverage of Macro difficult-to-reach communities, partnerships, M&E, » Maternal and child and learning) health and family planning approaches » Engagement and ownership (such as commitment of leaders, community participation and engagement of Health Lending citizens, demand for accountability, and transparent Lending information) Analytical work Institutional strengthening support Analytical work Country context: inequalities in the distribution of outcomes; poverty; health status; demographics; status of women; fragility and conflict; politics; environment World Bank Support to Reducing Child Undernutrition  Chapter 1 Sources: Adapted from Maternal and Child Nutrition Study Group 2013 and UNICEF 1990. Note: The assessment of the contribution of the World Bank’s nutrition support to human capital bene- fits is outside the scope of the evaluation. Ag = Agriculture; Edu = Education; Gov = Governance; Macro = Macroeconomics, Trade, and Investment; M&E = monitoring and evaluation; Social = Social Sustainability and Inclusion; SP = Social Protection and Jobs; Urban = Urban, Disaster Risk Management, Resilience, and Land; WASH = water, sanitation, and hygiene. 10  xplaining the Logic of the Conceptual Framework Box 1.1. E of Child Undernutrition The conceptual framework premises that nutrition outcomes for pregnant women and children (for example, anthropometric measures and micronutrient status) are better among women and children with adequate nutrition determinants. Immediate determinants of child nutrition relate to caregiving practices, dietary intake or diversity, and the health status of the mother and child. It is not possible to realize these factors when communities lack adequate access to underlying determinants of nutrition, in- cluding nutrient-rich food, caregiving resources, health services, and water, sanitation, and hygiene (WASH). Improvements in underlying determinants are interdependent; that is, access to food is not enough without adequate feeding, proper care, adequate and accessible health services, and clean water. Successfully addressing both the immediate and the underlying determinants of nu- trition requires transforming social norms relating to early marriage, early pregnancy, birth spacing, and women’s empowerment (decision-making regarding childcare, food production, health care seeking) and changing behaviors relating to feeding, care- giving, health, and WASH practices, including those related to gender relations and practices. Behavioral interventions are thus central to the framework and can target women, caregivers, children, and other agents of change (such as household mem- bers and community leaders) who can influence the prevailing social norms, and more broadly, behavior practices at the community and household levels. World Bank Group Independent Evaluation Group    11 The conceptual framework suggests that nutrition interventions within a country need to be multidimensional to address both the immediate and underlying nutrition deter- minants in their context; this may require synergizing interventions related to multiple sectors. Nutrition-specific interventions, such as adolescent nutrition, maternal nutrition, breastfeeding support, micronutrient supplementation, child disease prevention, and management and treatment of undernutrition, are expected to influence the immedi- ate determinants of nutrition. Nutrition-sensitive interventions, such as cash transfers, WASH approaches, girls’ education, and food system improvements, are expected to address the underlying determinants. Whereas nutrition-specific interventions are often delivered by the health system and target women and children, nutrition-sensitive interventions may be delivered by various sectors and target households and commu- nities or geographies with inadequate nutrition determinants (access to nutritious food, caregiving resources, health services, and WASH). (continued)  xplaining the Logic of the Conceptual Framework Box 1.1. E of Child Undernutrition (cont.) The country-specific situation, including the distribution of outcomes, frames the context in which to prioritize and target interventions, and the enabling environment frames interventions to strengthen institutional capacities at national and subnational levels over time in a country to support outcomes. Factors of fragility and distributional factors related to inequalities in nutritional outcomes, health and education status, and poverty can create different country scenarios in which to prioritize and target interven- tions to improve undernutrition. Moreover, the distribution of nutrition determinants in a population—that is, access to nutritious foods, caregiving resources, health services, and WASH—can provide information on investment needs. Institutional capacities in the enabling environment at the national and subnational levels can frame priorities for interventions to improve the delivery of services and programs, the engagement of communities, and the implementation of policies to address nutrition in countries. Source: Independent Evaluation Group. The evaluative findings and conclusions are a result of the triangulation of different evaluation components at the global, portfolio, and country lev- World Bank Support to Reducing Child Undernutrition  Chapter 1 els. The evaluation adopted several innovative practices and broadened the methodological applications to ensure construct validity, internal validity, external validity, and reliability of findings through a transparent method- ological design, with clear justification of choices made (see appendix A for the evaluation methodology). At the global level, the evaluation methods included a systematic review map (SRM) that synthesizes the existing evidence from systematic reviews (SRs) of the literature on the effectiveness of nutrition interventions across sectors to support nutrition outcomes, immediate nutrition determinants, or underlying nutrition determinants (appendix B). The SRM provides a tool to visualize the existing evidence and benchmark it against the nutrition portfolio to review the alignment of World Bank support in Global Practices (GPs) to the evidence base. Additionally, a structured literature review iden- tified and categorized behavior change concepts and evidence to develop a set of process maps describing a basic results chain for benchmarking behav- 12 ior change in projects (appendix C). The process maps provide a qualitative tool to review behavior change support to nutrition determinants, which is often not tracked in projects. At the portfolio level, the evaluation conducted a systematic identification, coding, extraction, and analysis of the World Bank’s nutrition lending port- folio based on its relevance, multidimensional approaches, and contributions to nutrition results in countries. The portfolio review and analysis combined a mapping of project indicators to measure nutrition achievements with artificial intelligence theory-based content analysis and unsupervised ma- chine learning techniques to develop a taxonomy of common success and failure factors that influenced the results of nutrition projects (appendix D). In addition, portfolio data were contrasted against a heat map on nutrition outcomes and determinants in countries to assess the alignment of projects’ interventions to the country needs and to understand the empirical links of the conceptual framework (appendix F). Moreover, a qualitative stocktaking exercise of 12 countries was conducted to understand multisectoral ap- proaches to nutrition in different country contexts and how the World Bank helped enhance multisectoral coordination through institutional capacity building (appendix H). Finally, a multivariate regression analysis was done to deepen learning on the portfolio data (appendix I). At the country level, central to the evaluation are eight country case studies (Ethiopia, Indonesia, Madagascar, Malawi, Mozambique, Nicaragua, Niger, and Rwanda) that include a review of nutrition in the country program World Bank Group Independent Evaluation Group    13 (including analytical work), semistructured interviews, and analysis of the World Bank’s contribution to results in each country (appendix G). The report is structured as follows: chapter 2 focuses on the World Bank’s approaches to multidimensionality; chapter 3 looks at the World Bank’s contribution to results; and chapter 4 presents conclusions, lessons, and recommendations to inform the design of future multidimensional nutrition support by the World Bank. 2 | Approach to Multidimensionality: Evidence-Based Interventions, Engagement of Global Practices, and Alignment with Country Needs Highlights The World Bank’s nutrition portfolio is growing quickly, with many new projects since 2014, and it increasingly uses multidimensional and multisectoral interventions. The portfolio engages with issues spanning the conceptual frame- work, with a mix of interventions, including nutrition-specific and nutrition-sensitive approaches. Institutional strengthening ac- counts for the largest share of support, with less attention to social norms and behavior change and limited attention to adolescent health—all of which have been shown to be effective in improving nutrition outcomes. The World Bank aligns its nutrition support with current evi- dence and helps generate knowledge and learning to promote evidence-based policies. However, a range of interventions can be delivered by health, social protection, agriculture, and water, sanitation, and hygiene sectors with consistent evidence, which could be better addressed in the portfolio. This approach can be balanced with knowledge and learning in countries to improve the 14   use of evidence in nutrition programming. Global Practices often collaborate in implementing nutrition inter- ventions, but case studies suggest that they need support in learn- ing how to design effective nutrition interventions in projects where nutrition is not the main priority. The World Bank’s institutional strengthening support can facilitate multisectoral arrangements in two ways. First, it can support the en- hancement of national leadership and subnational governments to coordinate multisectoral actions. Second, it can support the organi- zation of sectoral extension services or community actors to deliver an integrated package of interventions tailored to local needs. Better results also could be achieved through alignment of rele- vant interventions to address the disaggregated needs or priorities of countries. Recent efforts to improve the alignment of interven- tions with community needs are promising. 15  Portfolio of Nutrition-Related Interventions During FY08–19, the World Bank committed $22.7 billion in financing, including about $14.4 billion of International Development Association support, $2.5 billion of International Bank for Reconstruction and Develop- ment financing, and $5.8 billion in RETFs, to support reducing child under- nutrition. The nutrition lending portfolio is young, comprising 282 projects, more than half of which were approved since 2014. The portfolio, mostly channeled through International Development Association financing, sup- ports investments in 64 countries with high rates of stunted growth,1 mainly in the Africa Region (53 percent of projects). Most of the lending support is through investment project financing operations (90 percent), led mainly by the Health, Nutrition, and Population (HNP), Agriculture, and Social Protec- tion and Jobs (SPJ) GPs, among others.2 The portfolio includes a mix of interventions—nutrition-specific, nutrition-sensitive, social norms, and behavior change—and institutional strengthening support, which accounts for the largest share of investments. This mix of support is consistent with the premise that improving child nutrition requires support across the conceptual framework. Investments in World Bank Support to Reducing Child Undernutrition  Chapter 2 institutional strengthening include support to develop institutional capaci- ties at the national and subnational level to improve delivery of services and programs, engagement of communities, and implementation of policies to address nutrition in countries. Almost 40 percent of World Bank support is institutional strengthening, especially aimed at improved nutrition service delivery, such as quality assurance approaches, capacity building, and performance-based systems (figure 2.1). The World Bank is increasingly orienting its support toward nutrition- -sensitive interventions, namely those that address the underlying deter- minants to improve access to nutritious food, maternal resources, health, and WASH services. The share of nutrition-sensitive interventions increased from 27 percent at the beginning of the FY08–19 evaluation period to 39 percent toward the end. Health and family planning interventions con- tinue to be a major support area, and agricultural approaches (for example, home gardens, livestock production, and food fortification), and to a lesser 16 extent social safety nets, receive increased attention. Still, the portfolio con- tinues to have important gaps. The World Bank’s nutrition portfolio also supports nutrition-specific in- terventions, namely those that aim to address the immediate determinants of nutrition. Nutrition-specific interventions account for 23 percent of the portfolio, with more recent investments in dietary diversity and breastfeed- ing than in child disease prevention and treatment. Although there was some increase toward the end of FY08–19, particularly in the Europe and Central Asia and East Asia and Pacific Regions, nutrition-specific interven- tions targeting adolescents do not receive much attention. Behavior change interventions to address determinants of nutrition are cross-cutting in the World Bank portfolio. Successfully addressing determi- nants of child nutrition requires transforming behaviors relating to feeding, caregiving and stimulation, health care-seeking behaviors and treatment compliance, food production diversification, WASH practices, social norms, and service delivery practices (box 2.1). About 85 percent of projects had at least one behavior change intervention. Behavior change is most common in institutional strengthening support targeting service providers’ practices (29 percent), followed by food and care interventions targeting caregivers and households (14 percent; figure 2.2).3 Behavior change interventions in health, agriculture, and WASH sectors are less apparent (about 9 percent). World Bank Group Independent Evaluation Group    17 Figure 2.1. Nutrition Interventions in the Portfolio a. Areas of support in the portfolio by intervention or support area (2008–19) b. Intervention type over time c. Nutrition-specific interventions over time Share of interventions (percent) Share of interventions (percent) World Bank Support to Reducing Child Undernutrition  Chapter 2 d. Nutrition-sensitive interventions e. Social norms interventions over time over time Share of interventions (percent) 18 d. Nutrition-sensitive interventions e. Social norms interventions over time over time Share of interventions (percent) f. Institutional strengthening over time Share of interventions (percent) World Bank Group Independent Evaluation Group    19 Sources: Independent Evaluation Group; portfolio review and analysis. Note: In panel a, boxes report the percentages of total interventions represented by each area. In panel e, because social norms has no subcategories, the bar chart reports numbers of interventions. WASH = water, sanitation, and hygiene. Box 2.1. Examples of Interventions by Behavior Change Area Food and Care » Community or backyard garden promotion, agricultural skills training, promotion of fruits and vegetables or diversification of food production, promotion of local processing and conservation » Parent counseling and education, promotion of toys, promotion of early child- hood development, awareness campaigns, positive deviation modeling, breast- feeding, child feeding promotion and counseling, accompanying measures of conditional cash transfers Health Services » Health and nutrition promotion and counseling, information, education, and com- munication campaigns, accompanying measures of conditional cash transfers, sexually transmitted disease prevention education Water, Sanitation, and Hygiene » Communication campaigns, outreach activities, open defecation–free campaigns, hand washing and hygiene promotion World Bank Support to Reducing Child Undernutrition  Chapter 2 Social Norms » Women’s empowerment activities, awareness campaigns, life skills education, accompanying measures of conditional cash transfer Institutional Strengthening » Awareness campaigns, performance-based financing, coordination activities, continuing education programs for service providers, community mobilization and training on nutrition and health, sensitization of local community leaders Sources: Independent Evaluation Group; behavior change analysis. A notable gap in portfolio coverage is the limited attention to social norms. Despite the consensus that social norms can provide an understanding of gen- der roles, such as those related to decision-making regarding the care of chil- 20 dren, and social and cultural practices that may influence the nutrition status of children and pregnant and lactating women, the focus on women’s empower- ment, early marriage, and childbearing remains relatively narrow in the nutrition portfolio (only 3 percent of interventions and 6 percent of projects). Figure 2.2. Behavior Change Interventions in the Portfolio Sources: Independent Evaluation Group; behavior change analysis. Note: A project was coded as having an intervention in the behavior change category if it had at least one relevant intervention. Boxes report the percentage of total interventions within each area. WASH = water, sanitation, and hygiene. Evidence-Based Interventions: Is the World World Bank Group Independent Evaluation Group    21 Bank Doing the Right Thing? The evidence on what works to reduce child undernutrition and improve nu- trition determinants encompasses many options for interventions in projects. The SRM for this evaluation visually synthesizes the available evidence on the effectiveness of nutrition-specific and nutrition-sensitive interventions across multiple nutrition outcomes and determinants from SRs (appendix B). This synthesis is to benchmark knowledge on interventions that work against GP support in the portfolio. The SRM’s search strategy includes 227 SRs identi- fying 84 types of interventions and 24 nutrition-relevant outcomes (relating to nutrition outcomes, and more intermediate outcomes of immediate and underlying determinants) for children, women, and households. Approximately 36 percent are nutrition-specific interventions (30 out of 84 interventions with available evidence). Nutrition-sensitive interventions account for 64 percent, spanning across health (20 percent), agriculture (19 percent), and, to a lesser extent, WASH (14 percent) and social protection (11 percent) sectors. Although nutrition-sensitive interventions are more in number than nutrition-specific interventions, the evidence supporting these interventions is often weaker. Synthesizing the available evidence on nutri- tion-sensitive interventions is especially importance, since the list of inter- ventions that work for World Bank support in this area has been less clear. A large body of evidence suggests that some interventions have the potential to reduce the long-term effects of undernutrition, although the SRM could not identify a single intervention with a consistent and large amount of evidence of effectiveness to reduce stunted growth, emphasizing the need to mix a range of interventions in countries. Among nutrition-specific interventions, one SR found that social and behavior change communication (SBCC) on nutrition and health practices via community and support groups was an effective intervention to improve stunted and linear growth (figure 2.3). SBCC interventions through other channels (such as education or promotion, growth monitoring and promotion, and home visits and peer support) offer less conclusive evidence. Within inter- World Bank Support to Reducing Child Undernutrition  Chapter 2 ventions targeting children, most of the evidence studied the effects of providing supplementary energy-dense foods, followed by zinc supplementation, supple- mentary feeding with micronutrient-rich food, and multiple micronutrients, and showed mixed results, yet with mostly positive findings. Among nutrition-sensitive interventions in the health sector, few SRs found that family planning and contraception services, through its effects on birth spacing, and institutional strengthening policies and health insurance can contribute to reducing stunted growth. Deworming campaigns targeting children and child stimulation interventions were found to have mixed results. Few nutrition-sensitive interventions in the agriculture sector seem to be effective in improving child growth, although the evidence remains limited. A meta-analysis found that consumption of biofortified quality pro- tein maize led to an increase in the rate of growth in weight and height in infants and young children with mild to moderate undernutrition. Also, a significant and positive effect of land reforms conferring or providing land rights and autonomy to women in agricultural production was observed on the 22 long-term nutritional status of women and child nutrition. The study revealed that a mother owning land halved the probability of her child being severely underweight. Home gardening, small-scale livestock production, and provision of agricultural inputs and training interventions are shown to have mixed results on improving stunted and physical growth. In the social protection sector, the provision of daycare services and the facilitation of access to microfinance, credit, and banking were found to have mixed results. Evidence on the effect of nutri- tion-sensitive interventions in WASH is rather limited. One SR found evidence suggestive of a small benefit of improving quality of water supply, identifying a borderline statistically significant effect on height-for-age z score in children under five years old. Provision of latrines and potties for safe disposal of feces (4 SRs) and SBCC delivered through WASH (1 SR) show mixed results. The global knowledge also highlights other interventions with consistent evidence of effectiveness to improve particular nutrition outcomes and determinants. Many of the most effective interventions target the mother, underscoring the importance of engaging women early (preconception) and across all stages of early child development. For instance, the provision of io- dine supplementation to women has consistently worked for improving child micronutrients status, the provision of energy-dense food increases child birthweight, supplementation with iron folate improves maternal nutrition status and micronutrient deficiencies, and SBCC is effective for improving breastfeeding practices and maternal mental health. Also, there is consistent evidence that deployment of community health workers (CHWs) is effective World Bank Group Independent Evaluation Group    23 for improving child use of health services, family planning and contraception services are effective to reduce birth spacing, and health system strength- ening support shows positive effects in improving complementary feeding practices and household welfare. In agriculture, food fortification with vita- min A improves children’s complementary feeding, small-scale aquaculture is an effective intervention for increasing household income resources, and the provision of agriculture inputs and training improves knowledge and attitudes. In the social protection sector, conditional cash transfers are the only intervention with consistent and positive evidence to improve house- hold access to nutrient-rich food, schooling, and knowledge and attitudes. In the WASH sector, provision of safe water storage is the only intervention showing strong evidence of effectiveness in reducing child enteric infection and diarrhea. Figure 2.3. Effective Interventions to Improve Stunted and Linear Growth Nutrition and dietary support interventions Target Stunted and linear growth Micronutrient supplement: iodine Children Micronutrient supplement: MMNs (including omega�3) Children Micronutrient supplement: vitamin A Children Micronutrient supplement: zinc Children Supplementary feeding with energy�dense food Children (lipid, protein) Supplementary feeding with micronutrient�rich food Children Micronutrient supplement: MMNs Women SBCC via community or support groups Household SBCC via education or promotion Household SBCC via home visits or peer support Household SBCC via growth monitoring and promotion Household Health interventions Target Stunted and linear growth Deworming (single or periodic) Children Family planning and contraception Women Provision of early child stimulation Household WASH interventions Target Stunted and linear growth Sanitation (latrines, potties, safe disposal, and so on) Household Water supply (such as community standpipe or hand pump) Household SBCC via promotion, home visits or peer support, or mass Household communication Agriculture interventions Target Stunted and linear growth Home gardens (with or without livestock) Household World Bank Support to Reducing Child Undernutrition  Chapter 2 Small�scale livestock production Household Fortification/biofortification: quality protein maize Household Policy on land property rights Household Provision of inputs and training Household Social protection interventions Target Stunted and linear growth Daycare services Household Access to microfinance, credit, or banking Household Health care interventions Target Stunted and linear growth Health care demand: health insurance Institutions Health care supply: system strengthening (training, BFHI, Institutions mHealth) Intervention study results Number of references 1–2 No effect + Positive Positive Positive or (<3 studies) (3+ studies) 3–5 Inconsistent (P, NE) 6–9 10+ Sources: Independent Evaluation Group; systematic review map. Note: The legend combines the size of the evidence (number of systematic reviews) and the direc- tion of the evidence (positive, no effect, or inconsistent). Positive indicates that the pooled effect (for meta-analyses) or all underlying studies (for narrative syntheses) of the intervention are found to have a positive effect on the outcome of interest. No effect indicates that the intervention is neither significantly positive nor significantly negative on the outcome of interest. Inconsistent indicates that for a narrative 24 synthesis, the evidence of a particular intervention on a specific outcome shows a mix of positive (P) and no effects (NE) across the underlying studies. Given the direction of the evidence, the dark- and medium-green legends indicate that the evidence of an intervention on a particular outcome is found to be positive more than three systematic reviews or in up to three systematic reviews, respectively. The light-green legend indicates that the pool of evidence of a particular intervention on a specific outcome shows a mix of positive effects no effect, or a combination of both (inconsistent) in narrative synthesis. The full list of interventions reviewed in the systematic review map for stunted and linear growth is shown in appendix B. BFHI = Baby-Friendly Hospital Initiative; MMN = multiple micronutrients; SBCC = social and behavior change communication. The World Bank largely supports nutrition interventions that are known to work. An assessment of the alignment between the portfolio interventions and the literature on what works covering 47 percent of the portfolio shows that the World Bank has focused on interventions that have positive evidence of effectiveness to improve the nutrition outcomes of interest (figure 2.4).4  lignment of Nutrition Interventions with Evidence on What Figure 2.4. A Works, by Intervention Area Sources: Independent Evaluation Group; systematic review map, and portfolio review and analysis. World Bank Group Independent Evaluation Group    25 Note: The legend follows the systematic review map. Positive indicates that the pooled effect (for meta-analyses) or all underlying studies (for narrative syntheses) of the intervention are found to have a positive effect on the outcome of interest. No effect indicates that the intervention is neither significantly positive nor significantly negative on the outcome of interest. Inconsistent indicates that for a narrative synthesis the evidence of a particular intervention on a specific outcome shows a mix of positive and no effects across the underlying studies. Negative indicates that the intervention is found to have a nega- tive effect on the outcome of interest. Given the direction of the evidence, the dark- and medium-green legends indicate that the evidence of an intervention on a particular outcome is found to be positive in more than three systematic reviews or in up to three systematic reviews, respectively. Similarly, the dark-red legend indicates that the evidence of an intervention on a particular outcome is found to be negative in more than two systematic reviews. The light-green legend indicates that the pool of evidence of a particular intervention on a specific outcome shows a mix of positive effects, no effect, or a combination of both (inconsistent) in narrative synthesis. IS = institutional strengthening; WASH = water, sanitation, and hygiene. In health, the World Bank concentrates on supporting SBCC on nutrition and health practices known to work across different nutrition-relevant outcome ar- eas, including breastfeeding and complementary feeding, and child use of health care services. Other health interventions where the World Bank highly aligns with the literature are supporting health care approaches that implement health facilities outreach activities, the deployment of CHWs, and family planning and contraception services. Consistent with the findings of the portfolio review, the World Bank largely focuses on institutional strengthening support to improve the health system, expand health insurance, and implement performance-based financing and service integration approaches that the global evidence base shows to be effective for improving particular nutrition-relevant outcomes (such as use of health care services, knowledge and attitudes, complementary feeding, child health, stunted growth, and child cognitive development). The most frequent agriculture intervention supported by the World Bank’s nutrition portfolio is the provision of inputs and training. Biofortification of foods and the support for small-scale livestock production are also promi- nent in the portfolio within the group of interventions with consistent posi- tive evidence of effectiveness. Within social protection interventions, the World Bank aligns with evidence on what works by mainly focusing on countries’ cash transfer programs. Cash transfer programs have positive effects in improving households’ food security and welfare, schooling attendance, health care seeking, and child health and nutrition dietary practices. Support for access to center- or home- based care services, also supported by the World Bank, has been shown to be World Bank Support to Reducing Child Undernutrition  Chapter 2 effective to improve complementary feeding and child health outcomes. The portfolio includes effective WASH interventions, such as SBCC and commu- nity water supply. According to the literature on what works, effective interven- tions in the portfolio are SBCC to promote hand washing and safe drinking water, community water supply through standpipes or hand pumps, safe water storage, and provision of soap. These interventions have consistent evidence of effec- tiveness in improving access to safe water, improving household knowledge and attitudes, or reducing the incidence of childhood illness and diarrhea. Although World Bank interventions align well with global knowledge in many areas, more attention might be directed at particular interventions where evi- dence is consistently positive across a broad set of nutrition-relevant outcomes areas, such as energy-dense food supplements for women and micronutrient-rich food supplements for children. The SRM identifies interventions with broad positive evidence of effectiveness across multiple nutrition-relevant outcomes 26 areas (table 2.1). Although the World Bank emphasizes many of these interven- tions, some of them may not be receiving sufficient attention given their poten- tial benefits. Among nutrition-specific interventions, few projects in the HNP portfolio include women’s supplementary feeding with energy-dense food and children supplementary feeding with micronutrient-rich food. Within nutri- tion-sensitive interventions, vitamin A biofortification of foods in the agriculture portfolio and provision of soap to stimulate hygiene and sanitation practices in the WASH portfolio have received little attention. Furthermore, two interven- tions with broad positive impacts remain unexplored in the nutrition portfolio. The first refers to maternal emotional support interventions for which the global evidence suggests that they are effective in improving breastfeeding and parent- ing practices, women’s mental health, and use of health care services. The second intervention is land property right reforms that could be implemented through governance, macroeconomics, or the agriculture sector. Such reforms can be ef- fective in improving household welfare (consumption and income), empowering women (increased control of resources), reducing micronutrient deficiencies of women, and even stunted growth. Impact evaluations of World Bank projects contribute to increasing knowl- edge of what works by supporting evidence-based learning to design and im- prove nutrition interventions in operations. A review of the advisory services and analytics (ASA) portfolio in case study countries shows that even when the World Bank works in a small geographical area, impact evaluations facil- itate the mainstreaming of interventions or experiences leveraged from the World Bank Group Independent Evaluation Group    27 project support. In this way impact evaluations can support the institution- alization of interventions in the country’s own program. Some countries, like Madagascar, have given more consistent attention to evidence learning over a decade and have been using evidence to improve and strengthen the roll- out of nutrition interventions of the community-based nutrition (CBN) pro- gram. Impact evaluations on CBN programs have been important in Ethiopia, Malawi, Nicaragua, and Rwanda. In social sectors, some countries (Indonesia, Madagascar, Malawi, Nicaragua, Niger, and Rwanda) are using impact eval- uations to improve the design of interventions, specifically the links among community block grants or cash transfers and behavior nudges to improve the demand for maternal and child health services, parenting behaviors pro- grams, or child feeding practices. Impact evaluations also support learning on ECD programs in some countries (Ethiopia, Indonesia, Madagascar, Niger, and Rwanda) to integrate nutrition interventions across social sectors.  ystematic Review Map Interventions with a Broad Positive Table 2.1. S Impact Intervention Types Interventions (%) Projects (no.) Nutrition-specific Child supplementary feeding with 0.5 8 micronutrient-rich foods Maternal supplementary feeding with 0.3 4 energy-dense foods Women micronutrient supplementation: 0.9 15 iron folate (iron–folic acid) SBCC of nutrition and health promotion 21.5 107 (via community and groups, education, growth monitoring and promotion, home visits, mass communication, and IPC at health facility) Nutrition-sensitive Health Health system strengthening 8.8 101 Maternal emotional support 0.0 0 Family planning and contraception 2.1 32 World Bank Support to Reducing Child Undernutrition  Chapter 2 Health care approach: CHWs 0.8 11 Health facility community outreach 0.7 11 E-health communication 0.0 0 Health insurance 0.7 11 Agriculture Provision of agriculture inputs and 2.3 34 training Small-scale livestock 2.1 30 Vitamin A fortification 0.5 7 Land property rights 0.0 0 Social protection CCTs 2.2 32 WASH Provision of soap 0.2 4 Total interventions with a broad positive 43.6 impact 28 Sources: Independent Evaluation Group; systematic review map and portfolio review and analysis. Note: CCT = conditional cash transfer; CHW = community health worker; IPC = interpersonal communica- tion; SBCC = social and behavior change communication; WASH = water, sanitation, and hygiene. Trust funds and partnerships catalyze innovation and the adoption of novel approaches where learning is important to support expansion. For example, in Madagascar, the Knowledge for Change program and the Health Results Innovation Trust Fund provided critical support for impact evaluation and other operational learning activities to adaptively improve the CBN program, including for human-centered design learning to improve the effectiveness of interventions. In Rwanda, the Bill & Melinda Gates Foundation supports evidence-based learning on the national behavior change strategy (led by the Mind, Behavior, and Development Unit), which could help rethink behav- ior change interventions for nutrition. Multisectorality of the Nutrition Portfolio: Engagement across Global Practices The World Bank’s nutrition portfolio is multisectoral in that it engages dif- ferent GPs to implement interventions toward nutrition determinants. The portfolio is also multidimensional in that it includes a range of different in- terventions across the nutrition-specific and nutrition-sensitive dimensions of the conceptual framework. HNP leads most projects in the nutrition port- folio (42 percent), Agriculture leads about 21 percent, and SPJ leads 20 per- cent. Over time, the roles of Agriculture and SPJ have grown to account for about half of the active nutrition portfolio, but projects led by Water, Edu- World Bank Group Independent Evaluation Group    29 cation, and other GPs remain small (figure 2.5). Nutrition interventions are implemented by a combination of core projects that have a heavy focus on nutrition and noncore projects, which integrate nutrition interventions in their components. Increasingly, the portfolio has included more multidimensional projects that support a range of nutrition interventions. In these projects, GPs have inte- grated interventions that inherently belong to other sectors to work across silos to tackle nutrition determinants more comprehensively. These projects may be core nutrition projects that have a heavy focus on nutrition, with nu- trition explicit in the objectives or in the title.5 Interventions related to diet and breastfeeding, WASH, safety nets, health, agriculture, and institutional strengthening have been integrated across projects in all GPs (figure 2.6). Interventions related to social norms have been emerging across GPs, and ECD is increasingly being integrated in SPJ, Agriculture, and HNP operations. The emphasis on multidimensional projects is strongest in SPJ, Education, and other GPs, such as Macroeconomics, Trade, and Investment (figure 2.7). SPJ has integrated interventions across all dimensions of the conceptual framework as part of its support to lower-income households. Education has integrated interventions on diet and breastfeeding, child disease prevention (such as deworming), and WASH (such as SBCC). Figure 2.5. Projects by Approval Period and Global Practice Share of projects (percent) World Bank Support to Reducing Child Undernutrition  Chapter 2 Sources: Independent Evaluation Group; portfolio review and analysis. Note: Data are presented by fiscal years. Other Global Practices include Macroeconomics, Trade, and Investment; Social Sustainability and Inclusion; Urban, Disaster Risk Management, Resilience, and Land; and Governance. 30 Figure 2.6. Interventions in Projects by Global Practice Sources: Independent Evaluation Group; portfolio review and analysis. Note: WASH = water, sanitation, and hygiene. World Bank Group Independent Evaluation Group    31 Figure 2.7. Projects by Global Practice and Degree of Multidimensionality Source: Independent Evaluation Group. World Bank Support to Reducing Child Undernutrition  Chapter 2 Note: The multidimensionality score is the sum of the number of nutrition-specific and nutrition-sensitive intervention areas in a project divided by the total possible number of interventions. Noncore nutrition projects integrate sector-related nutrition interventions in their components. These projects often focus on a few interventions, but do not have an explicit focus on nutrition. For example, HNP projects focus on health and family planning interventions, Water projects on WASH interven- tions (such as SBCC and latrines), Agriculture projects on agriculture and food approaches (such as fortified crops, home gardens, livestock and poultry, and seasonal food access), and SPJ projects on safety nets. In case study countries, interventions that have been integrated in Agriculture and Water projects lacked an intentional design to target improvements in nutrition determi- nants, such as access to nutritious foods or hygiene and sanitation practices of households with children. Some countries (Ethiopia, Madagascar, Malawi, Nicaragua, Niger, and Rwanda) are using evaluation evidence to improve the design of integrated nutrition interventions in social protection and ECD. 32 Multidimensionality in Country Programs A World Bank country portfolio with a mix of nutrition-specific and nutrition-sensitive interventions and institutional strengthening provides a pathway to improve nutrition determinants and contribute to outcomes. Key for the country portfolio is that it successfully supports a mix of inter- ventions toward nutrition determinants and institutional strengthening to contribute to outcomes, in collaboration with other partners. About half of the countries have both multidimensional portfolios, with a mix of in- terventions, and medium-to-high support for institutional strengthening. (Figure D.7 in appendix D shows countries by the multidimensionality of their portfolio and the share of institutional strengthening support.) Coun- tries where the World Bank portfolio has had few interventions and low support for institutional strengthening, such as Burkina Faso and Sierra Leone, stand out as candidates to improve nutrition support. Among coun- tries where the World Bank’s portfolio has had high multidimensionality and medium-to-high institutional strengthening are the Comoros, Côte d’Ivoire, Haiti, India, Indonesia, Malawi, Madagascar, Nicaragua, Pakistan, Rwanda, and Senegal. Some of these countries, however, have newer investments, such as Côte d’Ivoire, Pakistan, and Rwanda. The success of the institutional strengthening in these countries will be important to support results toward nutrition. The country portfolios of fragile and conflict-affected situation countries on average have a slightly lower multidimensionality than other World Bank Group Independent Evaluation Group    33 countries. This is likely due to the implementation challenges in fragile and conflict-affected situations. In the case studies, country portfolios show a continuum of support to nu- trition led by different GPs and instrument types (investment project financ- ing, development policy loan, Program-for-Results, RETF) over the 10-year evaluation period. However, in most countries these interventions are frag- mented across projects and time, and coordination to ensure support to all relevant nutrition determinants is limited. Figure 2.8 presents the timeline of the Ethiopia portfolio, which since 2008 has had a series of HNP projects to support the national nutrition program, expanding health services, and a package of community-based interventions. Other projects have supported safety nets, nutritious food, and WASH. Ethiopia stands out for its increasing emphasis on multisector and partner coordination of nutrition efforts. In Malawi, the Nutrition and HIV/AIDS Project (P125237; FY12–19) has sup- ported development of a package of CBN interventions with other donors. Other GPs with relevant support have included Agriculture, SPJ, Water, and Urban, Disaster Risk Management, Resilience, and Land, but health services support was lacking. In Mozambique, the main support has been through HNP projects to health services and a CBN intervention package—with limited coordination with projects in other sectors. In Nicaragua, HNP and SPJ projects have coordinated support to community and family health care services, including ECD and adolescent health support linked to a social wel- fare model, focused on children. Projects in Water and Agriculture separately supported interventions. In Niger, the main support is through HNP and SPJ projects to health services, women and girls’ empowerment, and safety nets, with some recent support to WASH and ECD. In Rwanda, projects in Water, SPJ, Agriculture, and Macroeconomics, Trade, and Investment helped im- prove decentralized access to health, water, safety nets, and food. Since FY17, the country portfolio has emphasized coordinated GP projects (HNP, SPJ, Agriculture, Education, and Macroeconomics, Trade, and Investment). See appendix G for all country examples. World Bank Support to Reducing Child Undernutrition  Chapter 2 Many nutrition interventions in country portfolios are emerging and need further support and collaboration with partners to be institutionalized in country systems. In Indonesia, the community-driven development ap- proach is well established, but support to converge services across sectors is newer. In Ethiopia, support to ECD, maternal diet intake, women’s empow- erment, and adolescent nutrition is emerging. In Madagascar, although the community package supporting nutrition has been developed over many years, support to access nutritious food and WASH is less developed. In Malawi, the duration of World Bank support to develop community inter- ventions has been limited, and other partners have also provided support. In Mozambique, support to nutritious food, social norms, and WASH has received limited attention. In Nicaragua, support to develop child feeding and caregiving is ongoing. In Niger, there has been limited support to devel- op community-based interventions to reach the large rural populations and ensure access to nutritious food. In Rwanda, support to develop a package of nutrition-related intervention to reach communities, including ECD, mater- 34 nal health, home gardens, safety nets, and other support, is ongoing. Figure 2.8. Ethiopia Project Time for World Bank Nutrition Support 51 50 Childhood stunting (percent under 5) AFR average 44 33% (2018) 40 WASH SIL 40 Protection of Basic 38 Food security Productive Services Projects Series 30 SIL Safety Nets World Bank Young Women’s APL Series Health P4R lending projects Livelihood JSDF Lowlands Livelihoods IPF 20 Nutrition SIL WASH IPF Social Protection P4R One WASH IPF Pastoral Community Nutrition JSDF Development IPF Education P4R 10 Livestock and Agriculture IPF Series Fisheries IPF 0 2005 2010 2015 2020 SCD identifies Small area Combating nutrition Investing in the estimation Maintaining World Bank malnutrition in Ethiopia constraints early years Momentum analytics (WASH) Nutrition project Maternal and child Stunting All Health P4R project operations research health inequalities reduction hands operations research in SSA on deck Government National nutrition strategy, first national initiatives program, and establishment of Second national nutrition program, National food and coordination body, health extension Seqota Declaration, second growth and nutrition policy program, community-based nutrition transformation plan, CINUS Sources: Independent Evaluation Group; UNICEF, WHO, and World Bank 2019. Note: The box colors in the chart indicate the World Bank Global Practice responsible for the lending: brown = Social Protection and Jobs; gray = Water; green = Agri- culture; dark blue = Health, Nutrition, and Population; light blue = Education. AFR = Africa; APL = adaptable program loan; CINUS = Comprehensive Integrated Nutrition Services; IPF = investment project financing; JSDF = Japan Social Development Fund; P4R = Program-for-Results; SCD = Systematic Country Diagnostic; SIL = sector investment loan; SSA = Sub-Saharan Africa; WASH = water, sanitation, and hygiene. World Bank Group Independent Evaluation Group    35 Strengthening institutional capacities (stakeholder engagement, policy, and service delivery) is important to improve nutrition support in countries. In case studies, stakeholder engagement and ownership has included strengthen- ing leadership, knowledge, and participatory roles of networks of community volunteers, local leaders, farmers, nongovernmental organizations, and other local actors in SBCC, the implementation of interventions, results monitoring, and other approaches. In Indonesia, Rwanda, and Senegal, leadership building has been at all levels and across sectors, from the president to ministries, dis- tricts, and communities, and this has helped improve the accountability of nu- trition support. Strengthening of policy, financing, and coordination has included support to policy dialogue and strategies. Strengthening of service delivery has included support to design basic services and to build knowledge of service providers, monitoring and evaluation (M&E), and supervision. Experience points to a need for institutional strengthening of multisectoral arrangements for nutrition in countries to improve stakeholder engagement, policy, and services. Most institutional strengthening in case study countries has been in one sector (for example, to develop agriculture or health ser- vices), with emerging examples of how projects can strengthen multisectoral arrangements for nutrition. The stocktaking analysis of countries identified World Bank Support to Reducing Child Undernutrition  Chapter 2 factors that have facilitated multisectoral coordination efforts and results for nutrition in countries (appendix H; box 2.2). In Ethiopia, Indonesia, Malawi, Nepal, Rwanda, and Senegal, support to multisectoral nutrition coordination, strategies, planning, and financing at the national and decentralized levels has been key. However, the continuity of this support across projects is a challenge. In some countries, the World Bank has supported multisectoral arrangements for M&E. In Senegal, the World Bank has facilitated the M&E of the nutrition program in communities. Activities in Peru have helped build capacity for the social monitoring of nutrition results. Indonesia and Rwanda are improving the accountability and convergence of service deliv- ery by initiating village scorecards and child scorecards and developing the interoperability of sectoral M&E systems. In some countries (Ethiopia, Indo- nesia, Madagascar, Malawi, Mozambique, Rwanda, and Senegal), the World Bank has ongoing support to develop integrated nutrition intervention packages (integrating interventions from health, social protection, education, agri- culture). In Indonesia, Malawi, Rwanda, and Senegal, the World Bank has 36 supported multisectoral communication strategies to align nutrition messages across different sectors and actors involved in nutrition. Overall, more intentional planning of nutrition support (financing and ASA) is needed in the country portfolio and for multisector implementation with- in country portfolios to support nutrition determinants and institutional strengthening. This is already initiated in Ethiopia, Indonesia, and Rwanda to coordinate the implementation of World Bank support across GPs and projects and to synergize efforts with other partners. Although the strategies of all countries addressed nutrition in some way, most did not identify how different instruments collectively addressed nutrition needs in the country context. Moreover, country experiences point to the importance of better aligning World Bank support to strengthen multisectoral nutrition coordina- tion and local government and communities to deliver a multidimensional package of nutrition interventions. Box 2.2. Factors That Facilitate Multisectoral Coordination Efforts » Consistency of national leadership regarding a mandated program or framework to coordinate actors and roles of relevant sectoral ministries » Developed role of subnational government to coordinate multisectoral actions » Organization of sectoral extension services and community actors to deliver an inte- World Bank Group Independent Evaluation Group    37 grated package of interventions tailored to local needs, with consistent messaging » Strengthened financing and planning, monitoring and evaluation, and knowledge sharing approaches that support multisectoral interoperability of decisions, ac- tions, and learning (rather than single-sector systems) on nutrition interventions at different levels of implementation, horizontally and vertically Source: Independent Evaluation Group. The challenge is to coordinate the delivery of nutrition interventions by sectors—social, agriculture, and WASH—considering their different priorities and target groups in communities. Health interventions often target women and children in communities with low nutrition indicators, but coverage of remote areas is a challenge. Safety net and ECD interventions increasingly have coordinated with health interventions by focusing on lower-income households in the same communities, as in Nicaragua and Rwanda. However, agriculture interventions tend to target farmers and geographies important to the food supply, and WASH interventions are often in towns. In particular, food and agriculture approaches and social services (health, social protec- tion, education) support have often not coordinated support in the same communities (for example, Ethiopia, Madagascar, and Mozambique). This situation has likely limited the possibility of the country program to support results unless another partner is providing the relevant interventions in oth- er geographies. Accordingly, recent approaches in Indonesia and Rwanda to improve the coordinated implementation of interventions are learning how to converge interventions at the community level. These countries also have technical assistance to help coordinate nutrition support across projects and partners nationally and in districts and communities. Having multidimensional projects, which support a range of nutrition interventions in communities, and coordinating nutrition interventions led by different GPs are options to improve nutrition support in countries. The community level provides a platform where a project can support the World Bank Support to Reducing Child Undernutrition  Chapter 2 delivery of a multidimensional package of interventions (as in Madagascar and Malawi). Another option is the use of multisectoral nutrition action plans as internal coordination tools to improve synergy across portfolios with projects led by different GPs under the leadership of the World Bank country manager or director, such as in Indonesia, Rwanda, Vietnam, Papua New Guinea, and Ethiopia. (The countries listed are examples, and the list may not be exhaustive.) The evaluation did not examine the relative cost-effectiveness of coordinating sectoral support compared with hav- ing multidimensional projects with a mix of intervention. Also important is the consistency, quality, coverage, and expansion strategies to support interventions in communities—for example, Ethiopia and Madagascar have had multiple World Bank projects to help design nutrition intervention packages and institutionalize them in national programs over years. This consistent timeline to develop a quality package of interventions has been lacking in most other countries. 38 Are Interventions Supported by the World Bank Based on Country Needs? The evaluation confirms the logic of the conceptual framework, which guides the World Bank’s nutrition agenda. The heat map analysis assesses the countries’ access to the nutrition determinants and their empirical links with nutrition outcomes based on the conceptual framework and existing work on the drivers of undernutrition (Skoufias, Vinha, and Sato 2019). The heat map is based on cross-country data from the Joint Child Malnutrition esti- mates (UNICEF, WHO, and World Bank 2019) during the evaluation period for the 64 countries of the portfolio for which indicators related to food and care, access to health and WASH services, social norms determinants, and nutrition outcomes data were available. Principal components analysis has been used to construct composite measures for each of the determinants and overall nutrition outcomes based on selected indicators.6 The countries’ conditions in nutrition determinants matter for achieving better nutrition outcomes, reinforcing the importance of having synergized support across determinants to improve outcomes. Correlation analysis shows that countries that are better off in terms of food and care, access to WASH and health services, and social norms determinants tend to have bet- ter nutrition outcomes (no stunted growth, no wasting, no underweight, no anemia, and no LBW) at the beginning and at the end of the evaluation peri- World Bank Group Independent Evaluation Group    39 od (figure 2.9, panel a). The link between health determinants and outcomes is the strongest across all nutrition outcomes, followed by social norms, WASH, and food and care, which reinforces the importance of having inter- ventions in health synergized with multidimensional interventions across determinants to improve outcomes.7 This synergized support has not con- sistently happened in any of the case study countries. For example, although Malawi had a strong emphasis on community interventions that addressed a range of determinants, support to health services was largely absent in the portfolio. In Mozambique and Niger, by contrast, support in health has not been consistently synergized with support to other determinants. Intentional planning of World Bank support in countries to address needs re- lated to disadvantaged determinants (low levels of food and care, WASH, and health services) can help countries to catch up to improve outcomes. Cor- relation analysis suggests that countries at the bottom of the distribution in nutrition determinants at the beginning of the period are slowly converging in nutrition outcomes, with improvements in these determinants, and thus have potential to catch up over time (figure 2.9, panel b). These results are encouraging and suggest that the inequality in nutrition outcomes among countries could decrease with more intentional support to improve deter- minants. Conversely, improvements in nutrition outcomes to benefit vul- nerable populations may be slower, with countries taking longer to achieve adequate levels of determinants. The nutrition portfolio indeed has focused on low-income countries with high rates of stunted growth. The interventions of the nutrition portfolio align well with country needs at the national level, but there is room to strengthen support to nutrition de- terminants, particularly with regard to social norms.8 A mapping exercise us- ing portfolio review data on interventions addressing nutrition determinants (food and care, WASH and health services, and social norms) shows that about 79 percent of the interventions in the portfolio align with the country needs, suggesting that the World Bank supports the right areas of interven- tion. The World Bank support has been especially relevant in addressing World Bank Support to Reducing Child Undernutrition  Chapter 2 needs related to food and care (appropriate alignment in 95 percent of the cases), and access to health services (90 percent), which has the strongest as- sociation with country nutrition outcomes according to the heat map analy- sis. However, needs related to areas such as access to WASH and social norms have often not been addressed by interventions (64 percent and 52 percent, respectively; figure 2.10). Particular areas where the World Bank emphasis on social norms is thin include women’s empowerment, early marriage, and pregnancy, which currently account for only 6 percent of the portfolio across all GPs. Moreover, case studies suggest that support to address needs related to nutrition determinants has been inconsistent in countries over the 10- year evaluation period. 40 Figure 2.9. Undernutrition Determinants and Nutrition Outcomes a. Overall composite of four determinants (2008) and composite measure of outcomes (2018) 4 Five outcomes (2018) 2 0 �2 �4 �2 �1 0 1 2 Four determinants (2008) AFR ECA MNA EAP LAC SAR b. Overall composite of four determinants (2008) and composite measure of outcomes (trend 2008–18) 2 Trend of five outcomes 0 World Bank Group Independent Evaluation Group    41 �2 �4 �2 �1 0 1 2 Four determinants (2008) AFR ECA MNA EAP LAC SAR Sources: Independent Evaluation Group; heat map analysis. Note: Composite measures are based on principal component analysis. Nutrition outcomes include stunted growth, wasting, underweight, anemia, and low birthweight. Determinants include food and care; health services; water, sanitation, and hygiene services; and social norms indicators. AFR = Africa; EAP = East Asia and Pacific; ECA = Europe and Central Asia; LAC = Latin America and the Caribbean; MNA = Middle East and North Africa; SAR = South Asia. Figure 2.10. Alignment of Portfolio Interventions with Country Needs 100 Food and care Health 80 Matching score WASH 60 Social norms 40 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Correlation of determinants and nutrition outcomes Sources: Independent Evaluation Group; heat map and portfolio review and analysis. Note: Matching score represents the degree of alignment of the portfolio interventions with the country needs in food and care; health; water, sanitation, and hygiene services; and social norms. WASH = water, sanitation, and hygiene. In addressing needs, the key is strengthening the World Bank’s within- country alignment across sectors and targeting relevant interventions to address disaggregated needs in particular geographies and populations. Proj- World Bank Support to Reducing Child Undernutrition  Chapter 2 ects in different GPs are often implemented in different geographical areas and for different target groups and have lacked mechanisms to integrate or converge actions or build on achievements to improve nutrition outcomes in the same communities. Similarly, behavior change interventions are frag- mented (across GPs, projects, timelines, and geography). To meaningfully improve outcomes, all priority needs should be addressed across targeted communities, given the synergistic nature of determinants. In Nicaragua, simultaneous support has been provided to needs in health, social protec- tion, water, and agriculture only in one region of the country with vulnerable groups (Jinotega between FY11 and FY17). In Mozambique, health support has focused on the northern provinces with high rates of stunted growth. Although there has been some coordination with health on biofortification, most agriculture support focused on emergency food distribution and did not synergize with health interventions. Similarly, behavior change interven- tions to promote health, WASH, caregiving, and nutritious foods have been supported by different projects and implemented in different communities. 42 1 The identification of the relevant nutrition portfolio has focused on countries with rates of stunted growth at or above 20 percent at any point in time of the evaluation period (see port- folio identification strategy in appendix D). 2  See appendix D on nutrition portfolio for more details. 3  See appendix E for a detailed analysis of behavior change interventions. 4  The alignment analysis is based on the systematic review map’s interventions that are also found in the nutrition portfolio in the areas of nutrition, health, social protection, water, agriculture, and institutional strengthening in the health sector for which there is existing evidence of their effectiveness. Twelve out of 84 interventions types of the systematic review map are not found in the nutrition portfolio. See appendix B for details on the scope of the alignment analysis. 5  Core nutrition projects are those that have nutri or stunt in their title or in their project development objectives and have a nutrition content share equal to or above the top two quintiles of the distribution (top 40 percent). 6  See appendix F on the heat map for more details and full correlation analysis. 7  These findings are consistent with those shown by Skoufias, Vinha, and Sato (2019) based on logit model estimates using 33 recent Demographic and Health Surveys from Sub- Saharan Africa. 8  For the purpose of the matching exercise, country need for a particular determinant (such World Bank Group Independent Evaluation Group    43 as food and care) is defined as any of its comprised indicators (such as minimum dietary diversity of children ages 6–23 months) falling below their corresponding threshold that has been established by the literature, when available, or falling in the bottom 50 percent of the distribution at the beginning of the evaluation period. This means, for example, that a country with a minimum dietary diversity index below the threshold has an inadequate level of food and care determinant for which it would be desirable to find interventions in the area of food and care in the World Bank nutrition portfolio matching this need. 3 | World Bank Contribution to Nutrition Results Highlights The World Bank’s nutrition portfolio has improved its overall per- formance over time, and support to institutional strengthening and underlying determinants shows better results than support to im- mediate determinants. The adequacy of the enabling environment underlies the potential for a country to improve the determinants of nutrition, which in turn can improve a country’s nutrition outcomes for mother and child. Community-based programs contribute to behavior changes that improve nutrition determinants. However, achieving sustained behavior change is challenging, and measurement of progress along the results chain is weak, undermining effective planning and evaluation to support behavior change. World Bank contributions to behaviors related to social norms are modest. The measurement of nutrition results for projects exhibits per- sistent gaps, especially in tracking expected achievements from nutrition-specific and social norms interventions requiring behavior changes. Expected results from nutrition-sensitive interventions are most frequently measured, especially for those projects tar- geting health and family planning services, social safety nets, and agriculture and food systems. Improving project performance in achieving nutrition results re- quires adequate monitoring and evaluation frameworks, sustained community-based implementation, strong government commit- ment and institutional capacity to support project activities, and a project design that intentionally integrates nutrition interventions 44   that aim to improve nutrition determinants. Nutrition Results: Project Performance and World Bank Contributions The World Bank’s nutrition portfolio overwhelmingly aims to improve the underlying determinants of nutrition and institutional strengthening in countries. This aim is consistent with the distribution of interventions that tend to concentrate at the foundations of the conceptual framework dis- cussed in chapter 2 (figure 3.1). Social norms (women’s empowerment and early marriage and pregnancy) and to a lesser extent nutrition outcomes (anthropometric measurements, micronutrient status, and cognitive devel- opment) and immediate determinants (child feeding practices, diet diversity, and maternal and child health) are rarely part of the projects’ objectives, ac- counting for 2 percent, 13 percent, and 18 percent, respectively. This tenden- cy to focus on nutrition determinants is not surprising given the relatively short duration of investment projects and the longer-term nature of nutri- tion outcomes and social norms. The World Bank’s nutrition portfolio has good overall performance, but support to institutional strengthening and underlying determinants shows better results than support to immediate determinants that are more chal- lenging to achieve. Project performance is measured by the achievement rates of results framework indicators for 131 closed projects that had rele- vant nutrition-related indicators across the dimensions of the conceptual World Bank Group Independent Evaluation Group    45 framework. Overall portfolio performance is about 70 percent, which also de- pends on the adequacy of the results framework in terms of quality of indi- cators and the ambitiousness of indicators targets. Information from country case study evidence, impact evaluations, and regression analysis have also been used to assess results achieved by the portfolio. Contributions to Institutional Strengthening The World Bank contributes to institutional strengthening, but there is lim- ited evidence on the continued application of knowledge gained by actors involved in training and other support to sustain change. Project performance in strengthening institutions has improved over time (figure 3.1). Overall portfolio investments in policy, service delivery, and stakeholder engagement have achieved 79 percent of the expected results. Moreover, multivariate re- gression analysis of the portfolio offers some evidence that successful results in institutional strengthening, and in particular policy, financing, and coordi- nation, are associated with better achievement of nutrition outcomes and its determinants at the project level.1 In addition, the World Bank’s contribution toward stakeholder engagement (strengthening knowledge and participatory roles of networks of community volunteers, local leaders, farmers, nongovern- mental organizations, and other local actors in SBCC) is also highlighted as a major factor behind project performance.2 Further, institutional strengthening emphasizes the success of the World Bank in engaging actors (90 percent), es- pecially service providers; in improving their knowledge (83 percent); and, to a lesser extent, in applying the acquired behavior (71 percent). However, there is limited tracking of these indicators beyond the level of engaging actors and learning, and no evidence is available on whether institutional strengthening has supported sustained behavior change, such as of frontline workers to con- sistently apply skills to deliver services. Figure 3.1. Portfolio Performance a. Disaggregated achievement rates of nutrition indicators World Bank Support to Reducing Child Undernutrition  Chapter 3 b. Achievement rates of nutrition indicators by time period 46 b. Achievement rates of nutrition indicators by time period Sources: Independent Evaluation Group; portfolio review and analysis. Note: WASH = water, sanitation, and hygiene. World Bank Group Independent Evaluation Group    47 In-depth analysis in country case studies provides good examples of the World Bank’s contribution to strengthening institutions at national and sub- national levels. Successful examples of institutional strengthening (through both ASA and lending) have included policy dialogue, leadership building, South-South knowledge exchange, evidence-based learning, support to M&E systems, and support to districts to oversee nutrition, use M&E, and strength- en extension services and community groups. A key variation across countries has been in the extent of support to policy and coordination relative to service delivery. At the national level, the World Bank has supported high-level lead- ership; coordination of nutrition, policies, financing, and strategies; and M&E systems, diagnostics, and research and evaluation. At the district level, it has supported learning, M&E, and supervision to oversee nutrition services. At the community level, the World Bank has strengthened the targeting of services, community groups, and extension workers (box 3.1). Box 3.1. Contributions to Institutional Strengthening in 12 Countries Policy, Financing, and Coordination » Multisectoral coordination, strategies, financing, and planning. In Ethiopia, Indonesia, Malawi, Nepal, Rwanda, and Senegal, the World Bank has strength- ened national nutrition coordination, policy dialogue, strategies, and planning. In Madagascar and Mozambique, however, the lack of continuity of this support across projects has limited the institutional strengthening of nutrition coordination capacities. In Senegal, the World Bank’s Nutrition Enhancement Program has improved coordination efforts by identifying areas of collaboration among sector ministries, with an emphasis on the delivery of multidimensional nutrition services in communities. In Indonesia and Rwanda, the World Bank has strengthened the multisectoral nutrition strategy, including district-level plans, and a communica- tion strategy through a range diagnostic work on the nutrition situation, financing, and policy options. For example, the development of the National Strategy for Stunting Reduction in Indonesia and the reforms to develop the interoperability of social sector information systems in Rwanda have been catalyzed through South- South knowledge sharing supported by the World Bank on Peru’s experience in combating undernutrition. World Bank Support to Reducing Child Undernutrition  Chapter 3 Nutrition Service Delivery » Strengthening decentralized and community-level interventions. In Nicaragua, the World Bank has strengthened the supervision and management capacities of local governments for the decentralized delivery of a multisector package of social ser- vices. In Madagascar, the World Bank has contributed to the refinement of communi- ty-based nutrition services through years of analytic work and advocacy. In Rwanda, the World Bank is helping in the development of community services, which engage community health workers, and the convergence of nutrition-sensitive interven- tions in social protection, early childhood development, and agriculture. In many countries (Ethiopia, Indonesia, Madagascar, Malawi, Mozambique, Peru, Rwan- da, and Senegal), the World Bank has contributed to strengthen an integrated multidimensional package of community-based nutrition interventions. However, learning to organize sectoral and community actors to integrate the delivery of services remains a challenge. 48 (continued)  ontributions to Institutional Strengthening in 12 Countries Box 3.1. C (cont.) » Monitoring and evaluation improvements. World Bank interventions in Ethiopia have helped develop nutrition surveillance capacity and geographic data. Sup- port in Senegal has facilitated the measurement of multidimensional nutrition programs in communities. Activities in Peru have helped build capacity for the social monitoring of nutrition results. Indonesia and Rwanda are improving the accountability of service delivery by initiating village scorecard and child scorecards, which are planned to become an input into the formal management information system. Stakeholder Engagement and Ownership » Leadership building and stakeholder mobilization. In many countries, the en- gagement of government actors at all levels in nutrition strategy and planning has been instrumental for building leadership and government commitment. In Rwanda, Indonesia, and Senegal, the World Bank has supported high-level leadership and local leadership on nutrition. In Rwanda, the World Bank supports the monitoring of Imihigo, which is a contract between the president and local government leaders on achieving targets for key programs. In Indonesia, stunting summits are used to secure and sustain political leadership at national, provin- cial, district, and village levels, and provide a cascading system of accountability. Moreover, in most case study countries, the World Bank has strengthened social World Bank Group Independent Evaluation Group    49 and behavior change communication to raise awareness and shape social norms at the community level. These ranged from awareness or advocacy campaigns (Bangladesh, Nicaragua, and Peru) to more intensive social mobilization programs (Madagascar, Malawi, and Senegal). Often these activities have involved multiple sectors and types of actors in communities. Source: Independent Evaluation Group. Contributions to the Immediate and Underlying Determinants of Nutrition The World Bank makes important contributions in determinants of nutri- tion—with performance improving for underlying determinants and slightly declining for immediate determinants. There is more evidence of results for underlying determinants because the World Bank has invested more in nutri- tion-sensitive interventions over time and measurement is better. The perfor- mance of projects in achieving underlying determinants results has slightly improved over the evaluation period, and the most successful area has been agriculture and food, which has included improvements in food supply and production (vegetables, legumes, dairy products, livestock) productivity, mar- ket access, and food storage and transformation. But project achievements in immediate determinants of nutrition resulting from nutrition-specific inter- ventions have slightly declined more recently. Although achieved results in adolescent health appear to be remarkable, that is due to a very small sample of indicators. As we have shown before, the World Bank’s attention to adolescent health interventions is limited and has considerable measure- ment gaps. World Bank Support to Reducing Child Undernutrition  Chapter 3 The World Bank has contributed to improving nutrition determinants in all case study countries. In terms of food and care, in Ethiopia, Madagascar, and Nicaragua this has included improvements in breastfeeding, child feeding, and diet. However, in most case study countries, feeding and dietary improve- ments are modest. Through SPJ, the World Bank has contributed in some countries to improved food consumption, parenting skills, access to health services, livelihoods, and school enrollment among lower-income households. Through Agriculture, the World Bank has improved seasonal availability of food and crops (in Ethiopia, Malawi, and Rwanda), and biofortification (in Mozambique, Nicaragua, and Rwanda). In terms of access to health, the World Bank has improved access to health services (in Ethiopia, Indonesia, Mozam- bique, Nicaragua, Niger, and Rwanda), including to immunizations, family planning, institutional delivery, and antenatal and postnatal care. In some countries, the World Bank has contributed to child health through expanding growth monitoring and promotion, screening, and treatment of malnourished children (in Ethiopia, Madagascar, Niger, and Rwanda). There were also gains 50 in the prevention and treatment of childhood diseases, including diarrhea, parasitic infection, and malaria, supported by health services. Contributions to maternal nutrition are limited across countries, whereas improvements have been made in the provision of iron–folic acid to pregnant women. In terms of access to WASH, in Ethiopia, Malawi, Nicaragua, and Rwanda, the World Bank has increased access to water and sanitation, such as piped water and latrines. Community programs in Madagascar likely improved WASH be- haviors. However, in Madagascar, Mozambique, and Niger the World Bank’s contribution to WASH has been modest. Health, food, and care interventions, such as nutrition counseling, parent edu- cation, breastfeeding promotion, and support to backyard gardens, account for most behavior change results in nutrition determinants. These interventions have been quite successful in tracking evidence of actual behavioral practices, achieving 55–70 percent of behavior change indicators at the apply level and 58–72 percent of sustain level changes in the behavior of mothers or caregivers and communities.3 Still, within most projects there is a lack of tracking incre- mental behavior changes along a results chain of engage-learn-apply-sustain, and sustained behaviors have been less measured overall. Community-based programs supported by the World Bank contribute to behavior changes to improve nutrition determinants, but there was limited evidence of longer-term sustained changes. The behavior change framework and process mapping have been applied in case studies to assess how behav- World Bank Group Independent Evaluation Group    51 ior changes have been supported by frontline workers, community groups, and nongovernmental organizations, among other types of stakeholders. Case study evidence suggests that the World Bank has contributed to engag- ing actors and learning (although this is not often measured), and in some cases to new practices by caregivers, farmers, and health workers, among others, but there was limited evidence that the World Bank has contributed to longer-term sustained changes in the behaviors of actors. In most of the countries, CBN programs are still being strengthened, providing an opportu- nity to improve evidence and learning regarding behavior change.4 Impact evaluations of specific project investments show positive results in improving nutrition determinants. For instance, a randomized controlled tri- al evaluation of over 3,000 villages and 1.8 million target beneficiaries of the Generasi program in Indonesia found that community block grants to rural communities are an effective tool where the use of basic health care services is constrained not just by demand but also by supply and access. Moreover, positive impacts on the use of basic health care services have been higher in communities whose grants were linked to performance-based incentives, suggesting that the attempt of the Generasi program to replicate the condi- tionality of cash transfers on a community-wide level can produce positive results (Olken, Onishi, and Wong 2011).5 Another randomized controlled trial evaluation of a Total Sanitation and Sanitation Marketing program in Indonesia, when the program had been implemented at scale in rural East Java, found that sanitation improvements were largely driven by an increase in the rate of toilet construction by nonpoor households, whereas improve- ments remain limited for lower-income households, which are more likely to be credit constrained. Self-reported open defecation has decreased and par- asitic infestations in the nonpoor sample with no sanitation at baseline has also been reduced. Diarrhea prevalence has dropped 30 percent in treatment communities compared with control communities among young children likely affected by differences in drinking water and hand washing behavior (Cameron, Shah, and Olivia 2013). World Bank Support to Reducing Child Undernutrition  Chapter 3 Contributions to Social Norms The World Bank is not contributing in a substantial way to improving social norms in the nutrition portfolio. The World Bank’s nutrition portfolio gives insufficient attention to social norms interventions relating to early mar- riage, early pregnancy, birth spacing, and women’s empowerment (decision-making regarding childcare, food production, and health care seeking) when designing nutrition projects. The alignment of the portfolio interventions falls short in addressing social norms needs given its relative importance for achieving better nutrition outcomes. Even when social norms are supported, expected results are rarely measured, further intensifying the lack of available results. The evidence base for project-level results is very thin, encompassing only 21 indicators. Evidence gathered through case studies also reflects the modest contributions of the World Bank toward social norms outcomes. Projects likely had some 52 contribution in improving knowledge on sexual and reproductive health and rights and in delaying pregnancy (Ethiopia, Nicaragua, and Niger); gender roles in agriculture (Madagascar, Niger, and Rwanda); girls’ enrollment in school (Niger); and family planning usage (Ethiopia, Niger, and Rwanda). For example, Nicaragua’s support to sexual and reproductive health and rights likely helped increase contraceptive usage and reduce teen pregnancy and gender-based violence. Agriculture support has likely been particularly important to improve women’s participation in food production, storage and transformation, and livestock farming for milk and meat. Contributions to Nutrition Outcomes In World Bank nutrition support, the pathway to improve nutrition out- comes for mothers and children has been through support to nutrition determinants. This approach is consistent with a low emphasis of nutrition outcomes in projects’ objectives and consequently the limited measurement of nutrition outcome indicators in projects’ results frameworks. In addition to their low frequency, improvements in nutrition outcomes at project level have been much harder to achieve (53 percent) compared with immedi- ate and underlying determinants and institutional strengthening results. Moreover, the multivariate regression analysis found that the inclusion of nutrition indicators in results frameworks (such as anthropometric measure- ments, micronutrient status, and cognitive development) is associated with a lower project performance.6 This finding is likely due to the time lag to see World Bank Group Independent Evaluation Group    53 movement in these indicators, which makes it difficult to measure improve- ment in outcomes in the time frame of a single project. Anthropometric measures and the micronutrients status of children under five have improved in most case study countries over the evaluation period, but these changes are more difficult to attribute to World Bank support. The prevalence of wasting and underweight has decreased in most of the coun- tries, likely because of investments in growth monitoring and promotion and treatment of malnutrition. In countries such as Malawi and Mozambique, repeat crises likely have led to limited improvements in nutrition indicators. Indicators of stunted growth, LBW, and anemia may have decreased, but levels remain high in most of the countries. The achievement of nutrition outcomes at the country level is more difficult to attribute to World Bank support given the multiplicity of factors and development partners involved, the synergies among multisectoral interventions, and the rather longer time span for the changes in nutrition outcomes to materialize. Impact evaluations for specific projects provide evidence on the positive impact of World Bank efforts on nutrition outcomes for target beneficiary groups. The impact evaluation of the Generasi program in Indonesia shows that nutrition outcomes for children in project implementation areas have improved compared with those of a control group. The impact was stronger in areas with higher undernutrition before project implementation, where underweight rates have declined by 8.8 percentage points (20 percent com- pared with control areas); severe underweight rates have dropped by 5.5 per- centage points (33 percent compared with control areas); and severe stunted growth has been reduced by 6.6 percentage points (21 percent compared with control areas; Olken, Onishi, and Wong 2011). Another example is an experimental design evaluation of the ECD project for expanding access to community-based early childhood services in rural Indonesia, which has found that the project led to improvements in lower-income children’s social competence, language, cognitive development, and their emotional maturity (Brinkman et al. 2015). World Bank Support to Reducing Child Undernutrition  Chapter 3 Measuring Nutrition Results Persistent gaps exist in the measurement of nutrition-related results within projects, especially when tracking expected achievements from nutrition- specific and social norms interventions. IEG has calculated a matching score reflecting the alignment between the supported nutrition interventions and the presence of indicators to track progress on results based on a classifica- tion of more than 2,500 indicators according to the dimensions of the con- ceptual framework. On average, the matching score between interventions and indicators for the entire nutrition portfolio is 57 percent, and it has slightly improved over time (figures 3.2 and 3.3). Expected results from nu- trition-sensitive interventions are the most frequently measured, especially in health and family planning services, social safety nets, and agriculture and food systems. The high measurement gap in social norms further intensifies the lack of available results in an area where the World Bank has not given enough attention when designing nutrition projects. Among GPs, Water, 54 HNP, and Agriculture have most consistently tracked progress on results from their interventions.  istribution of Interventions, Intended Outcomes, and Project Figure 3.2. D Indicators 100 80 Sensitive/underlying Intended outcomes—PDO (percent) 60 Institutional strengthening 40 Nutrition outcomes 20 Specific/immediate Social norms 0 0 20 40 60 80 100 Interventions (percent) Sources: Independent Evaluation Group; portfolio review and analysis. Note: The vertical axis represents the percentage of projects’ objectives per dimension, and the horizon- tal axis represents percentages of interventions. The size of each bubble represents the share of project World Bank Group Independent Evaluation Group    55 indicators by area. PDO = project development objective. Figure 3.3. Measurement of Nutrition Results at the Project Level a. Measurement of nutrition results by intervention area Diet and breastfeeding (n=126) 0.54 Child disease prevention and treatment (n=116) 0.46 Adolescent health (n=14) 0.07 Health and family services (n=118) 0.93 Social safety nets (n=74) 0.76 Agriculture and food systems (n=115) 0.72 WASH approaches (n=86) 0.53 Early childhood development (n=32) 0.06 Social norms (n=41) 0.32 Policy, financing, and coordination (n=102) 0.41 Improved nutrition service delivery (n=190) 0.52 Stakeholder engagement (n=125) 0.53 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 World Bank Support to Reducing Child Undernutrition  Chapter 3 Matching score b. Measurement of nutrition results by Global Practice 56 c. Measurement of nutrition results by time period c. Measurement of nutrition results by time period Sources: Independent Evaluation Group; portfolio review and analysis. Note: The matching score is the share of interventions with an indicator to track progress of results in the same area. Matching score levels: Low, matching score <= 0.40; Medium, 0.40 < matching score <= 0.80; High, matching score > 0.80. The mean matching score is 0.59, the median is 0.57, the standard deviation is 0.29, and the range is 0–1. WASH = water, sanitation, and hygiene. Similarly, weak measurement of the progression along the behavior change results chain hinders effective planning and evaluation. IEG has developed a framework and process map that traces evidence of behavior changes across different actors along the engage-learn-apply-sustain results chain (figure 3.4). The goal of behavior change interventions is to engage different actors to induce sustained practices in the long run. However, when track- ing evidence of behavior change results, World Bank projects rarely follow World Bank Group Independent Evaluation Group    57 the results chain, leading to an incomplete monitoring of change processes. Moreover, many projects do not measure behavior changes. For projects that measured behavior changes, they most often measured the apply level (43 percent), followed by the engage (23 percent), learn (14 percent), and sustain (23 percent) levels. For example, the case study found evidence of be- havior change, although it was not always measured in the project indicators (appendix G). The behavior change map for Nicaragua shows that to improve access to food and care, at the engage level, parents were engaged by nutri- tionists and promoters in communities; at the learn level, parents learned to prepare new foods and learned practices for caregiving of children; at the apply level, families increased their consumption of a variety of foods, and women and children increased the number of food groups consumed; at the sustain level, there was no evidence available on sustained diet diversity. Figure 3.4. Tracing Evidence of Behavior Change Levels in Actors Institutional Engage Learn Apply change Actor gains Actor develops Actor draws on Change in actor adequate new knowledge available to consistently Results awareness and and skills for resources and support chain motivation to changing programs to determinant of develop new behavior support behavior undernutrition capabilities change Sources: Independent Evaluation Group; behavior change analysis. Explaining Nutrition Results: Successes and Failures behind Project Performance Effective pathways to improve project performance in achieving nutrition re- sults are improving M&E of determinants across sectoral areas, community- World Bank Support to Reducing Child Undernutrition  Chapter 3 based implementation, country ownership, and designing projects to address nutrition determinants and to strengthen country coordination capaci- ty. Based on analysis of Implementation Completion and Results Reports, Implementation Completion and Results Report Reviews, and Project Performance Assessment Reports of closed projects, IEG has identified nutrition-relevant factors behind the success or failure in achievement of nutrition results. Unsupervised hierarchical clustering machine learning algorithms have been used to build a taxonomy of factors emerging from the text of project documents, suggesting that project failure factors are the reverse or absence of success factors in the nutrition portfolio.7 Successful projects had M&E frameworks that measured the expected con- tribution to nutrition determinants across sectoral areas. Important was measuring indicators that were related to nutrition determinant in a theory of change, such as feeding practices, micronutrient supplementation, and use of nutrition-sensitive services; that were at a geographical level sup- 58 ported by the project, such as the community level, rather than nationwide; that measured changes that could be observed in the project time frame and at different levels of the results; that measured the expected achievements of a range of interventions supported by the project; that had a number of sources for routine (such as administrative data) and periodic data collection (such as population surveys and operational research studies), which could be triangulated to review progress; and that included indicators to measure the quality of services. Moreover, as observed in the country stocktaking (appendix H), the strengthening of the multisectoral collection of M&E data on nutrition interventions has been critical for implementation monitoring in Malawi, Peru, Rwanda, and Senegal. Successful projects have a strong community-based implementation, which is also key to induce behavior changes that improve nutrition determinants. The strength of community engagement, participation, and leadership in implementing nutrition interventions; support for capacity building in com- munities for selecting and managing local subprojects; and collaboration between communities and local partners in delivering social services matter for the achievement of nutrition results. This evaluative finding is consistent with multivariate regression estimates.8 For example, community partici- pation has been key for a project in Burundi that sought to strengthen the capacities of local community groups to work together in selecting, imple- menting, financing, and monitoring and maintaining priority community services. Project implementation has included timely contribution by ben- World Bank Group Independent Evaluation Group    59 eficiaries to subproject costs so that they would own the subprojects and be very much involved in their maintenance and continued operation. Other important factors explaining projects’ success are country ownership and having a design to reinforce existing country structures and to develop country coordination capacity for nutrition. This includes supporting gov- ernment commitment to nutrition and aligning projects to develop insti- tutional capacity for nutrition in terms of coordinating adequate financing, supporting nutrition-related reforms, the availability and commitment of a skilled workforce to support nutrition, and developing the capacity of line ministries and executing agencies to coordinate action and service delivery. A project in Senegal that sought to reduce nutrition insecurity of children under five by expanding the country’s National Enhancement Program, for instance, has seen that the institutional setup in the prime minister’s office has enhanced coordination of the policy dialogue among multiple stakehold- ers and sectors. This has increased stakeholders’ shared responsibility of the observed nutrition-related problems, which in turn has contributed to the enhanced uptake of services. The intentional design of interventions to support results in nutrition de- terminants matters for project success and is often better in projects with a heavy focus on nutrition than in sectoral projects. Where interventions are integrated in sectoral projects with few nutrition interventions, the design of support to achieve expected results in nutrition determinants is often weak. Multivariate regression analysis found that sectoral projects that integrated a few nutrition interventions had a lower achievement of nutrition outcomes and determinants when they planned interventions beyond their area of expertise, but projects that were designed with a heavy nutrition focus were able to contribute to results for nutrition determinants across a range of sectors. In case study countries, interventions integrated in Agriculture and Water projects have often lacked an intentional design to improve nutrition determinants, such as access to nutritious foods or hygiene and sanitation practices of households with children. The use of diagnostics to inform project design and evidence-based policies World Bank Support to Reducing Child Undernutrition  Chapter 3 contributes to project success in achieving nutrition results and is particu- larly important in multisectoral approaches toward reducing undernutrition for country programs and policy. In Rwanda and Indonesia, a nutrition situa- tion analysis, rapid mapping on nutrition-specific and nutrition- sensitive interventions, and nutrition public expenditure review have sup- ported the government in developing its multisector strategy and identi- fying needs to improve nutrition financing for multisectoral coordination. Consistently, regression analysis has found a strong positive association between project performance and the use of diagnostic and analytical work at project design. Although some countries (Ethiopia, Indonesia, and Rwan- da) have better leveraged a mix of knowledge activities to help strengthen the nutrition results of both projects and the country’s program, countries’ support seldom strategically balances analytical work, knowledge sharing, and leadership building for improving evidence-based policies and nutrition programming. For example, in Mozambique, although services have been 60 emphasized, attention to relevant analytical work, support to leadership, knowledge sharing, and coordination activities has been limited. Moreover, although Madagascar has done extensive evidence learning to develop its services, less support has been given to develop leadership, policies, and coordination of nutrition. Over the years, a clear consensus has grown that the key to solving child un- dernutrition is multidimensionality in programming. Countries with a World Bank project portfolio that has a mix of nutrition-specific and nutrition- sensitive interventions and institutional strengthening provide a pathway to improve nutrition determinants and contribute to outcomes. Regression analysis offers some evidence that multidimensional country portfolios, having a mix of nutrition-specific and nutrition-sensitive interventions, are slightly positively associated with project performance. But strength- ening the World Bank’s within-country alignment and targeting of relevant interventions to address disaggregated needs or priorities is key, and case studies showed that projects in different GPs have been mostly implemented in different geographical areas and target groups, and they lacked mecha- nisms to integrate or converge actions or build on respective achievements to improve nutrition outcomes in the same communities. Although learning to integrate or converge the implementation of interventions of different sectors in the same geographical areas is emerging in countries, the World Bank needs to improve coordination across GPs in the implementation of interventions. Multidimensional projects offer one approach to overcome World Bank Group Independent Evaluation Group    61 coordinated targeting challenges, but they do not perform better on average than those focusing on a narrow set of intervention areas. Moreover, non- core nutrition projects that integrate a small nutrition component tend to perform worse in achieving nutrition results in areas that are not related to their sectors. 1  See appendix I on multivariate regression analysis for more details. 2  See section on factors of success and failure in project performance in appendix D. 3  See table C.1 for examples of behavior change indicators at the engage-learn-apply-sustain levels. 4  See table G.3 for behavior change assessment in selected countries. 5  The Generasi program target health indicators are (i) four prenatal care visits; (ii) taking iron tablets during pregnancy; (iii) delivery assisted by a trained professional; (iv) two postnatal care visits; (v) complete childhood immunizations; (vi) adequate monthly weight increases for infants; (vii) weighing monthly for children under three and biannually for children under five; and (viii) vitamin A twice a year for children under five. 6  See appendix I on multivariate regression analysis for estimation results. 7  The emerging taxonomy of factors has been reviewed and validated by the evaluation team. See appendix D for the complete taxonomy of success and failures factors. 8  See appendix I on multivariate regression analysis. World Bank Support to Reducing Child Undernutrition  Chapter 3 62 4 | Conclusions and Way Forward The evaluation confirms that the World Bank approach to nutrition— addressing nutrition determinants—provides a plausible pathway to improve nutrition outcomes. Among those determinants, the associations between access to health services and social norms and a country’s nutrition outcomes are the strongest, followed by access to WASH and food and care. Moreover, multivariate regression analysis using portfolio data suggests that institutional strengthening achievements contribute to the success of inter- ventions that address nutrition determinants in countries. The evaluation highlights encouraging bright spots, including an increase in the number of projects and improved nutrition outcomes in some coun- tries. In countries burdened by undernutrition, the World Bank invested an estimated $22 billion in nutrition across multiple sectors from FY08 to FY19 (including about $5.8 billion in RETFs). This financing has supported evidence-based interventions, with the number of projects tripling in recent years. Some countries, Madagascar and Senegal among them, now have more than a decade of experience using a combination of financing and knowledge work to improve nutrition outcomes through multidimensional nutrition programs, from which other countries can learn. In many countries, the World Bank has supported a mix of nutrition-specific and nutrition-sensitive interventions affecting access to food and care, health, and WASH to address multidimensional needs that can improve nutrition determinants. The World Bank has also consistently invested in institutional strengthening, particularly to improve services. Achieving a mix of interven- tions in a country portfolio can be accomplished through multidimension- al projects that deliver a broad set of nutrition interventions or through a multisectoral approach by coordinating and integrating nutrition interven- tions implemented by projects across GPs. Although most projects across GPs continue to support activities related to their respective sectors (about three-quarters of them), GPs increasingly attempt to integrate into projects 63 nutrition-related interventions that belong to other sectors, suggesting that the World Bank is beginning to work beyond sector silos. The overall performance of the World Bank’s nutrition portfolio is improv- ing over time and, based on the evaluation’s portfolio analysis, shows good results in improving nutrition determinants and institutional strengthening. At the same time, the nutrition portfolio is young, providing for opportuni- ties to further improve the evidence base of interventions, knowledge work, the addressing of nutrition in the country programs, and results achievement and measurement. Lessons The following five lessons arising from the evaluative evidence are offered to inform the design of the World Bank’s future multidimensional and multi- sectoral nutrition support and improve how the World Bank operationalizes the conceptual framework. Lesson 1: More intentional planning of nutrition support (financing and ASA) is needed in the country portfolio within countries to improve nutri- tion outcomes through supporting nutrition determinants, social norms, World Bank Support to Reducing Child Undernutrition  Chapter 4 behavior change, and institutional strengthening. The evaluation finds that the multidimensionality of the country portfolio matters for results. » Nutrition interventions can be supported by multidimensional projects that implement a range of interventions to address nutrition determinants or by trust funds and partnership, and better GP coordination. Interventions can also be supported by sector projects in GPs if accompanied by learning to de- sign nutrition interventions and efforts to coordinate implementation. Trust funds and partnerships have been especially catalytic to design new support in countries, which can be expanded with government ownership to develop comprehensive nutrition services. » Institutional strengthening can be done at national and local levels through support to coordinate and develop intervention packages and policies that engage multiple actors. At the national level, institutional strengthening can help develop multisectoral nutrition approaches, including arrange- 64 ments to coordinate, finance, plan, and communicate nutrition. Recent Program-for-Results investments in Indonesia focused on ensuring effective coordination and accountability mechanisms across sectors and levels of gov- ernment. At the local level, institutional strengthening has been important (for example, in Rwanda and Senegal) to operationalize the delivery of inter- ventions (such as nutrition counseling, child disease management, maternal health, home gardens, cash transfers, ECD, and hand washing) in country nutrition programs. » Addressing social norms is important for improving nutritional outcomes in countries. More emphasis on social norms, which currently accounts for only 6 percent of the portfolio across all GPs and sectors, is needed since it can also facilitate expanded actions on nutrition, for example, as women become more involved in food, health, family planning, and caregiving decisions in households. This emphasis links to the concept of accelerator actions from the behavior change findings, where supporting the empowerment of key change agents can influence other behaviors and facilitate changes toward nutrition determinants. Lesson 2: The targeting and continuity of support in countries matter to successfully influence nutrition determinants. The evaluation finds that the targeting, continuity, and sustainability of nutrition interventions are im- portant to achieve expected results from multisectoral nutrition approaches. » The quality and extent of subnational targeting of multisectoral interven- World Bank Group Independent Evaluation Group    65 tions matter for the ability to address (disaggregated) needs within countries. The evaluation’s heat map analysis shows that most countries have had interventions to address their needs across nutrition determinants. However, the case studies show that interventions are often small in scale. Moreover, interventions of different GPs or sectors often have not converged in the same communities to address synergistic needs for food, care, health, and WASH. To meaningfully improve outcomes, given the synergistic nature of determinants, all priority needs should be addressed across targeted com- munities. Recent approaches to improve the convergence of interventions and needs in countries, such as Indonesia and Rwanda, are promising. Mul- tidimensional projects offer one option to coordinate interventions to meet needs in the same community, but they have not performed better or worse overall. An alternative is improved coordination across GPs and with oth- er development partners in the implementation of multisectoral interventions within countries. The evaluation did not examine the relative cost-effectiveness of these two options. » Continuity of support, particularly at the community level, is important to successfully influence nutrition determinants for results. Madagascar and Senegal offer strong examples of the contribution made by continuity of support across projects and years to strengthening interventions. Community interventions involve building the capacity of a wide range of actors and pro- moting behavior change, which need to be sustained to successfully influence nutrition determinants. Strong community-based implementation is shown to be a success factor for improving project performance. Lesson 3: Improving the measurement of results of interventions addressing nutrition determinants and of behavior change will improve nutrition out- comes in countries. » Although the World Bank has improved its results measurement in the past 10 years, some areas still are not well measured. Projects measure only about 60 percent of achievements toward nutrition determinants. The evaluation consistently identifies M&E of nutrition indicators as a pathway to improve World Bank Support to Reducing Child Undernutrition  Chapter 4 project performance. » The World Bank’s nutrition-sensitive interventions increasingly have achieved results that contribute to nutrition determinants in countries. Yet, nutrition-specific interventions, mainly implemented by HNP, have not seen the same improvements, and these results are more challenging to achieve and require consistent support in countries. Areas where projects had limited success include diet diversity, child feeding, and micronutrient outcomes in women and children. » Most projects did not measure the results of behavior change. The evaluation’s behavior change process map and framework (engage-learn-apply-sustain) shows that the World Bank’s contributions to behavior change focus mostly on lower-level indicators related to the engagement of actors and stops short of looking at results. This limitation is a concern for interventions at the com- munity level and those that focus on the institutional strengthening of actors involved in delivering services, since the goal of behavior change interventions 66 is to engage various actors to induce sustained practices in the long run. There is a need for learning in countries to better track behavior change. Lesson 4: Refocusing the portfolio to have greater emphasis on nutrition-specific interventions, balanced with nutrition-sensitive inter- ventions across GPs can improve nutrition programs in countries. Although nutrition-sensitive interventions have increased in the portfolio, a similar proportional increase in nutrition-specific investments supported by health and other sectors is not seen except in some countries (such as Rwanda), despite the critical importance of supporting these interventions in countries. The eval- uation’s SRM shows that effective interventions can be delivered by health, social protection, agriculture, and WASH sectors. Investing in improving nutrition-specific interventions in countries is needed, together with nutri- tion-sensitive support. Lesson 5: Learning—the systematic generation and use of knowledge work— is important to help countries to design and expand effective nutrition policy and programming. Some case study countries use a combination of knowledge work—such as analytical work, knowledge sharing, and leadership activities—to help the development of nutrition interventions and policies. Key examples are Ethiopia, Indonesia, Madagascar, Rwanda, and Senegal. » Country-level learning requires a stream of analytical work (evaluations, diagnostics, and so on) to improve interventions and expand their targeted World Bank Group Independent Evaluation Group    67 delivery in national programs. For example, Madagascar had over a decade of ASA to develop its community-based program, which is being expanded; Rwanda has employed ASA to develop its nutrition-sensitive social protec- tion support and to roll out innovative tools, including a child length mat and child scorecard; and Nicaragua used years of evaluation evidence to develop its integrated models for health and social protection services. Consistent attention to learning is weaker in Malawi, Mozambique, and Niger. » Since nutrition is often not the objective of GP projects (such as in Agriculture and Water), interventions do not target to improve nutrition determinants and in some cases might even negatively affect child undernutrition (as in the example of support for cash crops). Attention to this issue and learning has already started at the global level, for example, through research on nu- trition-sensitive agriculture. In other sectors supporting nutrition-sensitive interventions (such as SPJ), case studies suggest learning to improve the design of interventions to better target nutrition determinants is more advanced. » Combining analytical work (evaluations, diagnostics) with knowledge shar- ing (within and across countries) and leadership building activities in coun- tries helps generate political commitment and the use of evidence to inform policies and programs and to leverage resources. In Malawi, multisectoral learning forums led to policies and strategies on nutrition. In Ethiopia, the convening of actors pooled resources for nutrition. In Indonesia, Madagas- car, and Rwanda, South-South learning led to commitment and technical knowledge to implement interventions. In Indonesia, Rwanda, and Senegal, high-level leadership activities pivoted the nutrition agenda to a priority for the country. Recommendations The preceding lessons support two broad recommendations to strengthen the support of the World Bank to nutrition: » Recommendation 1. Adjust nutrition programming in country portfolios to (i) give more priority to institutional strengthening for coordination and imple- World Bank Support to Reducing Child Undernutrition  Chapter 4 mentation of multisectoral nutrition interventions and (ii) increase focus on subnational targeting of interventions to reflect areas of greatest disadvan- tage and persistency of need. » Recommendation 2. Strengthen nutrition support in GPs to (i) rebalance investments to have greater emphasis on nutrition-specific interventions and (ii) increase focus on social norms interventions and behavior changes, with more attention to tracking expected achievements to improve nutrition determinants. 68 Glossary Behavior change interventions. Interventions that engage changes agents (such as frontline workers, opinion leaders, and households) and mothers and primary caregivers to shift behaviors to influence determinants of nutrition. Behavior change results chain. The delineation of levels along a results chain (engage-learn-apply-sustain) that lead from initial inputs and outputs all the way to sustained behavior change that could be expected to persist after interventions are completed. Core nutrition project. A project with an explicit focus on nutrition in its objectives or title and a heavy focus on supporting nutrition interventions. Institutional strengthening support. Support to strengthen nutrition-related stakeholder engagement and ownership, policy, financing, and coordination, and service delivery. Multidimensional portfolio. This describes the inclusion of a mix of nutrition-specific and nutrition-sensitive interventions in the World Bank’s nutrition portfolio or portfolio of support in a country. Multidimensional project. These are projects that include a mix of nutrition-specific and nutrition-sensitive interventions that are inherently World Bank Group Independent Evaluation Group    69 implemented by different sectors. Multisectoral nutrition support. This refers to World Bank nutrition support that involves multiple World Bank Global Practices to support sector-focused nutrition interventions. Noncore project. Projects that do not have an explicit focus on nutrition in their title or objectives and integrate nutrition interventions in their components. Nutrition determinants. There are immediate and underlying nutrition determinants. Immediate determinants of child nutrition relate to caregiv- ing practices, dietary intake or diversity, and the health status of the mother and child. It is not possible to realize these factors when communities lack adequate access to underlying determinants of nutrition, including nutrient-rich food, caregiving resources, health services, and water, sanita- tion, and hygiene. Nutrition results. Projects can improve underlying determinants (nutrient-rich food, caregiving resources, health services, and water, sani- tation, and hygiene); immediate determinants (child feeding practices, diet diversity, and maternal and child health); social norms (women’s empower- ment and early marriage and pregnancy); and institutional strengthening, and to a lesser extent nutrition outcomes (anthropometric measures and micronutrients deficiencies). Nutrition-sensitive interventions. These are interventions, such as cash transfers, water, sanitation and hygiene approaches, girls’ education, and food system improvements, that are expected to address the underlying de- terminants of nutrition. Nutrition-specific interventions. These are interventions, such as adoles- cent nutrition, maternal nutrition, breastfeeding support, micronutrient World Bank Support to Reducing Child Undernutrition  Glossary of Key Terms supplementation, child disease prevention, and management and treat- ment of undernutrition, that are expected to influence the immediate determinants of nutrition. Social norms interventions. These are interventions that address social norms relating to early marriage, early pregnancy, birth spacing, and wom- en’s empowerment (decision-making regarding childcare, food production, health care seeking) to influence nutrition determinants. 70 Bibliography ACC/SCN (United Nations Administrative Committee on Coordination/Sub-Commit- tee on Nutrition). 2000. Fourth Report on the World Nutrition Situation: Nutrition throughout the Life Cycle. Geneva: ACC/SCN in Collaboration with International Food Policy Research Institute. https://www.unscn.org/web/archives_resources/ html/resource_000135.html. Alderman, H. 2016. Leveraging Social Protection Programs for Improved Nutrition: Summary of Evidence Prepared for the Global Forum on Nutrition-Sensitive Social Protection Programs, 2015. World Bank, Washington, DC. Berg, A. 1987. Malnutrition: What Can Be Done? Lessons from World Bank Experience. Washington, DC: World Bank. Brinkman, S. A., A. Hasan, H. Jung, A. Kinnel, and M. Pradhan. 2015. “The Impact of Expanding Access to Early Childhood Services in Rural Indonesia (Evidence from Two Cohorts of Children).” Policy Research Working Paper 7372, World Bank, Washington, DC. Cameron, L., M. Shah, and S. Olivia. 2013. “Impact Evaluation of a Large-Scale Rural Sanitation Project in Indonesia.” Policy Research Working Paper 6360, Impact Evaluation Series 83, World Bank, Washington, DC. World Bank Group Independent Evaluation Group    71 Denboba, A., R. Sayre, Q. Wodon, L. Elder, L. Rawlings, and J. Lombardi. 2014. Step- ping Up Early Childhood Development: Investing in Young Children for High Re- turns. Washington, DC: World Bank. Development Initiatives. 2020. 2020 Global Nutrition Report: Action on Equity to End Malnutrition. Bristol: Development Initiatives. Galasso, E., and A. Wagstaff. 2018. “The Aggregate Income Losses from Childhood Stunting and the Returns to a Nutrition Intervention Aimed at Reducing Stunt- ing.” Policy Research Working Paper 8536, World Bank, Washington, DC. Gillespie, S., M. McLachlan, and R. Shrimpton, eds. 2003. Combating Malnutrition: Time to Act. Health, Nutrition, and Population Series. Washington, DC: World Bank. Hawkes, C., and M. T. Ruel. 2008. “From Agriculture to Nutrition: Pathways, Syner- gies and Outcomes.” Agricultural and Rural Development Notes 40, World Bank, Washington, DC. Heaver, R. 2005. Strengthening Country Commitment to Human Development: Lessons from Nutrition. Directions in Development Series. Washington, DC: World Bank. Horton, S., M. Shekar, C. McDonald, A. Mahal, and J. K. Brooks. 2010. Scaling Up Nutrition: What Will It Cost? Directions in Development Series. Washington, DC: World Bank. Laviolette, L., S. Gopalan, L. Elder, and O. Wouters. 2016. Incentivizing Nutrition: In- centive Mechanisms to Accelerate Improved Nutrition Outcomes. Washington, DC: World Bank. MacNally, W. 1983. “World Bank Assistance to Food and Agriculture 1974–84.” AGREP Division Working Paper 82, World Bank, Washington, DC. Maternal and Child Nutrition Study Group. 2013. “Maternal and Child Nutri- tion.” Series, Lancet 382 (9890). https://www.thelancet.com/series/mater- nal-and-child-nutrition. World Bank Support to Reducing Child Undernutrition  Bibliography Maternal and Child Undernutrition Study Group. 2008. “Maternal and Child Under- nutrition.” Series, Lancet 371 (9608). https://www.thelancet.com/series/mater- nal-and-child-undernutrition. Olken, B., J. Onishi, and S. Wong. 2011. Indonesia’s PNPM Generasi Program: Final Impact Evaluation Report. World Bank, Washington, DC. Pearson, R., M. Killedar, J. Petravic, J. Kakietek, N. Scott, K. L. Grantham, R. M. Stuart, D. J. Kedziora, C. C. Kerr, J. Skordis-Worrall, M. Shekar, and D. P. Wilson. 2018. “Optima Nutrition: An Allocative Efficiency Tool to Reduce Childhood Stunting by Better Targeting of Nutrition-Related Interventions.” BMC Public Health 18 (384). https://doi.org/10.1186/s12889-018-5294-z. Rokx, C. 2006. “Governance and Malnutrition, Exploring the Contribution of ‘Good Governance’ to Malnutrition Reduction in Developing Countries.” PhD thesis, Maastricht University. 72 Shekar, M., J. Kakietek, J. Dayton Eberwein, and D. Walters. 2017. An Investment Frame- work for Nutrition: Reaching the Global Targets for Stunting, Anemia, Breastfeeding, and Wasting. Directions in Development Series. Washington, DC: World Bank. Skoufias, E., K. Vinha, and R. Sato. 2019. All Hands on Deck: Reducing Stunting through Multisectoral Efforts in Sub-Saharan Africa. Africa Development Forum. Wash- ington, DC: World Bank. SUN (Scaling Up Nutrition) Movement. 2010. Scaling Up Nutrition: A Framework for Action. Geneva: SUN Movement. SUN (Scaling Up Nutrition) Movement. 2019. Nourishing People and Planet Together: Scaling Up Nutrition (SUN) Movement Progress Report 2019. Geneva: SUN Move- ment. https://scalingupnutrition.org/wp-content/uploads/2019/11/SUN-Annu- al-Report-2019-ENG_web_FINAL.pdf. UNCNC21 (United Nations Commission on the Nutrition Challenges of the 21st Century). 2000. Ending Malnutrition by 2020: An Agenda for Change in the Millennium. Geneva: United Nations. https://www.unscn.org/uploads/web/news/2000-FEB-Ending-Mal- nutrition-by-2020-Agenda-for-Change-in-the-Millennium-Report.pdf. UNICEF (United Nations Children’s Fund), WHO (World Health Organization), and World Bank. 2019. Levels and Trends in Child Malnutrition: Key Findings of the 2019 Edition of the Joint Child Malnutrition Estimates. Geneva: WHO. UNICEF (United Nations Children’s Fund). 1990. Strategy for Improved Nutrition of World Bank Group Independent Evaluation Group    73 Children and Women in Developing Countries. New York: UNICEF. UNICEF (United Nations Children’s Fund). 2015. “UNICEF’s Approach to Scaling Up Nutrition for Mothers and Their Children.” Discussion Paper, Programme Divi- sion, UNICEF, New York. WHO (World Health Organization). 2014. Comprehensive Implementation Plan on Maternal, Infant and Young Child Nutrition. Geneva: WHO. World Bank. 2006. Repositioning Nutrition as Central to Development: A Strategy for Large- Scale Action. Directions in Development Series. Washington, DC: World Bank. World Bank. 2013a. Improving Nutrition through Multisectoral Approaches. Washing- ton, DC: World Bank. World Bank. 2013b. World Bank Group Strategy. Washington, DC: World Bank. World Bank. 2014. Learning from World Bank History: Agriculture and Food-Based Approaches for Addressing Malnutrition. Washington, DC: World Bank. World Bank. 2017. SecureNutrition History, Adaptation, and Selected Impact from In- ception through December 2017. Washington, DC: World Bank. World Bank. 2018. The Human Capital Project. Washington, DC: World Bank. World Bank. 2020. The World’s Bank: An Evaluation of the World Bank Group’s Global Convening. Washington, DC: World Bank. World Bank Support to Reducing Child Undernutrition  Bibliography 74 APPENDIXES Independent Evaluation Group World Bank Support to Reducing Child Undernutrition Appendix A. Overall Methodology Evaluation Questions The overarching question that the evaluation answers is: What has been the contribution of World Bank support in improving outcomes and intermediate outcomes in reducing child undernutrition and improving nutrition determinants in countries that are burdened by undernutrition? Underlying this question are three main lines of inquiry (box A.1). Box A.1. Three Evaluation Questions Guide the Evaluation 1. To what extent is the World Bank supporting relevant interventions to improve outcomes and intermediate outcomes of child undernutrition and its determinants within the country context? 2. How is the World Bank implementing multidimensional approaches to support outcomes and intermediate outcomes that reduce child undernutrition and improve its determinants, and strengthen countries’ institutional capacities? 3. To what extent have World Bank interventions contributed to outcomes and intermediate outcomes of reducing child undernutrition and improving its determinants, and what were the factors of success and failure? Source: Independent Evaluation Group. Overarching Principles The evaluation design adopts a multilevel analysis at the global, portfolio, country, and intervention levels. It uses a mixed-methods approach that combines quantitative and qualitative evaluative evidence and applies participatory, theory-based, and case-based principles. The conceptual framework underpinning this evaluation is adapted from the United Nations Children’s Fund framework of child undernutrition. The framework models interlinked dimensions to sustainably address child undernutrition in a country context. These dimensions are nutrition-specific and nutrition-sensitive interventions addressing the immediate and underlying determinants of nutrition—respectively, social norms interventions and institutional strengthening support—considering factors within the country that are used to prioritize and target interventions. The conceptual framework guides the nutrition agenda of the World Bank and global efforts on nutrition. This framework is used to align the learning from the evaluation with existing efforts to support nutrition. 78 Appendix A Overall Methodology Evaluation Components Table A.1 lists the evaluation components, and figure A.1 shows their articulation within the overall evaluation design. The next two sections provide details on each component. The components are at the global, portfolio, country, and intervention levels to triangulate evidence from different methods and address the three evaluation questions. Table A.1. Evaluation Components Evaluation Component Description Systematic review map The systematic review map visually synthesizes the available evidence from systematic reviews of the literature on the effectiveness of nutrition-specific and nutrition-sensitive interventions on nutrition outcomes and determinants to benchmark this evidence against support in the portfolio (appendix B). Literature review and behavior change process map A structured literature review identifies and categorizes behavior change concepts and evidence to develop a set of process maps that reflect a basic results chain for benchmarking behavior change in projects (appendix C). Stocktaking of multisector approaches A qualitative stocktaking exercise of 12 countries is conducted to (i) understand multisectoral approaches to nutrition in different country contexts, and (ii) understand how the World Bank has helped enhance multisectoral coordination through institutional capacity building (appendix H). Portfolio identification, review, and analysis The systematic identification, coding, extraction, and analysis of the World Bank’s nutrition lending portfolio (282 projects), on its relevance, its multidimensional approaches, and its contributions to nutrition results in countries (appendix D). Indicators mapping Indicators to measure nutrition results in project results frameworks are mapped to the conceptual framework to assess how the World Bank measures its results and achievement rates (appendix D). Artificial intelligence: theory-based content analysis Lessons on success and failure factors explaining (topic modeling) project performance have been extracted from closed project documents in the portfolio review. These are used for topic modeling to develop a taxonomy of common success and failure factors that influenced the results of nutrition projects (appendix D). Heat map analysis The heat map (i) uses data on nutrition outcomes and determinants to understand the situation of countries and their empirical links based on the conceptual framework; and (ii) assesses the extent to which the World Bank’s nutrition interventions aligned with the country needs (appendix F). 79 Appendix A Overall Methodology Case studies Case-based analyses of eight countries include a review of the nutrition country portfolio (projects, analytical work, impact evaluations, and partnerships), interviews, and analysis of the World Bank’s contribution to results against the conceptual framework (appendix G). Multivariate regression analysis A multivariate regression analysis is anchored in the conceptual framework with the main objective of uncovering predictors of project performance, drawing on findings from the portfolio review and case studies (appendix I). Source: Independent Evaluation Group. 80 Appendix A Overall Methodology Figure A.1. Evaluation Design Matrix Source: Independent Evaluation Group. Note: EQ = evaluation question. 81 Appendix A Overall Methodology Ensuring Validity of Findings The Independent Evaluation Group (IEG) took several steps to guarantee a consistent approach across the evaluation team members—for example, using a case study template and interview protocols to ensure a common framework and evaluative lens across studies. Similarly, IEG uses consistent protocols and templates for the portfolio coding, multisector stocktaking exercise, and literature reviews. Furthermore, the team applies triangulation at multiple levels, first by cross-checking evidence sources within a given methodological component. Within case studies, for example, the team compares evidence from interviews with World Bank staff, country counterparts, beneficiaries and partners, project documents, and existing evaluations. Second, the team applies triangulation across evaluation components—for example, cross-validating findings from case studies with findings from the multisector stocktaking, systematic review map (SRM), heat map analysis, and portfolio analysis. Moreover, the multivariate regression analysis tests hypotheses based on findings from different evaluation methods. The evaluation team also applies external validation mechanisms at various intervals during the evaluation process. For example, the team identifies the portfolio of core activities through an iterative process in dialogue with the Global Practices. Peer reviewers provided feedback at the beginning, during, and end of the evaluation process. The team organized consultations with key stakeholders to validate the scope and methods of the evaluation, the design of the case studies, and at the end to ensure the relevance and feasibility of the evaluation messages. Limitations Notwithstanding these steps, the following are limitations of the evaluation design: • A key strength of the literature review for the SRM is that no interventions are ruled out ex ante. However, the review is limited to evidence from systematic reviews (SRs) and thus may miss recent studies that are not included in SRs and gray literature. Moreover, there is a risk of duplication of the underlying studies included in the SRs, which may skew the results reported in either a positive or negative direction. Also, outcomes may have been measured differently by different studies (see appendix B for more detailed limitations). • Although the behavior change map provides a useful results chain (engage- learn-apply-sustain) to benchmark behavior change in projects, the analysis of behavior change interventions in the portfolio and in case studies is limited by the lack of indicators and descriptive details on these interventions in project documents. 82 Appendix A Overall Methodology • The team purposefully centers the case study selection and multisector stocktaking analysis on countries in which the World Bank has a multidimensional portfolio of nutrition projects and on countries that Global Practice (GP) colleagues identify as being able to provide relevant learning. This purposive sampling of countries may not be representative of the total population of countries in which the World Bank is active (see appendix G for details on country selection). • The team faced several challenges in identifying the nutrition portfolio, given its spread across GPs and the limited quality of the theme codes to identify nutrition projects. Therefore, a machine learning exercise was used to identify the nutrition portfolio. This missed some projects where nutrition interventions are not clearly articulated in project documents, although some relevant projects were recovered during case study analysis based on the suggestions of GPs. Nonetheless, the machine learning created algorithms that could be used by the GPs to support routine efforts to identify the nutrition portfolio. • Moreover, given that projects often had other interventions not related to nutrition, the team calculated the total amount of World Bank financing of the nutrition portfolio based on an estimate of the proportion of nutrition interventions in each project from the portfolio coding. • The portfolio analysis is limited by a lack of data on the intensity and timeline of interventions in countries. Moreover, some project documents lack details on the specific interventions being implemented in countries to compare the nutrition portfolio with the evidence base. • Whereas the indicator mapping provides an estimate of World Bank achievements, some project indicators are not reported or are missing. Moreover, since the portfolio is young, many active projects and the achievements of these projects could not be assessed. • For the heat map analysis, data on nutrition-related indicators are at the national level, limiting the assessment of disaggregated needs within countries. Moreover, data are missing for some countries, and years of available data differ across countries, which limits the calculation of trends to assess the change in the nutrition situation over the evaluation period. 83 Appendix A Overall Methodology Description of Evaluation Methods Systematic Literature Review SRM. The SRM is based on the conceptual framework and synthesizes the available evidence on the effectiveness of nutrition-specific and nutrition-sensitive interventions. The review of the literature is limited to evidence from SRs reporting effects of any type of nutrition-specific and nutrition-sensitive interventions on the following nutrition- relevant outcomes: child undernutrition and development (birthweight, micronutrient status and deficiencies, stunted and linear growth, and cognitive development); child feeding and caregiving (breastfeeding, complementary feeding, and parenting practices); child health and disease (enteric infection and diarrhea, and childhood illness and infection); maternal health (nutrition status and deficiencies, nutrient intake and dietary diversity, healthy pregnancy, and mental health); access to health services, water, sanitation, and hygiene (WASH) services and nutrient-rich food (maternal use of health services, child use of health services, WASH, and household food and nutrition security); maternal and childcare resources (household welfare, schooling, knowledge and attitudes, and household safety); and social norms (women’s empowerment, early pregnancy, and birth spacing). Other manifestations of child malnutrition (such as overweight or obesity) are outside of the scope (appendix B). Literature review and behavior change process map. A structured literature review identifies and categorizes behavior change concepts and evidence of how interventions have supported behavior change toward nutrition determinants. The evidence from the literature is used to understand the incremental sequences of actions that can lead to sustained behaviors to improve nutrition determinants, such as access to food, caregiving resources, health services, and WASH. The review includes qualitative studies (such as qualitative SR and empirical studies) on behavior change interventions. The findings are used to develop process maps that reflect a basic results chain for benchmarking behavior change in projects in the portfolio and in the case studies (appendix C). Stocktaking of Multisector Approaches The stocktaking analysis focuses on a purposeful sample of 12 countries, 8 of which are also case studies: Bangladesh, Ethiopia, Indonesia, Madagascar, Malawi, Mozambique, Nepal, Nicaragua, Niger, Peru, Rwanda, and Senegal. These countries are of interest because their lending portfolios have a high degree of multidimensionality in their mix of nutrition-specific and nutrition-sensitive interventions over the 10-year evaluation period. A qualitative stocktaking template captures descriptive details consistently across countries at the national and subnational levels. Data collection for each country 84 Appendix A Overall Methodology is conducted by reviewing country documents on nutrition (such as plans), and published case studies. The stocktaking exercise reviews country institutional arrangements for the coordination of nutrition, the delivery of interventions, and behavior change communication. The portfolio review data and case study evidence are then used to understand how the World Bank has contributed to institutional strengthening of multisectoral arrangements in these countries. These findings provide the basis for developing typologies for characterizing multisectoral approaches to nutrition in different country contexts and for highlighting factors that help facilitate or hinder multisectoral coordination (appendix H). Portfolio Identification, Review, and Analysis The portfolio review and analysis are anchored in the dimensions of the conceptual framework, that is, nutrition outcomes for mothers and children, immediate and underlying determinants, and nutrition-specific, nutrition-sensitive, and social norms interventions and institutional strengthening support. It consists of a portfolio identification strategy followed by portfolio coding and analysis (appendix D). Portfolio identification strategy. The strategy consists of four stages—search, delimitation, inclusion, and verification—to progressively define the nutrition-relevant portfolio for the evaluation. • The search stage consists of data retrieval from the World Bank’s Business Intelligence repository on active and closed lending projects that fall within the evaluation period fiscal years 2008–19 and are financed through International Bank for Reconstruction and Development, International Development Association, and recipient-executed trust fund agreements. The project features include project identification, titles, countries, regions, lead GP, lending instruments, approval and revised closing years, sector and theme codes, and additional financing flags. Project indicator data are also retrieved from Implementation Status and Results Reports. • The delimitation stage uses relevant sector and theme codes as project filters and restricts the sample to those operations implemented in high countries with high rates of stunted growth. These countries are defined as those having stunted growth rates at or above 20 percent at any point during the evaluation period. • In the inclusion stage, IEG defines a list of key nutrition concepts and associated keywords based on the conceptual framework as input for a machine learning exercise to improve the accuracy of project identification through text analytics. The machine learning algorithms use corpus of text (project development objectives [PDOs], project indicators, and project components extracted from 85 Appendix A Overall Methodology project documents) from 4,260 projects (see appendix D). Inclusion criteria are then applied to filter projects with relevant title, PDO, components, or indicators. For determining relevant PDO, components, and indicators, a combination of different thresholds for saliency and similarity scores are used to ensure that the most relevant projects are included in the portfolio. • The verification stage consists of a manual verification of 291 projects against lists of nutrition projects from the Nutrition Global Solution Group, and the Agriculture; Health, Nutrition, and Population; and Water GPs. Portfolio coding and analysis. The nutrition portfolio is manually reviewed and coded. The coding template is based on the conceptual framework and administered through Survey Monkey. It extracts project information on nutrition challenges, PDOs, interventions, project beneficiaries, and factors of success and failure relevant for a project’s nutrition outcomes. Coders reviewing the projects also estimate each project’s share of nutrition content and identify any remaining misclassified false-positive projects. Coders had training, a piloting phase, and periodic quality assessment and spot checks to ensure the reliability of their coding. The final input for portfolio analysis consists of 282 parent projects and 133 additional financing. Portfolio data analysis is in Excel, Stata, and Tableau software. Indicators mapping. Indicators extracted from Implementation Status and Results Reports of projects in the final portfolio are coded in Excel. This codes nutrition-relevant indicators, measuring nutrition outcomes, immediate and underlying nutrition determinants, institutional strengthening, social norms, and behavior change. The evaluation uses the indicators to assess (i) the achievement rate of indicators for closed projects, and (ii) the extent that indicators in results frameworks measure the intended results of project interventions (open and closed projects). A total of 2,571 nutrition- related indicators are coded for the 282 projects (135 are from closed projects, of which 131 had information on project indicators). Artificial intelligence: theory-based content analysis (topic modeling). In the portfolio coding, lessons (classified as success and failure factors) are extracted from Implementation Completion and Results Reports, Implementation Completion and Results Report Reviews, and Project Performance Assessment Reports of closed projects. For 117 of the 135 closed projects, 562 factors are identified. These factors are then analyzed through unsupervised hierarchical clustering machine learning algorithms by Oxford Analytics and Endeavour to develop a taxonomy of common success and failure factors emerging from the projects’ texts. The taxonomy is manually reviewed to define a final list of 10 factors that influence nutrition project achievements in countries. 86 Appendix A Overall Methodology Behavior change portfolio analysis. The portfolio coding also codes behavior change indicators and interventions that support nutrition determinants. This is guided by the results chain (engage-learn-apply-sustain) and qualitative mapping of the behavior change toward nutrition determinants in the process map (appendix E). Heat Map Analysis The heat map summarizes country nutrition outcomes and determinants. The analysis uses the main dimensions of the conceptual framework—nutrition outcomes, determinants (including access to nutrient-rich food, maternal and child caregiving, WASH, and health services), and social norms—to guide data collection on indicators to assess countries’ situations. See appendix F for the list of indicators and secondary data sources for 64 countries. Principal component analysis is used to calculate a composite measure for each nutrition determinant (that is, access to food and care, WASH, and health services), social norms, and nutrition outcomes at the baseline level, the current level, and for their trends over the 10-year period. Pearson correlation analyses are conducted between the levels and trends of nutrition outcomes and their determinants in the evaluation countries to empirically test their links in the conceptual framework. Further, using portfolio data, nutrition-related interventions in the World Bank’s portfolio in each country are mapped to the determinants to assess whether country needs are matched by World Bank interventions. Country Case Studies Selection of cases. The evaluation includes a case-based analysis of the World Bank’s nutrition portfolio in eight countries (Ethiopia, Indonesia, Madagascar, Malawi, Mozambique, Nicaragua, Niger, and Rwanda), selected from the 65 countries in the evaluation’s portfolio. The inclusion criteria for the countries are (i) countries with at least one closed IEG-evaluated project with a nutrition focus in the title or PDO; (ii) countries with support for institutional strengthening and behavior change interventions related to nutrition; and (iii) countries with projects in at least three GPs. Other considerations are the availability of impact evaluation evidence on interventions in the country; whether the country has a Human Capital Index rating in the bottom or third quartile compared with other countries; the extent that the country’s experience is already documented; and the coverage of countries in different Regions. Criteria used to vary the selection of countries are the average annual change in stunted growth rates during the evaluation period (slow, medium, and fast, based on the quartiles of the data across countries) and the overall project performance based on achievement rates of nutrition indicators in the portfolio. Based on these criteria, 15 countries eligible for case studies were discussed with operational counterparts to finalize the country selection (appendix G). 87 Appendix A Overall Methodology Methods and data collection. The data collection in each country follows a case study protocol organized in relation to the conceptual framework and evaluation questions, looking at the relevance, multidimensionality, and results of World Bank support. The case study covers all active and closed lending projects and knowledge work in the country portfolio that supported nutrition-related interventions during the 10-year evaluation period (fiscal years 2008–19). Most data collection was remote because of travel restrictions related to the coronavirus pandemic. In each country, the IEG team worked with national consultants to facilitate country stakeholder interviews. Work on an IEG Project Performance Assessment Report was integral to the case study data collection in Madagascar and Malawi. Evidence sources triangulated for each country include the following: • A country portfolio review of relevant lending projects and analytical work, including a review of project appraisal documents, program documents, concept notes, Implementation Completion and Results Report Reviews, and knowledge work. • Semistructured interviews with World Bank staff, government counterparts, partners, and project beneficiaries. Among the World Bank staff interviewees are task teams, country management, and experts involved in the implementation of nutrition projects in countries. Among government interviewees are key actors who coordinate and implement project interventions. Among partner interviewees are donor agencies and nongovernmental organizations that support nutrition activities in the same period to map synergies with other interventions. Among beneficiary interviewees are local leaders and community agents. • Secondary data on nutrition-related indicators from the heat map analysis of country contexts and needs (appendix F). • Evidence from evaluations, including existing impact evaluations and IEG evaluations. • World Bank Country Partnership Frameworks and Strategies from the period, and each government’s national development plan or nutrition strategy and plan. • Evidence of behavior changes from project evaluations and interviews that were assessed using the behavior change process maps developed for the evaluation (appendix C). 88 Appendix A Overall Methodology Contribution analysis. Evidence sources are triangulated to assess the contribution of each country to nutrition improvements against the dimensions of the conceptual framework. A country-specific theory of change is developed to assess how the nutrition interventions in the country program contribute to the dimensions of the conceptual framework. The analysis includes (i) the assessment of nutrition-related interventions that are supported by World Bank projects in the portfolio against the conceptual framework, including target populations and geographies of interventions, and roles of other partners in supporting interventions; (ii) the assessment of the alignment of nutrition interventions in the portfolio against country context and needs; (iii) the identification of the achievements of World Bank support against outcomes, intermediate outcomes, and outputs in the conceptual framework; and (iv) mapping how behavior changes are supported by project interventions to contribute to improvements in nutrition determinants. Multivariate Regression Analysis The multivariate regression analysis provides additional evidence for answering the third evaluation question (figure A.1) and for understanding the main drivers of project performance. Regression models test hypotheses that are based on findings from different exercises of the evaluation, including country case studies and portfolio review, and the relevant empirical literature. The analysis is based on the cross-section of 131 closed nutrition projects. The analysis uses information coded in the portfolio review and analysis, including indicators, nutrition interventions, factors of success and failure, and secondary data (appendix I). 89 Appendix B. Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Objective and Scope This appendix assesses the relevance of the World Bank’s nutrition lending portfolio; that is, how well the supported nutrition interventions align with the available evidence of “what works,” summarized in an systematic review map (SRM). An SRM is a visual presentation of the existing relevant evidence on effectiveness collected from a systematic review (SR) of the literature for a particular topic. SRMs are useful to help decision makers invest in interventions that are effective, and they highlight areas where further learning may be important to improve the implementation of evidence-based interventions. The scope of the SRM is based on the conceptual framework of undernutrition and synthesizes the available evidence on effectiveness of nutrition-specific and nutrition- sensitive1 interventions that aim to reduce child undernutrition and improve nutrition determinants. The review of the literature is limited to evidence from SRs reporting effects of any type of nutrition-specific and nutrition-sensitive interventions on the following nutrition-relevant outcomes areas: child undernutrition and development (birthweight, micronutrient status and deficiencies, stunted growth, stunted and linear growth, and cognitive development); child feeding and caregiving (breastfeeding, complementary feeding, and parenting practices); child health and disease (enteric infection and diarrhea, and childhood illness and infection); maternal health (nutrition status and deficiencies, nutrient intake and dietary diversity, healthy pregnancy, and mental health); access to health services, water, sanitation, and hygiene (WASH) services and nutrient-rich food (maternal use of health services, child use of health services, WASH, and household food and nutrition security); maternal and childcare resources (household welfare, schooling, knowledge and attitudes, and household safety); and social norms (women’s empowerment, early pregnancy, and birth spacing). Other manifestations of child malnutrition (low weight-for-age z score, low weight-for- height z score, and overweight or obesity) are outside of the scope, as are maternal overweight or obesity and excessive prenatal weight gain. Thus, the intervention axis of the SRM matrix was not defined from the outset but built from the literature review findings. Search Strategy Twelve databases—3ie, Campbell Collaboration, Cochrane, Cochrane Nutrition, EconLit, eLENA, IFPRI, IPA, J-PAL, PubMed, Science Direct, and Wiley Online—were 90 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio searched between August and December 2019. Expert consultations and reference tracking were also used to identify SRs. The search strategy focused on SRs that provide either a narrative synthesis or meta- analysis of studies published after 1993 in English. Search keywords reflecting the outcomes of interest listed above were used to identify relevant SRs. To maximize the capture of relevant SRs, the number of exclusion keywords was kept small, and articles were manually screened using context-sensitive keywords.2 Inclusion and Exclusion Criteria Search results were aggregated and analyzed in Excel, removing duplicates and SRs that were updated. The remaining SRs were then reviewed in stages, first to remove articles outside of the initial search criteria using a text search. Next, the article titles and abstracts were manually screened. Finally, the remaining articles were reviewed in- depth. To be included, every SR’s underlying study must have used an experimental or quasi-experimental design with a counterfactual and have been conducted in low- or lower-middle-income countries.3 To keep the review manageable, the review excluded SRs comparing the efficacy of different drugs or therapeutic interventions, except for treatment of common childhood diseases, such as diarrhea. SRs focusing on other childhood conditions, such as birth defects, were excluded. SRs focusing on HIV/AIDS and humanitarian contexts were excluded, due to concern for lack of generalizability. SRs with bundled interventions were also excluded if the effect of discrete interventions could not be meaningfully differentiated. SRs of interventions exclusively targeting adolescent girls and school feeding programs in primary school children were excluded. Further, interventions targeting outcomes primarily associated with the mother, with no clear link to outcomes of child undernutrition, were also excluded.4 Due to the high volume of relevant SRs found, an in-depth quality assessment of each SR, such as the 3ie and SURE checklist, was not conducted. Notwithstanding, minimum quality standards were ensured by including only peer-reviewed SRs from highly reputable databases and requiring underlying studies to use experimental or quasi- experimental study designs. A nutrition expert reviewed the search results, which were validated by the IEG evaluation team to ensure the quality and consistency of findings. Extraction and Synthesis of Evidence The following parameters were extracted from each SR article reviewed: study objective(s), design, and research setting; intervention target and components; method of 91 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio synthesis, number of underlying studies; method of quality assurance; outcomes measured; and the effect of each intervention. For meta-analyses, the pooled effects results were extracted, whereas for narrative SRs, results were extracted for each underlying study that fits the inclusion criteria. For “combination” interventions, such as education plus supplement, each component of the intervention received “credit” for the documented effect. After extracting the parameters, results from the SRs were synthesized in the SRM by intervention and outcome after the conceptual framework of child undernutrition. For the SRM, the effectiveness of interventions in improving the outcomes of interest were categorized as positive, negative, no effect, inconsistent, or no evidence, and tabulated across SRs.5 Results of the Literature Search and Description of Studies The initial search yielded 6,324 SRs, which were reduced to 227 (figure B.1). Most of the SRs included in the review came from PubMed (22 percent), 3ie (16 percent), Cochrane (15 percent), and eLENA (14 percent). Expert consultation contributed 15 percent of included SRs. Figure B.1. Stages of Systematic Review Identification Source: Independent Evaluation Group. The search strategy yielded 84 interventions that are synthesized in the SRM. Among these, 30 are nutrition-specific interventions. Of the remaining 54 nutrition-sensitive 92 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio interventions, 17 are implemented by the health sector; 16 by the agriculture sector; 12 by the WASH sector; and 9 by the social protection sector. In the next sections, the evidence on what works is summarized by outcome area based on the conceptual framework, organized by sector, intervention types, and target group. Findings for each outcome area are followed by their SRM visual. Systematic Review Map: What Works for Improving Nutrition Outcomes and its Determinants Reducing Child Undernutrition Birthweight: Evidence on the effectiveness of interventions targeting improvements in birthweight comes from both nutrition-specific (19 SRs) and nutrition-sensitive interventions (16 SRs). Among nutrition-specific interventions, strong and consistent evidence (6 SRs) supports the effectiveness of providing supplementary energy-dense foods to pregnant women (that is, micronutrients, including lipids, protein, and so on) to increase birthweight and reduce the risk of low birthweight (LBW). One SR also shows that supplementary feeding with energy-dense foods to children can have an intertemporal impact on the birthweight for the next generation. Another SR found that micronutrient supplementation with iron folate (iron–folic acid) to women reduces the risk of LBW. The evidence is less conclusive on the effectiveness of other micronutrient supplementation interventions for women, including iron (5 SRs), multiple micronutrients (MMNs; 4 SRs), and zinc (5 SRs), on improving birthweight outcomes, combining both positive and evidence of no effect. One SR shows positive effects of social and behavior change communication (SBCC) promoting nutrition and health practices through information and communication technology (ICT) on the risk of LBW, although another SR found no effect on such intervention in birthweight. Other micronutrient supplements for women, such as magnesium, omega 3, and vitamins A, C, and E, do not seem to be effective in improving birthweight. Among nutrition-sensitive interventions, in the health sector, the provision of insecticide-treated bed nets (ITNs; 1 SR), e-health communications between providers and beneficiaries (1 SR), and performance-based incentives (1 SR) reduce the risk of LBW, but the evidence is limited. Preventive deworming (3 SRs) had mixed results on birth outcomes (birthweight and risk of LBW). In the social protection sector, the impact of conditional cash transfers (CCTs) on birth outcomes is widely studied (6 SRs) and the evidence combines both positive and lack of effect results. 93 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Micronutrient status and deficiencies: There is abundant evidence on the effectiveness of interventions to improve children micronutrient status (50 SRs). Most of the evidence is for nutrition-specific interventions involving the direct supplementation of micronutrients to children. Among nutrition-specific interventions, supplementation with iodine and MMNs, and SBCC on nutrition and health practices improves children micronutrients status, yet the evidence is still scarce (1 SR for each intervention). Other supplementation of micronutrients, such as iron (9 SRs), micronutrient powders (MNPs; 6 SRs), zinc (4 SRs), and vitamin A (3 SRs) show mixed but mostly positive results on micronutrient status of children. Similar results are found for supplementary energy-dense foods (4 SRs), and foods rich in micronutrients (2 SRs) for children. The available evidence is less conclusive on the indirect effects of micronutrient supplementation to pregnant women. Although iodine supplementation to women (3 SRs) decreases the risk of child cretinism, vitamin A supplementation (2 SRs) does not appear to affect child micronutrient status and deficiencies. At the household level, the use of iron cookpots does not have a clear effect on micronutrients status (1 SR). Among nutrition-sensitive interventions, health interventions, such as delayed cord clamping (1 SR) and provision of ITNs (1 SR) are effective to reduce children anemia and improve hemoglobin concentration. The evidence is less conclusive, however, regarding the effectiveness of deworming on children’s micronutrient status. In agriculture, there is limited evidence on the effectiveness of nutrition-sensitive value chains (1 SR), small- scale livestock production (1 SR), irrigation (1 SR), provision of agriculture inputs and training (1 SR), and SBCC to reduce micronutrients deficiencies. Home gardening (5 SRs), fortification with iron (4 SRs) and vitamin A (4 SRs), however, shows mixed results. In social protection, three SRs found also mixed results (both positive and no effect) on the impact of CCTs on anemia, ferritin, and hemoglobin levels. Stunted and linear growth: This area comprises the largest body of evidence on effectiveness (71 SRs and 45 interventions). However, the SRM could not identify a single intervention with consistent and large amount of evidence to improve stunted and linear growth. Among nutrition-specific interventions, one SR found that SBCC on nutrition and health practices via community and support groups was an effective intervention to improve stunted and linear growth. SBCC interventions through other channels (such as education or promotion; growth monitoring and promotion, and home visits and peer support) offers less conclusive evidence. Within interventions targeting children, most of the evidence studied the effects of providing supplementary energy-dense foods (11 94 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio SRs) followed by zinc supplementation (9 SRs), supplementary feeding with micronutrient-rich food (7 SRs), MMNs (3 SRs), vitamin A (2 SRs), and iodine (1 SRs). The evidence for these interventions shows mixed results, yet with mostly positive findings. Within interventions in the health sector, few SRs found that family planning and contraception services, through their effects on birth spacing (1 SR), institutional strengthening policies (1 SR), and health insurance (1 SR) can contribute to reducing stunted growth. Deworming campaigns targeting children (6 SRs) and child stimulation (2 SRs) were found to have mixed results. Few nutrition-sensitive interventions in the agriculture sector seem to be effective in improving child growth, although the evidence remains limited. A meta-analysis found that consumption of biofortified quality protein maize led to an increase in the rate of growth in weight and height in infants and young children with mild to moderate undernutrition. Also, a significant and positive effect of land reforms conferring or providing land rights and autonomy to women in agricultural production was observed on the long-term nutritional status of women and child nutrition. The study revealed that a mother owning land halved the probability of her child being severely underweight. Home gardening (6 SRs), small-scale livestock production (3 SRs), and provision of agricultural inputs and training (1 SRs) are shown to have mixed results on improving stunted and physical growth. Other agriculture interventions, such as small-scale aquaculture (1 SR), fortification with iron, vitamin A or MMNs (1 SR), and cash cropping (1 SR), are shown to not be effective to improve stunted growth. In the social protection sector, the provision of daycare services, and the facilitation of access to microfinance, credit, and banking, were found to have mixed results. Notwithstanding the well-known positive impacts of CCTs on nutrition-related outcomes in the SRM, SRs show CCTs have an inconsistent effect on reducing stunted and physical growth. Evidence on the effect of nutrition-sensitive interventions in WASH is rather limited. One SR found evidence suggestive of a small benefit of improving quality of water supply, identifying a borderline statistically significant effect on height-for-age z score in children under five years old. Provision of latrines and potties for safe disposal of feces (4 SRs) and SBCC delivered through WASH (1 SR) show mixed results. Cognitive development: The review identifies 22 interventions with evidence on their effectiveness in affecting child cognitive development (29 SRs). Among nutrition-specific interventions, few SRs found positive effects of children supplementary feeding with micronutrient-rich food (1 SR), micronutrient supplementation with MMNs (1 SR), MNPs (1 SR), and SBCC via home visits or health 95 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio facility (interpersonal communication [IPC]; 1 SR) on cognitive development. The impact of supplementary energy-dense food (5 SRs), micronutrient supplementation with iron (5 SRs), and iodine (1 SR) show mixed but mostly positive results. Less is known about the impact of dietary support to pregnant women on child cognitive development. A few interventions have a mixed and sometimes positive effect, such as iodine supplementation for women (1 SR) and the provision of supplementary energy- dense foods for women (3 SRs) or micronutrient-rich food (1 SR). Regarding nutrition-sensitive interventions in the health sector, one SR provided evidence that the service integration, such as including training in early infant stimulation programs into existing health services, can have a more significant effect on the development of young children. There is substantial evidence on child deworming (5 SRs) and early stimulation (5 SRs) interventions, yet results are mixed but with mostly positive effects. In social protection, by improving maternal and care resources, the evidence of CCTs (3 SRs) is also emerging, showing mixed results in enhancing children cognitive development. The evidence on childcare is overall inconsistent. 96 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.2. Nutrition-Specific Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. ICT = information and communication technology; IPC = interpersonal communication; MMN = multiple micronutrients; MNP = micronutrient powder; SBCC = social and behavior change communication. 97 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.3. Nutrition-Sensitive Interventions: Health and Water, Sanitation, and Hygiene Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. CHW = community health worker; IMCI = integrated management of childhood illness; ITN = insecticide-treated bed net; SBCC = social and behavior change communication; WASH = water, sanitation, and hygiene. 98 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.4. Nutrition-Sensitive Interventions: Agriculture and Social Protection; with Institutional Strengthening Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. BFHI = Baby-Friendly Hospital Initiative; mHealth = mobile health; MMN = multiple micronutrients; SBCC = social and behavior change communication. 99 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Improving Child Feeding and Caregiving Behaviors Breastfeeding: There is substantial evidence (21 SRs) on interventions to improve breastfeeding practices. Among nutrition-specific interventions, strong and consistent evidence indicates that SBCC delivered via home visits and peer support (7 SRs), mass media communication (6 SRs), IPC at health facilities (4 SRs), or education and promotion (2 SRs) are effective interventions for improving breastfeeding practices. SBCC through other channels, such as community support (9 SRs) or ICT (4SRs), shows mixed results with mostly positive evidence. Within nutrition-sensitive interventions in the health sector, there is also limited but positive evidence for maternal emotional support (1 SR) and the deployment of community health workers (CHWs) to improve breastfeeding practices. The evidence is less consistent with respect to the effectiveness of health system strengthening interventions (5 SRs) and integrated management of childhood illness (IMCI; 1 SR). In the agriculture sector, there is limited but positive evidence of interventions, such as small-scale livestock production (1 SR) in combination with inputs and nutrition education (1 SR) to increase reported maternal practices in breastfeeding since cultural preferences toward some animal products (such as chicken and eggs) are believed to increase breast milk production. In the social protection sector, the evidence on provision of access to microfinance, credit, or banking (1 SR) is insufficient to draw conclusions. The pooled estimate from two studies suggests that conditional microcredit programs produce an average increase in the percentage of newborns receiving colostrum. Yet evidence from another two microcredit studies suggests no statistically significant effect on the prevalence of breastfeeding among children under two years. Complementary feeding: There is some evidence (19 SRs) on interventions to improve complementary feeding practices, comprising 19 interventions. Among nutrition-specific interventions, limited evidence suggests that supplementary feeding of children with micronutrient-rich foods (1 SR) and SBCC via different channels (5 SRs) improves diet quality and responsive feeding. Among nutrition-sensitive interventions in the health sector, 3 SRs provide consistent evidence that strengthening health systems can improve complementary feeding through dietary diversity, feeding frequency, and energy intake. One SR also shows that IMCI has a positive effect. There is substantial evidence that some nutrition-sensitive agriculture interventions can improve dietary intake and diversity. Home gardens (6 SRs), vitamin A fortification (3 SRs), livestock production (2 SRs), and provision of inputs and training (1 SR) interventions have shown positive effects on complementary feeding practices. Although 100 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio limited, the available evidence for cash cropping (1 SR) and irrigation (1 SR) interventions may deteriorate children’s diet. A study in Kenya found that children had the lowest intakes of energy, protein, and iron in their diets. Evidence from the social protection sector is highly variable. The provision of child daycare services (1 SR) may have a positive effect on complementary feeding. A study in Guatemala found a positive effect of a daycare program on the percentage of daily requirements consumed by children while at daycare (for example, energy, protein, iron, vitamin A), and child dietary intake improved while not at daycare. The effect of CCTs (3 SRs) on complementary feeding is too inconsistent to draw conclusions. Parenting practices: There is little evidence (8 SRs) on interventions to improve parenting practices (such as stimulation, interaction, and other nonfeeding skills). Nutrition-specific SBCC via IPC at the health facility (1 SR) was found to improve parenting practices in terms of skills and child stimulation. Other forms of SBCC (via community groups, home visits or peer support, or education and promotion) had mixed results with mostly positive effects. Although the evidence of nutrition-sensitive interventions in the health sector is limited, IMCI (1 SR), child stimulation (1 SR), and maternal emotional support (1 SR) had a positive effect in improving parenting practices, whereas the evidence of deployment of CHWs (2 SRs) is less consistent. No evidence was found on other nutrition-sensitive interventions to effectively improve parenting practices. Improving Child Health and Disease Status Enteric infection and diarrhea: Abundant evidence exists (50 SRs) on interventions to reduce the incidence, prevalence, or duration of child enteric infection and diarrheal diseases, although not always effective. Among nutrition-specific interventions, one SR found that SBCC via mass communication reduced the incidence of diarrhea. Child supplementation with zinc (9 SRs) was widely studied, showing a mixed but often positive effect. Evidence on the effects of child supplementation with vitamin A (4 SRs) is still limited to draw conclusions. The evidence suggests that other interventions such as child supplementation with iron (3 SRs) and MNPs (3 SRs) may potentially increase diarrhea duration. Among nutrition-sensitive interventions in the health sector, on SR shows that infants whose mothers were treated to reduce maternal depression, through maternal emotional support, experienced fewer episodes of diarrhea. Contracting out service provision through performance-based financing (1 SR), was also associated with lower incidence of diarrhea 101 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio among young children, although evidence is scarce. In agriculture, biofortification of staple crops that are richer in essential micronutrients like vitamin A (1 SR) has been shown to reduce the prevalence and duration of diarrhea in children younger than five years, supporting the well-known role of vitamin A in protecting immunity. Evidence of home gardening (4 SRs) shows rather mixed results. In social protection, one SR found significant reductions in the prevalence of diarrhea for children with longer exposures to daycare services, whereas the evidence for CCT (2 SRs) is rather limited and inconclusive. In the WASH sector, provision of safe water storage (3 SRs) appears to reduce rates of enteric infection and diarrhea. Also, substantial and mostly positive evidence exists for the provision of latrines and potties for safe disposal of feces (10 SRs), point-of-use water treatment (10 SRs), source water treatment (7 SRs), and SBCC (9 SRs). There is less but still mostly positive evidence for provision of insect control (1 SR), piped water (4 SRs), sewerage (2 SRs), and community water supply (3 SRs). The evidence on the provision of soap (2 SRs) is too inconsistent to draw conclusions. Childhood illness and infection: There is also abundant evidence (45 SRs) on the effectiveness of interventions aiming to reduce childhood illness and infections. Among the nutrition-specific interventions, two SRs reported positive impacts of SBCC education on nutrition and health practices on reducing the incidence of respiratory tract infections in young children. The evidence offers mixed results for the effects of supplementation with zinc (8 SRs), MNP (3 SRs), and supplementary feeding with micronutrient-rich-food (3 SRs) on reducing childhood illnesses. The evidence for nutrition-sensitive health interventions is somewhat limited. One SR found that performance-based financing improved parent-reported health status among children under five discharged after treatment for pneumonia. The provision of ITNs (2 SRs), health insurance (1 SR), and health system strengthening (1 SR) have rather limited evidence and mixed results on their effectiveness to reduce child illnesses. Within the agriculture sector, home gardening (5 SRs) is the most widely studied intervention. The results are mixed, but for respiratory tract infections especially, mostly positive. For other agriculture interventions (cash cropping, small-scale livestock production, provision of inputs and training, and irrigation), the evidence is limited and mixed. In the social protection sector, the evidence for CCTs (3 SRs), unconditional cash transfers (UCTs) (1 SR), and child daycare services (1 SR) is mostly positive, although rather limited. Within the WASH sector, the provision of sanitation (such as latrines and potties; 3 SRs) and SBCC (3 SRs) receive the most attention. The evidence for sanitation is too inconsistent to draw conclusions. For SBCC, the evidence is mixed but mostly positive. Provision of soap (1 SR) also seems to have a positive effect on childhood illness and infection. 102 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.5.Child Feeding and Caregiving: Nutrition and Dietary Support Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. ICT = information and communication technology; IPC = interpersonal communication; MMN = multiple micronutrients; MNP = micronutrient powder; SBCC = social and behavior change communication. 103 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.6. Child Feeding and Caregiving: Health and Water, Sanitation, and Hygiene Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. CHW = community health worker; IMCI = integrated management of childhood illness; ITN = insecticide-treated bed net; SBCC = social and behavior change communication; WASH = water, sanitation, and hygiene. 104 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.7. Child Feeding and Caregiving: Agriculture and Social Protection Interventions; with Institutional Strengthening Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. BFHI = Baby-Friendly Hospital Initiative; mHealth = mobile health; MMN = multiple micronutrients; SBCC = social and behavior change communication. 105 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Improving Maternal Health Status Nutrition status and deficiencies: There is substantial evidence for interventions aiming to improve maternal nutrition status and reduce micronutrient deficiencies (30 SRs), comprising 24 interventions. Among nutrition-specific interventions, strong and consistent evidence shows that iron folate supplementation to women (3 SRs) reduces anemia and improves ferritin and hemoglobin levels. A lot of attention has been paid to maternal supplementation with iron alone (5 SRs), vitamin A (6 SRs), and MMNs (3 SRs), most of which show positive effects. For maternal zinc supplementation (1 SR) and provision of supplementary energy-dense foods (3 SRs), the results are too inconsistent to draw conclusions. Among nutrition-sensitive interventions in the health sector, there is mixed evidence with mostly positive results on the effectiveness of deworming during pregnancy (4 SRs) on anemia and hemoglobin levels. Although limited, the evidence for use of e-health communications by health workers (such as for data collection, reporting, and decision- making; 1 SR) and prophylactic medication (such as intermittent preventive treatment for malaria; 1 SR) also appear to contribute to improved maternal nutrition status and deficiencies outcomes. In agriculture, the evidence is limited but positive effects were found for other agriculture interventions, such as small-scale livestock production (1 SR), land property rights (1 SR), and inputs and training (2 SRs). Most of the evidence concentrates in home gardening (6 SRs) reporting mixed results for their potential to improve maternal nutrition status and deficiencies. For fortification with folic acid (1 SR), iodine plus iron (1 SR), iron alone (2 SRs), and MMNs (1 SR), as well as nutrition promotion delivered through the agriculture sector (1 SR), the evidence is too limited or inconsistent to draw conclusions. Nutrient intake and dietary diversity: Little evidence is available for interventions aiming to affect maternal nutrient intake and dietary diversity (7 SRs). Scarce evidence shows that SBCC via mass communication delivered through the health sector (1 SR) was effective to improve women diet. Among nutrition-sensitive interventions, most of the evidence is for home gardening (5 SRs), which demonstrated a mostly positive effect on maternal diet quality and micronutrients intake. The evidence for other nutrition-sensitive agriculture interventions is more limited. Vitamin A fortification (1 SR) and provision of agriculture inputs and training (1 SR) appear to have a positive effect on maternal nutrient intake and dietary diversity. The evidence for small-scale aquaculture (1 SR) and SBCC delivered through the agriculture sector (3 SRs) is too limited or inconsistent to draw conclusions on the effect on maternal 106 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio nutrient intake and diet diversity. In social protection sector, one SR shows that CCTs (1 SR) can be an effective intervention to improve diet intake and diversity. Healthy pregnancy: There is some evidence on interventions aiming to promote healthy pregnancy (18 SRs), comprising 15 interventions. The evidence for nutrition-specific interventions suggests that SBCC through community support (1 SR) or mass communication (2 SRs) can reduce maternal morbidity during pregnancy, yet the evidence is limited. Supplementary energy-dense foods for women (3 SRs) seem to prevent giving birth to babies that are small for gestational age. The evidence for maternal supplementation with antioxidants (such as vitamin C, vitamin E, or selenium; 1 SR), calcium (2 SRs), MMNs (3 SRs), and zinc (2 SRs) offers mixed results from which it is impossible to draw conclusions. There is no evidence of effect on gestational growth from maternal supplementation with folic acid (1 SR), iron (1 SR), or vitamin A (1 SR). Evidence for nutrition-sensitive interventions is limited. The provision of ITNs (2 SRs) to reduce malaria incidence during pregnancy and the use of prophylactic medication (2 SRs) to reduce the risk of preterm birth are each inconclusive. In the WASH sector, although the evidence is limited, the provision of clean cookstoves (1 SR) appears to decrease respiratory tract infections for women through reducing household air pollution. Mental health: There is some evidence on interventions aiming to improve maternal mental health (such as anxiety, confidence, depression, self-esteem, and stress; 12 SRs), comprising 12 interventions. The strongest evidence is for SBCC through home visits or peer support (4 SRs). There is also more limited evidence that IPC delivered through the health sector (2 SRs), maternal emotional support (2 SRs), and health system strengthening (1 SR) can work to improve maternal mental health. In agriculture, a study in East Africa found that a farmer field school program in Kenya, Tanzania, and Uganda, which used education and training as a tool to support capacity development, increased competence and enhanced well-being of participating women. Limited evidence shows some positive effects from social protection interventions. Two SRs show that CCT can improve maternal emotional health. Women exposed to the Mexican Oportunidades program had lower depressive symptom scores, and in Brazil and Nicaragua CCT programs also showed improvements in women’s enhanced self-esteem. A SR on the impacts of microfinance programs in South Asia revealed that the duration and depth of involvement in microfinance activities would make a difference in women’s mental health and not just receiving loans (for example, lower levels of self-reported emotional stress, higher autonomy). 107 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.8. Maternal Health: Nutrition and Dietary Support Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. ICT = information and communication technology; IPC = interpersonal communication; MMN = multiple micronutrients; MNP = micronutrient powder; SBCC = social and behavior change communication. 108 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.9. Maternal Health: Health and Water, Sanitation, and Hygiene Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. CHW = community health worker; IMCI = integrated management of childhood illness; ITN = insecticide-treated bed net; SBCC = social and behavior change communication; WASH = water, sanitation, and hygiene. 109 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.10. Maternal Health: Agriculture and Social Protection Interventions; with Institutional Strengthening Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. BFHI = Baby-Friendly Hospital Initiative; mHealth = mobile health; MMN = multiple micronutrients; SBCC = social and behavior change communication. 110 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Increasing Maternal and Childcare Resources Household welfare: There is ample evidence on the effectiveness of interventions aiming to improve household welfare (27 SRs), comprising 22 interventions. Among nutrition-sensitive interventions in the health sector, there is strong and consistent evidence that health systems reforms (3 SRs), such as contracting out service provision and prospective payments lower individual out-of-pocket expenditures having an indirect effect on households’ income. However, evidence also shows that health insurance (3 SRs) and performance payments (1 SR) may have unintended effects on income. For example, the combination of performance payments and reduced user fees to attract patients negatively affects health facilities budgets and therefore staff salaries. In agriculture, there is strong and consistent evidence that small-scale aquaculture (3 SRs) improves household income and welfare. There is limited evidence that other agriculture interventions also have a positive effect, such as contract farming (1 SR), land property rights (1 SR), provision of irrigation (1 SR), and nutrition promotion (1 SR). The evidence for home gardening (3 SRs), small-scale livestock production (2 SRs), and agriculture inputs and training (2 SRs) is mixed but with mostly positive effects on households’ assets and income. In social protection, much attention has been paid to provision of access to microfinance, credit, and banking (9 SRs) and CCTs (4 SRs); however, the evidence for both is too inconsistent to draw conclusions. The evidence for other social protection interventions, such as UCTs (3 SRs), vouchers or other in-kind subsidies (4 SRs), community block grants (1 SR), and SBCC (3 SRs) is also mixed but mostly positive. Schooling: There is some evidence on interventions aiming to increase schooling (17 SRs), comprising 15 mostly nutrition-sensitive interventions. Among nutrition-specific interventions, provision of supplementary energy-dense (1 SR) and micronutrient-rich (2 SRs) foods for women seems to have a positive effect on education enrollment and attainment, whereas the evidence on provision of supplementary energy-dense (1 SR) and micronutrient-rich (1 SR) foods for children is too inconsistent to draw conclusions. Among nutrition-sensitive interventions in the health sector, the evidence is limited yet it suggests that child stimulation (2 SRs), user fee elimination or reduction (1 SR), and health insurance (1 SR) can improve schooling outcomes. Deworming (2 SRs), however, does not appear to be an effect on school attainment. In social protection, CCT (6 SRs) is the most effective intervention to improve school enrollment, attendance, and attainment. The evidence for other social protection interventions, such as UCTs (3 SRs), 111 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio vouchers and other in-kind subsidies (4 SRs), and SBCC (3 SRs) is mixed but mostly shows a positive effect. There is also much evidence for provision of access to microfinance, credit, or banking (6 SRs), but it is too inconsistent to draw conclusions. Knowledge and attitudes: There is substantial evidence on interventions that aim to affect knowledge and attitudes (29 SRs), comprising 29 interventions. Numerous interventions delivered through the health sector appear to be effective in improving women’s knowledge and attitudes toward nutrition practices. The evidence for SBCC delivered through the health sector (11 SRs) is mixed but mostly positive, except for health and nutrition promotion via ICT (1 SR), which thus far does not seem to influence knowledge and attitudes. Other interventions delivered through the health sector, such as user fee elimination or reduction (1 SR), health insurance (1 SR), health system strengthening (1 SR), CHWs (1 SR), health facility-to-community outreach (1 SR), child stimulation (i), maternal emotional support (1 SR), and family planning and contraception services (1 SR), have limited but still positive evidence. For instance, a quasi-experimental evaluation showed that an adolescent health program in Ethiopia supporting basic medical care services free of charge increased young girls’ knowledge of HIV and where to get tested for HIV. Another study in the Arab Republic of Egypt showed that a program supporting adolescents to obtain health insurance improved family planning knowledge and successfully changed attitudes about family size. In Benin, health system strengthening through activities like task shifting (which involves equipping a cadre of staff with the appropriate skills to provide services that would otherwise be provided by higher cadre providers, who are often scarce), was an effective intervention to improve maternal knowledge on prenatal care, birth preparedness and recognition of danger signs. Among nutrition-sensitive interventions in agriculture, there is consistent evidence indicating the effectiveness of agriculture inputs and training (3 SRs) nutrition-related knowledge and behavior, such as improved knowledge on balanced diets. However, there is limited evidence that home gardening (1 SR) and small-scale livestock production (1 SR) have a positive effect. In the social protection sector, CCTs (3 SRs) is the most studied intervention with consistent evidence on effectiveness. Although the evidence is limited, community block grants (1 SR) and UCTs (1 SR) also appear to positively affect knowledge and attitudes There is also positive evidence for vouchers and other in-kind subsidies (4 SRs); however, results are mixed though mostly positive. In the WASH sector, provision of soap (1 SR) and SBCC (4 SRs) on handwashing promotion seems to improve knowledge and attitudes toward better hygiene practices among households. 112 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Household safety: There is little evidence on interventions that aim to affect household safety (6 SRs). The most abundant evidence, from the social protection sector, is for provision of access to microfinance, credit, or banking (4 SRs), for which the evidence is mixed but mostly positive in reducing intimate partner violence. There is also limited evidence that child stimulation (1 SR), CCTs (1 SR), UCTs (1 SR), and SBCC (1 SR) an improve household safety outcomes. 113 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.11. Maternal and Childcare Resources: Nutrition and Dietary Support Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. ICT = information and communication technology; IPC = interpersonal communication; MMN = multiple micronutrients; MNP = micronutrient powder; SBCC = social and behavior change communication. 114 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.12. Maternal and Childcare Resources: Health and Water, Sanitation, and Hygiene Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. CHW = community health worker; IMCI = integrated management of childhood illness; ITN = insecticide-treated bed net; SBCC = social and behavior change communication; WASH = water, sanitation, and hygiene. 115 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.13. Maternal and Childcare Resources: Agriculture and Social Protection Interventions; with Institutional Strengthening Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. BFHI = Baby-Friendly Hospital Initiative; mHealth = mobile health; MMN = multiple micronutrients; SBCC = social and behavior change communication. 116 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Increasing Access to Health Services Maternal use of health services: There is substantial evidence on interventions aiming to improve maternal use of health services (31 SRs). SBCC delivered through the health sector (16 SRs) though different channels has received the most attention. The evidence shows mixed results, but with mostly positive effects on care-seeking behaviors and adherence to care. Relatively limited evidence also indicates that CHWs (1 SR), health facility-to-community outreach (1 SR), service integration (1 SR), and maternal emotional support (1 SR) can improve maternal use of health services through enhancing care-seeking behaviors. However, the evidence shows mixed results, but mostly positive impacts for user fee elimination or reduction (5 SRs), health insurance (5 SRs), system strengthening (5 SRs), and e-health communications (4 SRs). In the social protection sector, CCTs (8 SRs) and vouchers and other in-kind subsidies (6 SRs) seem to increase the use of preventive and curative health and nutrition services. Child use of health services: Similarly, substantial evidence exists on interventions (22) aiming to improve child use of health services (33 SRs). In the health sector, SBCC via education, growth monitoring and promotion, IPC at health facilities or ICTs are shown to increase care-seeing behavior and in particular the use of immunization services, although the extent of the evidence is still limited. Three SRs report consistent evidence on the effectiveness of CHWs to increase the use of child health care services. There is also limited but positive evidence for health facility-to- community outreach (1 SR), e-health communications (1 SR), and maternal emotional support (1 SR) to improve the use of immunization services. The evidence for other interventions in the health sector, such as service integration (4 SRs), IMCI (1 SR), user fee elimination or reduction (1 SR), and health insurance (1 SR) offers rather mixed results is inconclusive. Among nutrition-sensitive interventions delivered through the social protection sector, CCTs has been widely studied (6 SRs) and similar to the use of maternal health care services, CCTs programs and vouchers and other in-kind subsidies (2 SRs) can improve the uptake of child health care services. However, child daycare (1 SR) interventions may potentially reduce vaccination rates. A study of a daycare program had an unexpected negative impact on the proportion of children who were completely immunized. 117 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Increasing Access to Water, Sanitation, and Hygiene Services There is some evidence on the effectiveness of interventions aiming to improve access to WASH services (10 SRs). SBCC interventions have been widely studied. SBCC via community or support groups (2 SRs) and home visits (1 SR) seems to improve access to WASH services; whereas results for other SBCC channels (6 SRs) are more variable but sometimes positive. Although the evidence is limited, provision of point-of-use water treatment (1 SR), soap (2 SRs), and community water supply (1 SR) also seem to have a positive effect. The evidence on the provision of sanitation (3 SRs), source water treatment (1 SR), and insect control (1 SR) is too limited or less consistent to draw conclusions. Increasing Access to Nutrient-Rich Food There is some evidence on interventions that aim to improve access to nutrient-rich food (18 SRs), comprising 14 interventions. Strong and consistent evidence shows that CCTs (5 SRs) are effective for improving access to nutrient-rich food. Among social protection interventions, the evidence on provision of access to microfinance, credit, or banking (3 SRs) is mixed but mostly positive, and SBCC delivered through the social protection sector (1 SR) also seems effective. The evidence on UCTs (1 SR) is inconclusive. Among nutrition-sensitive interventions in the agriculture sector, limited evidence indicates that provision of inputs and training (2 SRs), contract farming (1 SR), and fortification with iron (1 SR) and vitamin A (1 SR) can improve households’ access to nutrient-rich food. Similarly, small-scale livestock production (4 SRs) and aquaculture (3 SRs) interventions have shown to increase households’ food security although results are mixed. Home gardens (5 SRs) has been widely studied with mostly positive impacts on household food consumption, although in a handful of studies declines in household pulse consumption (dry beans, dry peas, lentils, and chickpeas), and lower consumption of staple cereals and animal food were also observed. The evidence for irrigation (3 SRs) and cash cropping (1 SR) is rather limited and less consistent to draw conclusions. 118 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.14. Nutrition-Specific Interventions: Nutrition and Dietary Support Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. ICT = information and communication technology; IPC = interpersonal communication; MMN = multiple micronutrients; MNP = micronutrient powder; SBCC = social and behavior change communication; WASH = water, sanitation, and hygiene. 119 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.15. Nutrition-Sensitive Interventions: Health and Water, Sanitation, and Hygiene Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. CHW = community health worker; IMCI = integrated management of childhood illness; ITN = insecticide-treated bed net; SBCC = social and behavior change communication; WASH = water, sanitation, and hygiene. 120 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.16. Nutrition-Sensitive Interventions: Agriculture and Social Protection Interventions; with Institutional Strengthening Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. BFHI = Baby-Friendly Hospital Initiative; mHealth = mobile health; MMN = multiple micronutrients; SBCC = social and behavior change communication; WASH = water, sanitation, and hygiene. 121 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Changing Social Norms Women’s empowerment: There is some evidence on interventions aiming to effect women’s empowerment (19 SRs), comprising 16 nutrition-sensitive interventions. Among nutrition-sensitive interventions in the health sector, there is limited evidence on the effects of SBCC interventions on improving women’s empowerment and decision- making. There is also limited evidence in the health sector that facility-to-community outreach (1 SR), user fee elimination or reduction (1 SR), and health insurance (1 SR) may improve women’s empowerment, in theory by reducing the barriers to care seeking. In the agriculture sector, the limited evidence on home gardening (2 SRs)—in theory, from income earned and capacity to feed nutritious homegrown foods—mostly indicates positive effects on women’s income, control over resources, or influence in decision-making on a range of issues. However, in some cases like a project promoting orange sweet potato production in Kenya among women farmers showed that women gained control over selling the product, whereas men maintained control over income. Evidence on land property rights (1 SR) is limited but still suggests that land reforms conferring or providing land rights and autonomy to women in agricultural production have a positive influence on women’s empowerment. Among social protection interventions it is expected that the improved access to financial resources would enhance women’s power within the household to make and act on decisions that benefit her and her children. Microfinance, credit, or banking interventions (10 SRs) have been widely studied, and although the evidence is mixed, mostly positive effects were observed on women’s empowerment, control of resources and assets, decision-making, and reduced risk of interpersonal violence. There is also evidence that vouchers or other in-kind subsidies (2 SRs) have a positive effect, whereas the evidence on CCTs (3 SRs) and UCTs (1 SR) is rather limited to draw conclusions. Early pregnancy: There is little evidence on interventions that aim to affect early pregnancy (8 SRs), comprising 12 interventions. In the health sector, limited evidence suggests that user fee elimination or reduction (2 SRs) and family planning and contraception services (1 SR) are protective against early pregnancy. This is likely due to easier access to contraception counseling or methods. The evidence on the effect of SBCC delivered through the health sector (2 SRs) is mostly positive. 122 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio In social protection, transfers (cash or in-kind) are expected to reduce the financial pressures on households that contribute to early marriage and thereafter early pregnancy. Most of the available evidence is for CCTs (5 SRs) and vouchers and other in-kind subsidies (4 SRs), which show mixed but mostly positive effects on delaying early pregnancy and marriage. The evidence on provision of access to microfinance, credit, or banking (1 SR), UCTs (2 SRs), and SBCC delivered through the social protection sector (1 SR) is too limited or inconsistent to draw conclusions. Birth spacing: There is some evidence on the effectiveness of interventions aiming to affect birth spacing (24 SRs), comprising 20 interventions. In the health sector, there is consistent evidence showing that provision of family planning and contraception services (4 SRs) promotes birth spacing. There is also some evidence for other health sector interventions, such as CHWs (4 SRs), SBCC (7 SRs), service integration (3 SRs), and system strengthening (3 SRs), which shows a mixed but mostly positive effect. The evidence for health facility-to-community outreach (1 SR) is limited but positive. In the social protection sector, the evidence on CCTs (7 SRs), UCTs (2 SRs), and vouchers or other in-kind subsidies (3 SRs) is too inconsistent to draw conclusions. 123 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.17. Nutrition-Specific Interventions: Social Norms and Behaviors Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. ICT = information and communication technology; IPC = interpersonal communication; MMN = multiple micronutrients; MNP = micronutrient powder; SBCC = social and behavior change communication. 124 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.18. Nutrition-Sensitive Interventions: Health and Water, Sanitation, and Hygiene Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. CHW = community health worker; IMCI = integrated management of childhood illness; ITN = insecticide-treated bed net; SBCC = social and behavior change communication; WASH = water, sanitation, and hygiene. 125 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.19. Nutrition-Sensitive Interventions: Agriculture and Social Protection Interventions; with Institutional Strengthening Interventions Source: Independent Evaluation Group. Note: See the “Notes for Figures B.2–B.19 section for an explanation of superscript numbers. BFHI = Baby-Friendly Hospital Initiative; mHealth = mobile health; MMN = multiple micronutrients; SBCC = social and behavior change communication. 126 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Systematic Review Map Discussion The range of what works for reducing child undernutrition and improving nutrition determinants outcomes is broad. Here, 227 SRs are reviewed. Studies cover nutrition- specific interventions, nutrition-sensitive interventions, or both. Of the 84 interventions included in the SRM, 30 are nutrition-specific interventions, and of the remaining 54 nutrition-sensitive interventions 17 are implemented by the health sector, 16 by the agriculture sector, 12 by WASH sector, and 9 by social protection sector. Interventions with a Broad Positive Impact Across the nine nutrition-relevant outcome areas reviewed, some interventions show consistent positive influence across two or more outcome areas. These interventions would be well suited for programs aiming to achieve impact at immediate and underlying levels, and even sometimes nutrition outcomes (box B.1). Box B.1. Interventions with a Broad Positive Impact Nutrition-Specific • Child supplementary feeding with micronutrient-rich foods • Maternal supplementary feeding with energy-dense foods • Women micronutrient supplementation: iron folate (iron–folic acid) • Social and behavior change communication of nutrition and health promotion (via community and groups, education, growth monitoring and promotion, home visits, mass communication, and interpersonal communication at health facility) Nutrition-Sensitive Health • Health system strengthening • Maternal emotional support • Family planning and contraception • Health care approach: community health workers • Health facility community outreach • E-health communication • Health insurance Agriculture • Provision of agriculture inputs and training • Small-scale livestock 127 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio • Vitamin A fortification • Land property rights Social Protection • Conditional cash transfers WASH • Provision of soap Source: Independent Evaluation Group. Nutrition and health: For nutrition-specific interventions the evidence suggests that children supplementary feeding with micronutrient-rich foods, maternal supplementary feeding with energy-dense foods, and iron folate (iron–folic acid) supplementation have a broad positive impact across outcomes areas. Also, SBCC on health and nutrition promotion via different channels have the most widespread positive impact across several outcome areas. Health systems strengthening and maternal emotional support interventions are shown to be widely effective. Agriculture: Interventions, such as vitamin A fortification, provision of agriculture inputs and training, land property rights, and small-scale livestock, have a broad positive influence, highlighting the importance of agriculture in providing essential nutrients for home consumption and in affecting social norms. It is notable, however, that among agriculture interventions, cash cropping is particularly prone to negatively affecting child undernutrition and its associated outcomes. The cause for the consistently negative effect is often attributed to traditional gender roles. In low- and middle-income country contexts, men are more often responsible for—and receive the income from— cash crop farming, whereas income earned by women is more likely to be put toward nutritious food consumption. Social protection: Regarding interventions that cut across multiple outcome areas, the best bet from the social protection sector is CCTs. However, along with provision of access to microfinance, credit, or banking, CCTs are also among the nutrition-sensitive interventions that have commonly demonstrated potential to do harm. This may indicate the need for more research to understand the contextual factors—or other program attributes—associated with positive and negative impact, and to account for them in program design. Also, although SBCC is among the most effective interventions when delivered by the health sector, the evidence is weak for the effectiveness of SBCC when delivered through the social protection sector. This finding underscores the importance of multisector collaboration in the delivery of high-quality health and nutrition messaging. 128 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio WASH: Overall, the evidence for improving nutrition through WASH is limited. The WASH sector covers a wide array of interventions to improve access, knowledge, and practices regarding the use of latrines, safe feces disposal, water supply and quality, hand washing, and pest control. From a nutrition-sensitive perspective, the goal of WASH interventions is to reduce exposure to the germs that cause child disease, especially diarrhea, which closely relates to malnutrition. Among the interventions included in this study, provision of soap is the key WASH intervention, with positive effects across multiple outcome groups. Interventions with a Consistent Impact on Particular Outcomes Some interventions have positive and consistent evidence for just one or two outcomes. Many of these are nutrition-specific interventions that target the mother, such as maternal supplementation with iodine (to reduce the risk of cretinism), iron folate (to improve maternal nutrition status and reduce micronutrient deficiencies), and energy- dense food (to increase birthweight), and SBCC (to improve breastfeeding practices and maternal mental health). From the health sector, system strengthening interventions are effective for improving complementary feeding practices and household welfare, CHWs are beneficial for improving child use of health services, and family planning and contraception services are best for improving birth spacing. Within the agriculture sector, small-scale aquaculture is a consistently effective intervention for improving household welfare, and provision of inputs and training is best for improving knowledge and attitudes. Best options from the nutrition-sensitive social protection and WASH sectors are a little less clear, since few social protection or WASH interventions offer strong and consistent evidence. In the social protection sector, CCTs have proven effective for improving household access to nutrient-rich food, schooling, and knowledge and attitudes. Other promising interventions include provision of child daycare services and UCTs. In the WASH sector, provision of safe water storage appears to be the best option for reducing child enteric infection and diarrhea. Other promising interventions include provision of soap and community water supply. That most interventions only have a narrow area of focus underscores the importance of designing programs with a clear understanding of which outcomes they aim to improve. Also, the few interventions that have the potential for negative effects underscore the importance of monitoring and designing programs to prevent unintended harmful effects. Many of the most effective interventions target the mother, underscoring the 129 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio importance of engaging women early (preconception) and across all stages of early child development. Numerous interventions across all sectors show great promise, but more research into their impact on child nutrition and associated outcomes is needed. From the nutrition sector, this includes child supplementation with MMNs, MNPs, and maternal supplementary feeding with micronutrient-rich food. From the health sector, this includes IMCI, delayed cord clamping, health facility-to-community outreach, service integration, maternal emotional support, prophylactic medication during pregnancy, and health and nutrition promotion through information, communication, and technology. For agriculture, this includes contract farming, home gardening, nutrition- sensitive value chains, small-scale livestock production, fortification with iron and protein, and irrigation. For social protection, this includes community block grants and nutrition promotion delivered through the social protection sector. In the WASH sector, this includes only safe water storage. Gaps in Knowledge The SRM highlights areas where further learning may be important to improve the implementation of evidence-based interventions. For some outcomes, numerous interventions are evaluated but with few clear winners. This includes birth spacing, gestational growth, early pregnancy, (nonfeeding) parenting practices, birthweight, and maternal use of health services. Either more innovative approaches, information about the contextual constraints and factors of success, or improved program implementation is needed. The least studied outcomes are household safety, maternal nutrient intake and dietary diversity, (nonfeeding) parenting practices, prevention of early pregnancy, and maternal health. Limitations of the Systematic Review Map The scope of the literature review is comprehensive and ambitious, focusing on child undernutrition outcomes and its determinants guided by the conceptual framework. A strength of this review is that no interventions were ruled out ex ante. However, several limitations are noteworthy. • The review is only based on SR evidence due to time constraints. SRs are conducted after several individual impact evaluation studies have been published. Consequently, this literature review may omit relevant nutrition- specific or nutrition-sensitive interventions that are not yet included in an SR. 130 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio • The review may be subject to publication bias. To ensure that the results of this literature review reflect high-quality research, “gray” literature (project reports, reports from international organizations, unpublished dissertations, and so on) is not included. It is then possible that relevant SRs not included in the searched databases are missed. • There is a risk of duplication of the underlying studies included in the SRs. This would happen if a single impact evaluation is included in multiple SRs. The longer the maturity of the intervention and the higher the quality of the impact evaluation, the more likely the risk of duplication. Consequently, if the effect detected by the underlying study is strong, it may skew the results reported in this review in either a positive or negative direction. • Some outcomes represent broader concepts that may have been operationalized differently by different studies. The interpretation of specific outcomes, such as fat intake and energy intake, might change depending on the context. Since the focus of this literature review is undernutrition in low- and lower-middle-income countries, increased fat intake and increased energy intake are both considered pronutrition effects. • Evidence on the effectiveness of interventions should be interpreted exclusively for the nutrition-related outcomes of interest. For example, folic acid supplementation for women is found to have no effect on nutrition-related outcomes but is known to be highly effective for prevention of birth defects (neural tube defects), which are not included as outcomes of interest in this literature review. • Several SRs report results on combined, potentially synergistic interventions, such as provision of soap and point-of-use water treatment, health and nutrition promotion via home visits, community support groups, and mass media. Thus, it becomes challenging, if not impossible, to attribute the effect to any single intervention and addressing all combinations of interventions quickly becomes unwieldy. Bundling of this sort is more common for SBCC and nutrition- sensitive interventions in the agriculture, social protection, and WASH sectors. Alignment of the World Bank Nutrition Portfolio with the Global Evidence on What Works The SRM provides a useful visual of what works to help decision makers invest in those interventions that have been proved to be effective. In this section, we assess the extent 131 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio to which the interventions supported by the World Bank’s lending nutrition portfolio align with the evidence summarized in the SRM. Scope and Methodology The alignment analysis focuses on the intersection of the SRM and the World Bank’s nutrition portfolio. Therefore, interventions that are not reflected in the SRM are outside the scope. The interventions supported through the World Bank’s portfolio are individually classified according to the SRM’s areas of nutrition and health, social protection, water, agriculture, and institutional strengthening in the health sector, for which there is evidence of their effectiveness. Overall, the alignment analysis covers about half of the portfolio (47 percent; figure B.20). The remaining interventions could not be mapped to the SRM either because it is outside the SRM set or the intervention description did not provide sufficient detailed information. Twelve out of 84 intervention types of the SRM are not found in the nutrition portfolio. Since the evidence on what works for a particular intervention could vary for different outcomes (that is, the provision of zinc micronutrient supplementation to children is shown to improve micronutrients status of children, but has no effect on cognitive development outcomes) the alignment analysis makes the assumption that the World Bank’s interventions were meant to affect the outcomes reflected in projects’ results frameworks. Therefore, the analysis uses indicators classification from the portfolio review as proxies for intended outcomes. Three outcomes regarding the effectiveness of arrangements to deliver interventions—the efficiency of nutrition policies; financing and coordination; and the strength of stakeholder engagement and ownership—are excluded from the analysis since they were outside the scope of the SRM. The alignment analysis therefore expresses the World Bank’s portfolio as a combination of interventions- outcomes to be mapped to the existing evidence summarized in the SRM. 132 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.20. Scope of the Alignment Analysis Source: Independent Evaluation Group. Note: WASH = water, sanitation, and hygiene. Findings The World Bank largely supports nutrition interventions that are known to work, because the nutrition portfolio as a whole mainly focuses on interventions that have positive evidence of effectiveness to improve the nutrition-relevant outcomes of interest. This holds for each of the intervention areas (nutrition, health, social protection, agriculture, WASH, and institutional strengthening; figure B.21). Nutrition and health: Within the nutrition-specific interventions with consistent evidence of effectiveness, the World Bank’s efforts mainly concentrate on supporting SBCC on nutrition and health practices known to work across different nutrition- relevant outcome areas. Other health interventions where the World Bank highly aligns with the literature are supporting health care approaches that implement health facilities outreach activities; the deployment of CHWs; and family planning and contraception services. Consistent with the findings of the portfolio review, the World Bank largely focuses on institutional strengthening support to improve the health system, expand health insurance, and implement performance-based financing and service integration approaches that the global evidence base shows to be effective for improving particular nutrition-relevant outcomes. 133 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Agriculture: The most frequent agriculture intervention supported by the World Bank’s nutrition portfolio is the provision of inputs and training. Biofortification of foods and the support for small-scale livestock production are also prominent within the group of interventions with consistent positive evidence of effectiveness. Social protection: The World Bank aligns well with the evidence on what works by mainly focusing on supporting country cash transfer programs, which had positive effects in improving household food security and welfare, schooling attendance, health care use and child health and nutrition dietary practices. Support for the access to center- or home-based care services has been shown to be effective to improve complementary feedings and child health outcomes. WASH: Within the most effective interventions according to evidence, the World Bank’s support in the WASH sector has mainly focused on SBCC to promote handwashing and safe drinking water, community water supply through standpipes or hand pumps, safe water storage, and provision of soap. These interventions have consistent evidence of effectiveness in improving access to safe water, or household knowledge and attitudes, or reducing the incidence of childhood illness and diarrhea. 134 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Figure B.21. Alignment of World Bank’s Nutrition Interventions with Evidence on What Works Source: Independent Evaluation Group; systematic review map and portfolio review. Note: Values show the percentage of the total intervention outcomes represented by each intervention –evidence type combination. Positive indicates that the pooled effect (for meta-analyses) or all underlying studies (for narrative syntheses) of the intervention are found to have a positive effect on the outcome of interest. No effect indicates that the intervention is neither significantly positive nor significantly negative on the outcome of interest. Inconsistent indicates that for a narrative synthesis the evidence of a particular intervention on a specific outcome shows a mix of positive and no effects across the underlying studies. Negative indicates that the intervention is found to have a negative effect on the outcome of interest. Given the direction of the evidence, the dark- and medium-green legends indicate that the evidence of an intervention on a particular outcome is found to be positive in more than three systematic reviews or in up to three systematic reviews, respectively. Similarly, the dark-red legend indicates that the evidence of an intervention on a particular outcome is found to be negative in more than two systematic reviews. The light-green legend indicates that the pool of evidence of a particular intervention on a specific outcome shows a mix of positive effects, no effect, or a combination of both (inconsistent) in narrative synthesis. IS = institutional strengthening; WASH = water, sanitation, and hygiene. However, the World Bank could increase its attention to particular nutrition-specific and nutrition-sensitive interventions where evidence is consistently positive across a broad set of nutrition-relevant outcomes areas. Despite the World Bank supports many of the interventions identified as having a broad positive impact, some of them may not be receiving sufficient attention given their potential benefits (table B.1). Among nutrition-specific interventions, few projects in the HNP portfolio include women’s supplementary feeding with energy-dense food and children supplementary feeding with micronutrient-rich food. Within nutrition-sensitive interventions, vitamin A biofortification of foods in the agriculture portfolio, and provision of soap to stimulate hygiene and sanitation practices in the WASH portfolio have received little attention. 135 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Furthermore, two interventions with broad positive impacts remain unexplored in the nutrition portfolio. The first refers to maternal emotional support interventions for which the global evidence suggest they are effective in improving breastfeeding and parenting practices, women’s mental health, and use of health care services. The second intervention is land property right reforms that could be implemented through governance, macroeconomics, or the agriculture sector. Such reforms can be effective in improving household welfare (consumption and income), empowering women (increased control of resources), reducing micronutrient deficiencies of women, and even stunted growth. 136 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Table B.1. Systematic Review Map Interventions with a Broad Positive Impact Intervention Types Interventions (%) Projects (number) Nutrition-specific Child supplementary feeding with micronutrient-rich foods 0.5 8 Maternal supplementary feeding with energy-dense foods 0.3 4 Women micronutrient supplementation: iron folate (iron– 0.9 15 folic acid) SBCC of nutrition and health promotion (via community and 21.5 107 groups, education, growth monitoring and promotion, home visits, mass communication, and interpersonal communication at health facility) Nutrition-sensitive Health Health system strengthening 8.8 101 Maternal emotional support 0.0 0 Family planning and contraception 2.1 32 Health care approach: community health workers 0.8 11 Health facility community outreach 0.7 11 E-health communication 0.0 0 Health insurance 0.7 11 Agriculture Provision of agriculture inputs and training 2.3 34 Small-scale livestock 2.1 30 Vitamin A fortification 0.5 7 Land property rights 0.0 0 Social protection Conditional cash transfers 2.2 32 Water, santitation, and hygiene Provision of soap 0.2 4 Total interventions with a broad positive impact 43.6 Source: Independent Evaluation Group; systematic review map and portfolio review. Note: SBCC = social and behavior change communication. Limitations of the Analysis on the Alignment with the Evidence Base • The alignment analysis focuses on the intersection of the SRM and the nutrition portfolio, covering about half of the interventions in the portfolio. Project documents sometimes lack detailed information on interventions, making it difficult to map them against the SRM. 137 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio • The SRM identifies interventions targeted to women or children or households. In a few cases, however, the same intervention can be targeted to both women and children, having different effects on relevant outcomes. These are the cases of supplementation with certain micronutrients and deworming. Since it is not possible to distinguish the target groups of particular interventions in the portfolio, the alignment analysis assumes that such interventions targeted both women and children. As a result, the analysis may overestimate the World Bank’s support to these interventions. • The alignment analysis uses data from the portfolio review to identify the intended outcomes of projects, proxied by the projects’ indicators. Given the portfolio review’s finding that the World Bank falls short of measuring the outcomes of certain interventions, the use of project indicators as proxies for outcomes likely introduces some bias. • Many intervention-outcome pairs found in the portfolio have no available evidence in the SRM. This is either because the amount of impact evaluations for such intervention is not sufficient to be summarized in an SR (that is, there is a knowledge gap and the intervention has not been studied enough), or because there is not a theoretical causal pathway linking the intervention and the outcome (such as the effects of micronutrient supplementation in children on the mother’s health). • The SRM represents the stock of knowledge during a certain period and does not incorporate the time dimension. The alignment analysis, therefore, cannot make inferences about the evolution of the alignment between the portfolio interventions and the literature. Systematic Reviews Aboud, F. E., and A. K. Yousafzai. 2015. “Global Health and Development in Early Childhood.” Annual Review of Psychology 66 (1): 433–57. doi:10.1146/annurev-psych-010814-015128. Acharya, A., S. Vellakkal, F. Taylor, E. Masset, A. Satija, M. Burke, and S. Ebrahim. 2012. Impact of National Health Insurance for the Poor and the Informal Sector in Low- and Middle-Income Countries: A Systematic Review. London: The EPPI-Centre, Social Science Research Unit, Institute of Education, University of London. Aggarwal, R., J. Sentz, and M. A. Miller. 2007. “Role of Zinc Administration in Prevention of Childhood Diarrhea and Respiratory Illnesses: A Meta-analysis.” Pediatrics 119 (6): 1120– 30. doi:10.1542/peds.2006-3481. 138 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Angermayr, L., and C. Clar. 2004. “Iodine Supplementation for Preventing Iodine Deficiency Disorders in Children.” Cochrane Database of Systematic Reviews 2004 (2): CD003819. doi:10.1002/14651858.CD003819.pub2. Arnold, B. F., and J. M. Colford Jr. 2007. “Treating Water with Chlorine at Point-of-Use to Improve Water Quality and Reduce Child Diarrhea in Developing Countries: A Systematic Review and Meta-analysis.” American Journal of Tropical Medicine and Hygiene 76 (2): 354–64. doi:10.4269/ajtmh.2007.76.354. Arrowsmith, M. E., C. R. H. Aicken, A. Majeed, and S. Saxena. 2012. “Interventions for Increasing Uptake of Copper Intrauterine Devices: Systematic Review and Meta-analysis.” Contraception 86 (6): 600–5. doi:10.1016/j.contraception.2012.05.015. Ashworth, A., R. Shrimpton, and K. Jamil. 2008. “Growth Monitoring and Promotion: Review of Evidence of Impact. Maternal and Child Nutrition 4 (Suppl 1): 86–117. doi:10.1111/j.1740- 8709.2007.00125.x. Athe, R., M. V. V. Rao, and K. M. Nair. 2014. “Impact of Iron-Fortified Foods on Hb Concentration in Children (<10 Years): A Systematic Review and Meta-analysis of Randomized Controlled Trials.” Public Health Nutrition 17 (3): 579–86. doi:10.1017/S1368980013000062. Baird, S., F. H. G. Ferreira, B. Özler, and M. Woolcock. 2013. “Relative Effectiveness of Conditional and Unconditional Cash Transfers for Schooling Outcomes in Developing Countries: A Systematic Review.” Campbell Systematic Reviews 9 (1): 1–124. doi:10.4073/csr.2013.8. Bassani, D. G., P. Arora, K. Wazny, M. F. Gaffey, L. Lenters, and Z. A. Bhutta. 2013. “Financial Incentives and Coverage of Child Health Interventions: A Systematic Review and Meta- analysis. BMC Public Health 13 (Suppl 3): S30. doi:10.1186/1471-2458-13-S3-S30. Bentley, M. E., H. M. Wasser, and H. M. Creed‐Kanashiro. 2011. “Responsive Feeding and Child Undernutrition in Low- and Middle-Income Countries.” Journal of Nutrition 141 (3): 502– 7. doi:10.3945/jn.110.130005. Berti, P. R., J. Krasevec, and S. FitzGerald. 2004. “A Review of the Effectiveness of Agriculture Interventions in Improving Nutrition Outcomes.” Public Health Nutrition 7 (5): 599–609. doi:10.1079/PHN2003595. Bhutta, Z. A., T. Ahmed, R. E. Black, S. Cousens, K. Dewey, E. Giugliani, B. A. Haider, B. Kirkwood, S. S. Morris, H. P. S. Sachdev, and M. Shekar. 2008. “What Works? Interventions for Maternal and Child Undernutrition and Survival.” Lancet 371 (9610): 417–40. doi:10.1016/S0140-6736(07)61693-6. Bhutta, Z. A., G. L. Darmstadt, B. S. Hasan, and R. A. Haws. 2005. “Community-Based Interventions for Improving Perinatal and Neonatal Health Outcomes in Developing 139 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Countries: A Review of the Evidence.” Pediatrics 115 (Suppl 2): 519–617. doi:10.1542/peds.2004-1441. Bird, F. A., A. Pradhan, R. V. Bhavani, and A. D. Dangour. 2019. “Interventions in Agriculture for Nutrition Outcomes: A Systematic Review Focused on South Asia.” Food Policy 82: 39–49. doi:10.1016/j.foodpol.2018.10.015. Blacklock, C., E. MacPepple, S. Kunutsor, and S. Witter. 2016. “Paying for Performance to Improve the Delivery and Uptake of Family Planning in Low and Middle Income Countries: A Systematic Review.” Studies in Family Planning 47 (4): 309–24. doi:10.1111/sifp.12001. Bourey, C., W. Williams, E. E. Bernstein, and R. Stephenson. 2015. “Systematic Review of Structural Interventions for Intimate Partner Violence in Low- and Middle-Income Countries: Organizing Evidence for Prevention.” BMC Public Health 15: 1165. doi:10.1186/s12889-015-2460-4. Brody, C., T. de Hoop, M. Vojtkova, R. Warnock, M. Dunbar, P. Murthy, and S. L. Dworkin. 2015. “Economic Self‐Help Group Programs for Improving Women’s Empowerment: A Systematic Review.” Campbell Systematic Reviews 11 (1): 1–182. doi:10.4073/csr.2015.19. Brown, K. H., J. M. Peerson, S. K. Baker, and S. Y. Hess. 2009. “Preventive Zinc Supplementation among Infants, Preschoolers, and Older Prepubertal Children.” Food and Nutrition Bulletin 30 (Suppl 1): S12–40. doi:10.1177/15648265090301S103. Brown, T. W., F. C. van Urk, R. Waller, and E. Mayo-Wilson. 2014. “Centre-Based Day Care for Children Younger than Five Years of Age in Low- and Middle-Income Countries.” Cochrane Database of Systematic Reviews 2014 (9): CD010543. doi:10.1002/14651858.CD010543.pub2. Buppasiri, P., P. Lumbiganon, J. Thinkhamrop, C. Ngamjarus, M. Laopaiboon, and N. Medley. 2015. “Calcium Supplementation (Other Than for Preventing or Treating Hypertension) for Improving Pregnancy and Infant Outcomes.” Cochrane Database of Systematic Reviews 2015 (2): CD007079. doi:10.1002/14651858.CD007079.pub3. Byrne, A., A. Hodge, E. Jimenez-Soto, and A. Morgan. 2014. “What Works? Strategies to Increase Reproductive, Maternal and Child Health in Difficult to Access Mountainous Locations: A Systematic Literature Review.” PLoS ONE 9 (2): e87683. doi:10.1371/journal.pone.0087683. Caminha, M. de F. C., M. Batista Filho, T. F. dos S. Fernandes, I. K. G. de Arruda, and A. da S. Diniz. 2009. “Vitamin A Supplementation during Puerperium: Systematic Review.” Revista de Saúde Pública 43 (4): 699–706. doi:10.1590/S0034-89102009005000038. Chen, H., Q. Zhuo, W. Yuan, J. Wang, and T. Wu. 2008. “Vitamin A for Preventing Acute Lower Respiratory Tract Infections in Children up to Seven Years of Age.” Cochrane Database of Systematic Reviews 2008 (1): CD006090. doi:10.1002/14651858.CD006090.pub2. 140 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Choi, L., S. Majambere, and A. L. Wilson. 2019. “Larviciding to Prevent Malaria Transmission.” Cochrane Database of Systematic Reviews 2019 (8): CD012736. doi:10.1002/14651858.CD012736.pub2. Clasen, T. F., K. T. Alexander, D. Sinclair, S. Boisson, R. Peletz, H. H. Chang, F. Majorin, and S. Cairncross. 2015. “Interventions to Improve Water Quality for Preventing Diarrhoea.” Cochrane Database of Systematic Reviews 2015 (10): CD004794. doi:10.1002/14651858.CD004794.pub3. Colaci, D., S. Chaudhri, and A. Vasan. 2017. “mHealth Interventions in Low-Income Countries to Address Maternal Health: A Systematic Review.” Annals of Global Health 82 (5): 922–35. doi:10.1016/j.aogh.2016.09.001. Curtis, V., and S. Cairncross. 2003. “Effect of Washing Hands with Soap on Diarrhoea Risk in the Community: A Systematic Review.” Lancet Infectious Diseases 3 (5): 275–81. doi:10.1016/S1473-3099(03)00606-6. Dangour, A. D., L. Watson, O. Cumming, S. Boisson, Y. Che, Y. Velleman, S. Cavill, E. Allen, and R. Uauy. 2013. “Interventions to Improve Water Quality and Supply, Sanitation and Hygiene Practices, and Their Effects on the Nutritional Status of Children.” Cochrane Database of Systematic Reviews 2013 (8): CD009382. doi:10.1002/14651858.CD009382.pub2. Darvesh, N., J. K. Das, T. Vaivada, M. F. Gaffey, K. Rasanathan, and Z. A. Bhutta. 2017. “Water, Sanitation and Hygiene Interventions for Acute Childhood Diarrhea: A Systematic Review to Provide Estimates for the Lives Saved Tool.” BMC Public Health 17 (Suppl 4): 776. doi:10.1186/s12889-017-4746-1. Das, J. K., Y. B. Hadi, R. A. Salam, M. Hoda, Z. S. Lassi, and Z. A. Bhutta. 2018. “Fly Control to Prevent Diarrhoea in Children.” Cochrane Database of Systematic Reviews 2018 (12): CD011654. doi:10.1002/14651858.CD011654.pub2. Das, J. K., Z. S. Lassi, R. A. Salam, and Z. A. Bhutta. 2013. “Effect of Community Based Interventions on Childhood Diarrhea and Pneumonia: Uptake of Treatment Modalities and Impact on Mortality.” BMC Public Health 13 (Suppl 3): S29. doi:10.1186/1471-2458-13- S3-S29. Das, J. K., R. A. Salam, Y. B. Hadi, S. S. Sheikh, A. Z. Bhutta, Z. W. Prinzo, and Z. A. Bhutta. 2019. “Preventive Lipid-Based Nutrient Supplements Given with Complementary Foods to Infants and Young Children 6 to 23 Months of Age for Health, Nutrition, and Developmental Outcomes.” Cochrane Database of Systematic Reviews 2019 (5): CD012611. doi:10.1002/14651858.CD012611.pub3. Dawson, A. J., J. Buchan, C. Duffield, C. S. E. Homer, and K. Wijewardena. 2014. “Task Shifting and Sharing in Maternal and Reproductive Health in Low-Income Countries: A Narrative Synthesis of Current Evidence.” Health Policy and Planning 29 (3): 396–408. doi:10.1093/heapol/czt026. 141 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio De Buck, E., H. Van Remoortel, K. Hannes, T. Govender, S. Naidoo, B. Avau, A. V. Veegaete, A. Musekiwa, V. Lutje, M. Cargo, H. -J. Mosler, P. Vandekerckhove, and T. Young. 2017. “Approaches to Promote Handwashing and Sanitation Behaviour Change in Low‐ and Middle‐Income Countries: A Mixed Method Systematic Review.” Campbell Systematic Reviews 13: 1–447. doi:10.4073/csr.2017.7. De Buck, E., H. Van Remoortel, A. Vande Veegaete, and T. Young. 2017. “Promoting Handwashing and Sanitation Behaviour Change in Low- and Middle-Income Countries, 3ie Systematic Review Summary 10.” London: International Initiative for Impact Evaluation (3ie). De-Regil, L. M., M. E. D. Jefferds, and J. P. Peña-Rosas. 2017. “Point-of-Use Fortification of Foods with Micronutrient Powders Containing Iron in Children of Preschool and School-Age.” Cochrane Database of Systematic Reviews 2017 (11): CD009666. doi:10.1002/14651858.CD009666.pub2. De-Regil, L. M., M. E. D. Jefferds, A. C. Sylvetsky, and T. Dowswell. 2011. “Intermittent Iron Supplementation for Improving Nutrition and Development in Children under 12 Years of Age.” Cochrane Database of Systematic Reviews 2011 (12): CD009085. doi:10.1002/14651858.CD009085.pub2. De-Regil, L. M., P. S. Suchdev, G. E. Vist, S. Walleser, and J. P. Peña-Rosas. 2013. “Home Fortification of Foods with Multiple Micronutrient Powders for Health and Nutrition in Children under Two Years of Age.” Evidence-Based Child Health 8 (1): 112–201. doi:10.1002/ebch.1895. Dewey, K. G., and S. Adu‐Afarwuah. 2008. “Systematic Review of the Efficacy and Effectiveness of Complementary Feeding Interventions in Developing Countries.” Maternal and Child Nutrition 4 (Suppl 1): 24–85. doi:10.1111/j.1740-8709.2007.00124.x. Dewey, K. G., Z. Yang, and E. Boy. 2009. “Systematic Review and Meta-analysis of Home Fortification of Complementary Foods.” Maternal and Child Nutrition 5 (4): 283–321. doi:10.1111/j.1740-8709.2009.00190.x. Dickson, K., and M. Bangpan. 2012. Providing Access to Economic Assets for Girls and Young Women in Low- and Lower-Middle Income Countries. A Systematic Review of the Evidence. London: EPPI-Centre, Social Science Research Unit, Institute of Education, University of London. Dudley, L., and P. Garner. 2011. “Strategies for Integrating Primary Health Services in Low- and Middle-Income Countries at the Point of Delivery.” Cochrane Database of Systematic Reviews 2011 (7): CD003318. doi:10.1002/14651858.CD003318.pub3. Eaton, J. C., P. Rothpletz-Puglia, M. R. Dreker, L. Iannotti, C. Lutter, J. Kaganda, and P. Rayco- Solon. 2019. “Effectiveness of Provision of Animal-Source Foods for Supporting Optimal Growth and Development in Children 6 to 59 Months of Age.” Cochrane Database of Systematic Reviews 2019 (2): CD012818. doi:10.1002/14651858.CD012818.pub2. 142 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Eichler, R., K. Agarwal, I. Askew, E. Iriarte, L. Morgan, and J. Watson. 2013. “Performance-Based Incentives to Improve Health Status of Mothers and Newborns: What Does the Evidence Show?” Journal of Health, Population and Nutrition 31 (4 Suppl 2): S36–47. Elder, J. P., W. Pequegnat, S. Ahmed, G. Bachman, M. Bullock, W. A. Carlo, V. Chandra-Mouli, N. A. Fox, S. Harkness, G. Huebner, J. Lombardi, V. M. Murry, A. Moran, M. Norton, J. Mulik, W. Parks, H. H. Raikes, J. Smyser, C. Sugg, and M. Sweat. 2014. “Caregiver Behavior Change for Child Survival and Development in Low- and Middle-Income Countries: An Examination of the Evidence.” Journal of Health Communication 19 (Suppl 1): 25–66. doi:10.1080/10810730.2014.940477. Fernández-Gaxiola, A. C., and L. M. De-Regil. 2019. “Intermittent Iron Supplementation for Reducing Anaemia and Its Associated Impairments in Adolescent and Adult Menstruating Women.” Cochrane Database of Systematic Reviews 2019 (1): CD009218. doi:10.1002/14651858.CD009218.pub3. Fewtrell, L., and J. M. Colford Jr. 2004. “Water, Sanitation and Hygiene: Interventions and Diarrhoea. A Systematic Review and Meta-analysis.” Health, Nutrition, and Population (HNP) Discussion Paper. Washington, DC: World Bank. Fewtrell, L., R. B. Kaufmann, D. Kay, W. Enanoria, L. Haller, and J. M. Colford. 2005. “Water, Sanitation, and Hygiene Interventions to Reduce Diarrhoea in Less Developed Countries: A Systematic Review and Meta-analysis.” Lancet Infectious Diseases 5 (1): 42–52. doi:10.1016/S1473-3099(04)01253-8. Finkelstein, J. L., J. D. Haas, and S. Mehta. 2017. “Iron-Biofortified Staple Food Crops for Improving Iron Status: A Review of the Current Evidence.” Current Opinion in Biotechnology 44: 138–45. doi:10.1016/j.copbio.2017.01.003. Freeman, M. C., J. V. Garn, G. D. Sclar, S. Boisson, K. Medlicott, K. T. Alexander, G. Penakalapati, D. Anderson, A. G. Mahtani, J. E. T. Grimes, E. A. Rehfuess, and T. F. Clasen. 2017. “The Impact of Sanitation on Infectious Disease and Nutritional Status: A Systematic Review and Meta-analysis.” International Journal of Hygiene and Environmental Health 220 (6): 928– 49. doi:10.1016/j.ijheh.2017.05.007. Gajaria, A., and A. V. Ravindran. 2018. “Interventions for Perinatal Depression in Low and Middle-Income Countries: A Systematic Review.” Asian Journal of Psychiatry 37: 112–20. doi:10.1016/j.ajp.2018.08.014. Gamble, C. L., J. P. Ekwaru, and F. O. ter Kuile. 2006. “Insecticide-Treated Nets for Preventing Malaria in Pregnancy.” Cochrane Database of Systematic Reviews 2006 (2): CD003755. doi:10.1002/14651858.CD003755.pub2. Garcia-Casal, M. N., J. P. Peña-Rosas, L. M. De-Regil, J. A. Gwirtz, and S. R. Pasricha. 2018. “Fortification of Maize Flour with Iron for Controlling Anaemia and Iron Deficiency in 143 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Populations.” Cochrane Database of Systematic Reviews 2018 (12): CD010187. doi:10.1002/14651858.CD010187.pub2. Garn, J. V., G. D. Sclar, M. C. Freeman, G. Penakalapati, K. T. Alexander, P. Brooks, E. A. Rehfuess, S. Boisson, K. O. Medlicott, and T. F. Clasen. 2017. “The Impact of Sanitation Interventions on Latrine Coverage and Latrine Use: A Systematic Review and Meta- analysis.” International Journal of Hygiene and Environmental Health 220 (2): 329–40. doi:10.1016/j.ijheh.2016.10.001. Garner, P., R. Panpanich, and S. Logan. 2000. “Is Routine Growth Monitoring Effective? A Systematic Review of Trials.” Archives of Disease in Childhood 82 (3): 197–201. doi:10.1136/adc.82.3.197. Gebreselassie, S. G., and F. E. Gashe. 2011. “A Systematic Review of Effect of Prenatal Zinc Supplementation on Birthweight: Meta-analysis of 17 Randomized Controlled Trials.” Journal of Health, Population and Nutrition 29 (2): 134–40. Geerligs, P. D. P., B. J. Brabin, and A. A. A. Omari. 2003. “Food Prepared in Iron Cooking Pots as an Intervention for Reducing Iron Deficiency Anaemia in Developing Countries: A Systematic Review.” Journal of Human Nutrition and Dietetics 16 (4): 275–81. doi:10.1046/j.1365-277X.2003.00447.x. Gera, T., and H. P. S. Sachdev. 2002. “Effect of Iron Supplementation on Incidence of Infectious Illness in Children: Systematic Review. BMJ 325 (7373): 1142. doi:10.1136/bmj.325.7373.1142. Gera, T., H. P. S. Sachdev, and P. Nestel. 2009. “Effect of Combining Multiple Micronutrients with Iron Supplementation on Hb Response in Children: Systematic Review of Randomized Controlled Trials.” Public Health Nutrition 12 (6): 756–73. doi:10.1017/S1368980008003145. Gera, T., H. P. S. Sachdev, P. Nestel, and S. S. Sachdev. 2007. “Effect of Iron Supplementation on Haemoglobin Response in Children: Systematic Review of Randomised Controlled Trials. Journal of Pediatric Gastroenterology and Nutrition 44 (4): 468–86. doi:10.1097/01.mpg.0000243440.85452.38. Gera, T., D. Shah, P. Garner, M. Richardson, and H. S. Sachdev. 2016. “Integrated Management of Childhood Illness (IMCI) Strategy for Children under Five. Cochrane Database of Systematic Reviews 2016 (6): CD010123. doi:10.1002/14651858.CD010123.pub2. Gera, T., D. Shah, and H. S. Sachdev. 2018. “Impact of Water, Sanitation and Hygiene Interventions on Growth, Non-diarrheal Morbidity and Mortality in Children Residing in Low- and Middle-Income Countries: A Systematic Review.” Indian Pediatrics 55 (5): 381– 93. doi:10.1007/s13312-018-1279-3. Ghogomu, E. T., S. Suresh, P. Rayco-Solon, A. Hossain, J. McGowan, J. P. Peña-Rosas, and V. Welch. 2018. “Deworming in Non-pregnant Adolescent Girls and Adult Women: A 144 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Systematic Review and Meta-analysis.” Systematic Reviews 7: 239. doi:10.1186/s13643-018- 0859-6. Gilmore, B., and E. McAuliffe. 2013. “Effectiveness of Community Health Workers Delivering Preventive Interventions for Maternal and Child Health in Low- and Middle-Income Countries: A Systematic Review.” BMC Public Health 13: 847. doi:10.1186/1471-2458-13- 847. Girard, A. W., J. L. Self, C. McAuliffe, and O. Olude. 2012. “The Effects of Household Food Production Strategies on the Health and Nutrition Outcomes of Women and Young Children: A Systematic Review.” Paediatric and Perinatal Epidemiology 26 (Suppl 1): 205– 22. doi:10.1111/j.1365-3016.2012.01282.x. Glassman, A., D. Duran, L. Fleisher, D. Singer, R. Sturke, G. Angeles, J. Charles, B. Emrey, J. Gleason, W. Mwebsa, K. Saldana, K. Yarrow, and M. Koblinsky. 2013. “Impact of Conditional Cash Transfers on Maternal and Newborn Health.” Journal of Health, Population and Nutrition 31 (4 Suppl 2): S48–66. Glasziou, P. P., and D. E. Mackerras. 1993. “Vitamin A Supplementation in Infectious Diseases: A Meta-analysis.” British Medical Journal 306 (6874): 366–70. doi:10.1136/bmj.306.6874.366. Gogia, S., and H. S. Sachdev. 2009. “Neonatal Vitamin A Supplementation for Prevention of Mortality and Morbidity in Infancy: Systematic Review of Randomised Controlled Trials.” BMJ 338 (7699): b919. doi:10.1136/bmj.b919. Gogia, S., and H. S. Sachdev. 2010. “Maternal Postpartum Vitamin A Supplementation for the Prevention of Mortality and Morbidity in Infancy: A Systematic Review of Randomized Controlled Trials.” International Journal of Epidemiology 39 (5): 1217–26. doi:10.1093/ije/dyq080. Gogia, S., and H. S. Sachdev. 2012. “Zinc Supplementation for Mental and Motor Development in Children.” Cochrane Database of Systematic Reviews 2012 (12): CD007991. doi:10.1002/14651858.CD007991.pub2. Gopalaswamy, A. K., M. S. Babu, and U. Dash. 2016. Systematic Review of Quantitative Evidence on the Impact of Microfinance on the Poor in South Asia. London: EPPI-Centre, Social Science Research Unit, UCL Institute of Education, University College London. Goudet, S. M., P. L. Griffiths, B. A. Bogin, and N. J. Madise. 2015. “Nutritional Interventions for Preventing Stunted Growth in Children (0 to 5 Years) Living in Urban Slums. Cochrane Database of Systematic Reviews 2015 (5): CD011695. doi:10.1002/14651858.CD011695. Grantham-McGregor, S. M., L. C. H. Fernald, R. M. C. Kagawa, and S. Walker. 2014. “Effects of Integrated Child Development and Nutrition Interventions on Child Development and Nutritional Status.” Annals of the New York Academy of Sciences 1308 (1): 11–32. doi:10.1111/nyas.12284. 145 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Gunaratna, N. S., H. De Groote, P. Nestel, K. V. Pixley, and G. P. McCabe. 2010. “A Meta-analysis of Community-Based Studies on Quality Protein Maize.” Food Policy 35 (3): 202–10. doi:10.1016/j.foodpol.2009.11.003. Hagen-Zanker, J., A. McCord, R. Holmes, F. Booker, and E. Molinari, E. 2011. Systematic Review of the Impact of Employment Guarantee Schemes and Cash Transfers on the Poor. London: Overseas Development Institute. Haider, B. A., R. Sharma, and Z. A. Bhutta. 2017. “Neonatal Vitamin A Supplementation for the Prevention of Mortality and Morbidity in Term Neonates in Low and Middle Income Countries.” Cochrane Database of Systematic Reviews 2017 (2): CD006980. doi:10.1002/14651858.CD006980.pub3. Hall, J. 2011. “Effective Community-Based Interventions to Improve Exclusive Breast Feeding at Four to Six Months in Low- and Low-Middle-Income Countries: A Systematic Review of Randomised Controlled Trials.” Midwifery 27 (4): 497–502. doi:10.1016/j.midw.2010.03.011. Hatt, L., B. Johns, C. Connor, M. Meline, M. Kukla, and K. Moat. 2015. Impact of Health Systems Strengthening on Health. Bethesda: Health Finance and Governance Project, Abt Associates. Hermoso, M., V. Vucic, C. Vollhardt, A. Arsic, B. Roman-Viñas, I. Iglesia-Altaba,M. Gurinovic, and B. Koletzko. 2011. “The Effect of Iron on Cognitive Development and Function in Infants, Children and Adolescents: A Systematic Review.” Annals of Nutrition and Metabolism 59 (2–4): 154–65. doi:10.1159/000334490. Hess, S., L. Tecklenburg, and K. Eichler. 2016. “Micronutrient Fortified Condiments and Noodles to Reduce Anemia in Children and Adults—A Literature Review and Meta-analysis.” Nutrients 8 (2): 88. doi:10.3390/nu8020088. Hindin, M. J., A. M. Kalamar, T. -A. Thompson, and U. D. Upadhyay. 2016. “Interventions to Prevent Unintended and Repeat Pregnancy among Young People in Low- and Middle- Income Countries: A Systematic Review of the Published and Gray Literature.” Journal of Adolescent Health 59 (3 Suppl): S8–15. doi:10.1016/j.jadohealth.2016.04.021. Hombali, A. S., J. A. Solon, B. T. Venkatesh, N. S. Nair, and J. P. Peña-Rosas. 2019. “Fortification of Staple Foods with Vitamin A for Vitamin A Deficiency.” Cochrane Database of Systematic Reviews 2019 (5): CD010068. doi:10.1002/14651858.CD010068.pub2. Imdad, A., Z. Ahmed, and Z. A. Bhutta. 2016. “Vitamin A Supplementation for the Prevention of Morbidity and Mortality in Infants One to Six Months of Age.” Cochrane Database of Systematic Reviews 2016 (9): CD007480. doi:10.1002/14651858.CD007480.pub3. Imdad, A., and Z. A. Bhutta. 2011a. “Effect of Balanced Protein Energy Supplementation during Pregnancy on Birth Outcomes.” BMC Public Health 11 (Suppl 3): S17. doi:10.1186/1471- 2458-11-S3-S17. 146 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Imdad, A., and Z. A. Bhutta. 2011b. “Effect of Preventive Zinc Supplementation on Linear Growth in Children under 5 Years of Age in Developing Countries: A Meta-analysis of Studies for Input to the Lives Saved Tool.” BMC Public Health 11 (Suppl 3): S22. doi:10.1186/1471-2458-11-S3-S22. Imdad, A., and Z. A. Bhutta. 2012a. “Maternal Nutrition and Birth Outcomes: Effect of Balanced Protein-Energy Supplementation.” Paediatric and Perinatal Epidemiology 26 (Suppl 1): 178– 90. doi:10.1111/j.1365-3016.2012.01308.x. Imdad, A., and Z. A. Bhutta. 2012b. “Routine Iron/Folate Supplementation during Pregnancy: Effect on Maternal Anaemia and Birth Outcomes.” Paediatric and Perinatal Epidemiology 26 (Suppl 1), 168–77. doi:10.1111/j.1365-3016.2012.01312.x. Imdad, A., E. Mayo-Wilson, K. Herzer, and Z. A. Bhutta. 2017. “Vitamin A Supplementation for Preventing Morbidity and Mortality in Children from Six Months to Five Years of Age.” Cochrane Database of Systematic Reviews 2017 (3): CD008524. doi:10.1002/14651858.CD008524.pub3. Imdad, A., M. Y. Yakoob, and Z. A. Bhutta. 2011. “Impact of Maternal Education about Complementary Feeding and Provision of Complementary Foods on Child Growth in Developing Countries. BMC Public Health 11 (Suppl 3): S25. doi:10.1186/1471-2458-11-S3- S25. Imhoff-Kunsch, B., and V. Briggs. 2012. “Antihelminthics in Pregnancy and Maternal, Newborn and Child Health.” Paediatric and Perinatal Epidemiology 26 (Suppl 1): 223–38. doi:10.1111/j.1365-3016.2012.01280.x. Jenkins, M., C. Byker Shanks, and B. Houghtaling. 2015. “Orange-Fleshed Sweet Potato: Successes and Remaining Challenges of the Introduction of a Nutritionally Superior Staple Crop in Mozambique.” Food and Nutrition Bulletin 36 (3): 327–53. doi:10.1177/0379572115597397. Jeong, J., H. O. Pitchik, and A. K. Yousafzai. 2018. “Stimulation Interventions and Parenting in Low- and Middle-Income Countries: A Meta-analysis.” Pediatrics 141 (4): e20173510. doi:10.1542/peds.2017-3510. Jones-Hughes, T., J. Peters, R. Whear, C. Cooper, H. Evans, M. Depledge, and M. Pearson. 2013. “Are Interventions to Reduce the Impact of Arsenic Contamination of Groundwater on Human Health in Developing Countries Effective? A Systematic Review.” Environmental Evidence 2: 11. doi:10.1186/2047-2382-2-11. Kabeer, N., and H. Waddington. 2015. “Economic Impacts of Conditional Cash Transfer Programmes: A Systematic Review and Meta-analysis.” Journal of Development Effectiveness 7 (3): 290–303. doi:10.1080/19439342.2015.1068833. Kågesten, A., J. Parekh, Ö. Tunçalp, S. Turke, and R. W. Blum. 2014. “Comprehensive Adolescent Health Programs that Include Sexual and Reproductive Health Services: A Systematic Review.” American Journal of Public Health 104 (12): e23–36. doi:10.2105/AJPH.2014.302246. 147 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Kalamar, A. M., S. Lee-Rife, and M. J. Hindin. 2016. “Interventions to Prevent Child Marriage among Young People in Low- and Middle-Income Countries: A Systematic Review of the Published and Gray Literature.” Journal of Adolescent Health 59 (3 Suppl): S16–21. doi:10.1016/j.jadohealth.2016.06.015. Karra, M., D. Canning, J. Hu, M. Ali, and C. Lissner. 2016. “Community-Based Financing of Family Planning in Developing Countries: A Systematic Review.” Studies in Family Planning 47 (4): 325–39. doi:10.1111/sifp.12000. Khan, M. E., A. Hazra, A. Kant, and M. Ali. 2016. “Conditional and Unconditional Cash Transfers to Improve Use of Contraception in Low and Middle Income Countries: A Systematic Review.” Studies in Family Planning 47 (4): 371–83. doi:10.1111/sifp.12004. Kim, S. K., S. Park, J. Oh, J. Kim, and S. Ahn. 2018. “Interventions Promoting Exclusive Breastfeeding up to Six Months after Birth: A Systematic Review and Meta-analysis of Randomized Controlled Trials.” International Journal of Nursing Studies 80: 94–105. doi:10.1016/j.ijnurstu.2018.01.004. Korachais, C., E. Macouillard, and B. Meessen. 2016. “How User Fees Influence Contraception in Low and Middle Income Countries: A Systematic Review.” Studies in Family Planning 47 (4): 341–56. doi:10.1111/sifp.12005. Kristjansson, E., D. K. Francis, S. Liberato, M. Benkhalti Jandu, V. Welch, M. Batal, T. Greenhalgh, T. Rader, E. Noonan, B. Shea, L. Janzen, G. A. Wells, and M. Petticrew, M. 2015. “Food Supplementation for Improving the Physical and Psychosocial Health of Socio- Economically Disadvantaged Children Aged Three Months to Five Years.” Cochrane Database of Systematic Reviews 2015 (3): CD009924. https://doi.org/:10.1002/14651858.CD009924.pub2. Kristjansson, E., D. Francis, S. Liberato, T. Greenhalgh, V. Welch, M. Benkhalti Jandu, M. Batal, T. Rader, E. Noonan, L. Janzen, B. Shea, G. A. Wells, and M. Petticrew. 2015. “Supplementary Feeding for Improving the Health of Disadvantaged Infants and Young Children: A Systematic and Realist Review.” 3ie Systematic Review 15. London: International Initiative for Impact Evaluation (3ie). Lagarde, M., and N. Palmer. 2011. “The Impact of User Fees on Access to Health Services in Low- and Middle-Income Countries.” Cochrane Database of Systematic Reviews 2011 (4): CD009094. doi:10.1002/14651858.CD009094. Lamstein, S., T. Stillman, P. Koniz-Booher, A. Aakesson, B. Collaiezzi, T. Williams, K. Beall, and M. Anson. 2014. “Evidence of Effective Approaches to Social and Behaviour Change Communication for Preventing and Reducing Stunting and Anemia: Findings from a Systematic Literature Review.” Arlington: USAID/Strengthening Partnerships, Results, and Innovations in Nutrition Globally (SPRING) Project. https://www.spring- nutrition.org/publications/series/evidence-effective-approaches-social-and-behavior- change-communication. 148 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Langer, L., Y. Erasmus, N. Tannous, E. Obuku, Z. Ravat, C. Chisoro, M. Opondo, P. Nduku, J. Tripney, C. van Rooyen, and R. Stewart. 2018. “Women in Wage Labour: A Systematic Review of the Effectiveness and Design Features of Interventions Supporting Women’s Participation in Wage Labour in Higher-Growth and/or Male-Dominated Sectors in LMICs.” Technical report. London: EPPI-Centre, Social Science Research Unit, UCL Institute of Education, University College London. Lassi, Z. S., and Z. A. Bhutta. 2015. “Community-Based Intervention Packages for Reducing Maternal and Neonatal Morbidity and Mortality and Improving Neonatal Outcomes.” Cochrane Database of Systematic Reviews 2015 (3): CD007754. doi:10.1002/14651858.CD007754.pub3. Lassi, Z. S., J. K. Das, G. Zahid, A. Imdad, and Z. A. Bhutta. 2013. “Impact of Education and Provision of Complementary Feeding on Growth and Morbidity in Children less than 2 Years of Age in Developing Countries: A Systematic Review.” BMC Public Health 13 (Suppl 3): S13. doi:10.1186/1471-2458-13-S3-S13. Lassi, Z. S., P. F. Middleton, Z. A. Bhutta, and C. Crowther. 2016. “Strategies for Improving Health Care Seeking for Maternal and Newborn Illnesses in Low- and Middle-Income Countries: A Systematic Review and Meta-analysis.” Global Health Action 9 (1): 31408. doi:10.3402/gha.v9.31408. Lassi, Z. S., A. Moin, and Z. A. Bhutta. 2016. “Zinc Supplementation for the Prevention of Pneumonia in Children Aged 2 Months to 59 Months. Cochrane Database of Systematic Reviews 2016 (12): CD005978. doi:10.1002/14651858.CD005978.pub3. Lassi, Z. S., A. Moin, J. K. Das, R. A. Salam, and Z. A. Bhutta. 2017. “Systematic Review on Evidence-Based Adolescent Nutrition Interventions.” Annals of the New York Academy of Sciences 1393 (1): 34–50. doi:10.1111/nyas.13335. Lassi, Z. S., G. Zahid, J. K. Das, and Z. A. Bhutta. 2013. Systematic Review of Complementary Feeding Strategies amongst Children Less Than Two Years of Age. HEART, UK. 76 pp. https://www.heart-resources.org/assignment/systematic-review-of-complementary- feeding-strategies-amongst-children-less-than-two-years-of-age/. Lawry, S., C. Samii, R. Hall, A. Leopold, D. Hornby, and F. Mtero. 2014. “The Impact of Land Property Rights Interventions on Investment and Agricultural Productivity in Developing Countries: A Systematic Review.” Campbell Systematic Reviews 10 (1): 1–104. doi:10.4073/csr.2014.1. Lawson, T. M. 2012. “Impact of School Feeding Programs on Educational, Nutritional, and Agricultural Development Goals: A Systematic Review of Literature.” Master’s Degree Paper, Michigan State University. doi:10.22004/ag.econ.142466. Lazzerini, M., and H. Wanzira. 2016. “Oral Zinc for Treating Diarrhoea in Children.” Cochrane Database of Systematic Reviews 2016 (12): CD005436. doi:10.1002/14651858.CD005436.pub5. 149 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Lee, S. H., U. B. Nurmatov, B. I. Nwaru, M. Mukherjee, L. Grant, and C. Pagliari. 2016. “Effectiveness of mHealth Interventions for Maternal, Newborn and Child Health in Low- and Middle-Income Countries: Systematic Review and Meta-analysis.” Journal of Global Health 6 (1): 010401. doi:10.7189/jogh.06.010401. Leroy, J. L., P. Gadsden, and M. Guijarro. 2012. “The Impact of Daycare Programmes on Child Health, Nutrition and Development in Developing Countries: A Systematic Review.” Journal of Development Effectiveness 4 (3): 472–96. doi:10.1080/19439342.2011.639457. Leroy, J. L., M. Ruel, and E. Verhofstadt. 2009. “The Impact of Conditional Cash Transfer Programmes on Child Nutrition: A Review of Evidence Using a Programme Theory Framework.” Journal of Development Effectiveness 1 (2): 103–29. doi:10.1080/19439340902924043. Leung, B. M. Y., K. P. Wiens, and B. J. Kaplan. 2011. “Does Prenatal Micronutrient Supplementation Improve Children’s Mental Development? A Systematic Review.” BMC Pregnancy and Childbirth 11: 12. doi:10.1186/1471-2393-11-12. Liberato, S. C., G. Singh, and K. Mulholland. 2015. “Zinc Supplementation in Young Children: A Review of the Literature Focusing on Diarrhoea Prevention and Treatment.” Clinical Nutrition 34 (2): 181–88. doi:10.1016/j.clnu.2014.08.002. Liu, E., L. Pimpin, M. Shulkin, S. Kranz, C. P. Duggan, D. Mozaffarian, and W. W. Fawzi. 2018. “Effect of Zinc Supplementation on Growth Outcomes in Children under 5 Years of Age. Nutrients 10 (3): 377. doi:10.3390/nu10030377. Lucas, P. J., C. Cabral, and J. M. Colford Jr. 2011. “Dissemination of Drinking Water Contamination Data to Consumers: A Systematic Review of Impact on Consumer Behaviors.” PLoS ONE 6 (6): e21098. doi:10.1371/journal.pone.0021098. Lynch, U., S. McGrellis, M. Dutschke, M. Anderson, P. Arnsberger, and G. Macdonald. 2017. What Is the Evidence That the Establishment or Use of Community Accountability Mechanisms and Processes Improves Inclusive Service Delivery by Governments, Donors and NGOs to Communities? London: EPPI-Centre, Social Science Research Unit, Institute of Education, University of London. Madhani, F. I., C. Tompkins, S. M. Jack, and A. Fisher. 2015. “Participation in Micro-finance Programmes and Women’s Mental Health in South Asia: A Modified Systematic Review.” Journal of Development Studies 51 (9): 1255–70. doi:10.1080/00220388.2015.1036037. Mahomed, K., and A. M. Gülmezoglu. 1997. “Maternal Iodine Supplements in Areas of Deficiency.” Cochrane Database of Systematic Reviews 1997 (4): CD000135. doi:10.1002/14651858.CD000135. 150 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Manley, J., S. Gitter, and V. Slavchevska. 2013. “How Effective Are Cash Transfer Programmes at Improving Nutritional Status?” World Development 48: 133–55. doi:10.1016/j.worlddev.2013.03.010. Masset, E., L. Haddad, A. Cornelius, and J. Isaza-Castro. 2011. A Systematic Review of Agricultural Interventions That Aim to Improve Nutritional Status of Children. London: EPPI-Centre, Social Science Research Unit, Institute of Education, University of London. Matangila, J. R., P. Mitashi, R. A. Inocêncio da Luz, P. T. Lutumba, and J. P. Van Geertruyden. 2015. “Efficacy and Safety of Intermittent Preventive Treatment for Malaria in Schoolchildren: A Systematic Review.” Malaria Journal 14: 450. doi:10.1186/s12936-015- 0988-5. Mayo-Wilson, E., A. Imdad, K. Herzer, M. Y. Yakoob, and Z. A. Bhutta. 2011. “Vitamin A Supplements for Preventing Mortality, Illness, and Blindness in Children Aged under 5: Systematic Review and Meta-analysis.” BMJ 343: d5094. doi:10.1136/bmj.d5094. Mayo-Wilson, E., J. A. Junior, A. Imdad, S. Dean, X. H. S. Chan, E. S. Chan, A. Jaswal, and Z. A. Bhutta. 2014. “Zinc Supplementation for Preventing Mortality, Morbidity, and Growth Failure in Children Aged 6 Months to 12 Years of Age.” Cochrane Database of Systematic Reviews 2014 (5): CD009384. doi:10.1002/14651858.CD009384.pub2. McCauley, M. E., N. van den Broek, L. Dou, and M. Othman. 2015. “Vitamin A Supplementation during Pregnancy for Maternal and Newborn Outcomes.” Cochrane Database of Systematic Reviews 2015 (10): CD008666. doi:10.1002/14651858.CD008666.pub3. McCorriston, S., D. J. Hemming, J. D. Lamontagne-Godwin, M. J. Parr, J. Osborn, and P. D. Roberts. 2013. What Is the Evidence of the Impact of Agricultural Trade Liberalisation on Food Security in Developing Countries? A Systematic Review. London: EPPI-Centre, Social Science Research Unit, Institute of Education, University of London. McGready, R., N. White, and F. Nosten. 2011. “Parasitological Efficacy of Antimalarials in the Treatment and Prevention of Falciparum malaria in Pregnancy 1998 to 2009: A Systematic Review.” BJOG 118 (2): 123–35. doi:10.1111/j.1471-0528.2010.02810.x. McQueston, K., R. Silverman, and A. Glassman. 2013. “The Efficacy of Interventions to Reduce Adolescent Childbearing in Low- and Middle-Income Countries: A Systematic Review.” Studies in Family Planning 44 (4): 369–88. doi:10.1111/j.1728-4465.2013.00365.x. Menon, K., S. Puthussery, A. Ravalia, P. Panchal, R. Rana, S. Mistry, P. Tseng, J. Bhandol, and D. Mavalankar. 2018. “Effectiveness of Nutrition Interventions in Low and Middle Income Countries: An Evidence Summary.” Indian Institute of Public Health Gandhinagar; Public Health Foundation of India; University of Bedfordshire, UK and BRAC, Bangladesh. 151 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Meyer, C., N. Bellows, M. Campbell, and M. Potts. 2011. The Impact of Vouchers on the Use and Quality of Health Goods and Services in Developing Countries: A Systematic Review. London: EPPI-Centre, Social Science Research Unit, Institute of Education, University of London. Middleton, P. F., Z. S. Lassi, T. S. Tran, Z. Bhutta, T. K. Bubner, V. Flenady, and C. A. Crowther. 2013. “Nutrition Interventions and Programs for Reducing Mortality and Morbidity in Pregnant and Lactating Women and Women of Reproductive Age: A Systematic Review,” in “Abstracts of the 17th Congress of the Perinatal Society of Australia and New Zealand (PSANZ), 14–17 April 2013, Adelaide Convention Centre, Adelaide, South Australia,” special issue, Journal of Paediatrics and Child Health 49 (Suppl 2): 71. Mitra, S., M. Palmer, S. Pullaro, D. Mont, and N. Groce. 2017. “Health Insurance and Children in Low- and Middle-income Countries: A Review.” Economic Record 93 (302): 484–500. doi:10.1111/1475-4932.12331. Morgan, C., A. Petrosino, and T. Fronius. 2012. A Systematic Review of the Evidence of the Impact of Eliminating School User Fees in Low-Income Developing Countries. London: EPPI-Centre, Social Science Research Unit, Institute of Education, University of London. Morgan, C., A. Petrosino, and T. Fronius. 2013. A Systematic Review of the Evidence of the Impact of School Voucher Programmes in Developing Countries. London: EPPI-Centre, Social Science Research Unit, Institute of Education, University of London. Morita, T., S. Godfrey, and C. M. George. 2016. “Systematic Review of Evidence on the Effectiveness of Safe Child Faeces Disposal Interventions.” Tropical Medicine and International Health 21 (11): 1403–19. doi:10.1111/tmi.12773. Muanda, F. T., S. Chaabane, T. Boukhris, F. Santos, O. Sheehy, S. Perreault, L. Blais, and A. Bérard. 2015. “Antimalarial Drugs for Preventing Malaria during Pregnancy and the Risk of Low Birth Weight: A Systematic Review and Meta-analysis of Randomized and Quasi- randomized Trials. BMC Medicine 13: 193. doi:10.1186/s12916-015-0429-x. Murray, S. F., B. M. Hunter, R. Bisht, T. Ensor, and D. Bick. 2012. “Demand-Side Financing Measures to Increase Maternal Health Service Utilisation and Improve Health Outcomes: A Systematic Review of Evidence from Low- and Middle-Income Countries.” JBI Library of Systematic Reviews 10 (58): 4165–67. doi:10.11124/jbisrir-2012-408. Murray, S. F., B. M. Hunter, R. Bisht, T. Ensor, and D. Bick. 2014. “Effects of Demand-Side Financing on Utilisation, Experiences and Outcomes of Maternity Care in Low- and Middle-Income Countries: A Systematic Review.” BMC Pregnancy and Childbirth 14: 30. doi:10.1186/1471-2393-14-30. Mwaikambo, L., I. S. Speizer, A. Schurmann, G. Morgan, and F. Fikree. 2011. “What Works in Family Planning Interventions: A Systematic Review.” Studies in Family Planning 42 (2): 67–82. doi:10.1111/j.1728-4465.2011.00267.x. 152 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Nelson, K. N., A. S. Wallace, S. V. Sodha, D. Daniels, and V. Dietz. 2017. “Assessing Strategies for Increasing Urban Routine Immunization Coverage of Childhood Vaccines in Low and Middle-Income Countries: A Systematic Review of Peer-Reviewed Literature.” Vaccine 34 (46): 5495–503. doi:10.1016/j.vaccine.2016.09.038. Neuberger, A., J. Okebe, D. Yahav, and M. Paul. 2016. “Oral Iron Supplements for Children in Malaria-Endemic Areas.” Cochrane Database of Systematic Reviews 2016 (2): CD006589. doi:10.1002/14651858.CD006589.pub4. Nissensohn, M., A. Sánchez Villegas, D. Fuentes Lugo, P. Henríquez Sánchez, J. Doreste Alonso, N. M. Lowe, V. Hall Moran, A. L. Skinner, M. Warthon Medina, and L. Serra-Majem. 2013. “Effect of Zinc Intake on Serum/Plasma Zinc Status in Infants: A Meta-analysis. Maternal and Child Nutrition 9: 285–98. doi:10.1111/mcn.12045. Nissensohn, M., A. Sánchez Villegas, D. Fuentes Lugo, P. Henríquez Sánchez, J. Doreste Alonso, L. Peña Quintana, C. Ruano, N. L. Lowe, V. Hall Moran, A. L. Skinner, M. Warthon Medina, and L. Serra-Majem. 2016. “Effect of Zinc Intake on Growth in Infants: A Meta- analysis.” Critical Reviews in Food Science and Nutrition 56 (3): 350–63. doi:10.1080/10408398.2013.802661. Nissensohn, M., A. Sánchez Villegas, D. Fuentes Lugo, P. Henríquez Sánchez, J. Doreste Alonso, A. L. Skinner, M. Warthon Medina, N. M. Lowe, V. Hall Moran, and L. Serra-Majem. 2013. “Effect of Zinc Intake on Mental and Motor Development in Infants: A Meta- Analysis.” International Journal for Vitamin and Nutrition Research 83 (4): 203–15. doi:10.1024/0300-9831/a000161. Odendaal, W. A., K. Ward, J. Uneke, H. Uro-Chukwu, D. Chitama, Y. Balakrishna, and T. Kredo. 2018. “Contracting Out to Improve the Use of Clinical Health Services and Health Outcomes in Low- and Middle-Income Countries.” Cochrane Database of Systematic Reviews 2018 (4): CD008133. doi:10.1002/14651858.CD008133.pub2. Oliveira, I. B. B., L. P. Leal, M. W. de L. Coriolano-Marinus, A. H. da S. Santos, B. L. Horta, and C. M Pontes. 2017. “Meta-analysis of the Effectiveness of Educational Interventions for Breastfeeding Promotion Directed to the Woman and Her Social Network. Journal of Advanced Nursing 73 (2): 323–35. doi:10.1111/jan.13104. Oliveira, J. M., R. Allert, and C. E. East. 2016. “Vitamin A Supplementation for Postpartum Women.” Cochrane Database of Systematic Reviews 2016 (3): CD005944. doi:10.1002/14651858.CD005944.pub3. Olufunlayo, T. F., A. A. Roberts, C. MacArthur, N. Thomas, K. A. Odeyemi, M. Price, and K. Jolly. 2019. “Improving Exclusive Breastfeeding in Low and Middle‐Income Countries: A Systematic Review.” Maternal and Child Nutrition 15 (3): e12788. doi:10.1111/mcn.12788. Orton, L., A. Pennington, S. Nayak, A. Sowden, M. White, and M. Whitehead. 2016. “Group- Based Microfinance for Collective Empowerment: A Systematic Review of Health 153 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Impacts.” Bulletin of the World Health Organization 94 (9): 694–704A. doi:10.2471/BLT.15.168252. Owusu-Addo, E., and S. B. Owusu-Addo. 2014. “Effectiveness of Health Education in Community-Based Malaria Prevention and Control Interventions in Sub-Saharan Africa: A Systematic Review.” Journal of Biology, Agriculture and Healthcare 4 (3): 22–34. http://hdl.handle.net/123456789/8835. Owusu-Addo, E., A. M. N. Renzaho, A. S. Mahal, and B. J. Smith. 2016. “The Impact of Cash Transfers on Social Determinants of Health and Health Inequalities in Sub-Saharan Africa: A Systematic Review Protocol.” Systematic Reviews 5: 114. doi:10.1186/s13643-016- 0295-4. Oyo-Ita, A., C. S. Wiysonge, C. Oringanje, C. E. Nwachukwu, O. Oduwole, and M. M. Meremikwu. 2016. “Interventions for Improving Coverage of Childhood Immunisation in Low- and Middle-Income Countries.” Cochrane Database of Systematic Reviews 2016 (7): CD008145. doi:10.1002/14651858.CD008145.pub3. Pandey, V. L., S. M. Dev, and U. Jayachandran. 2016. “Impact of Agricultural Interventions on the Nutritional Status in South Asia: A Review.” Food Policy 62: 28–40. doi:10.1016/j.foodpol.2016.05.002 Panjwani, A., and R. Heidkamp. 2017. “Complementary Feeding Interventions Have a Small but Significant Impact on Linear and Ponderal Growth of Children in Low- and Middle- Income Countries: A Systematic Review and Meta-analysis.” Journal of Nutrition 147 (11 Suppl): 2169–78S. doi:10.3945/jn.116.243857. Panpanich, R., and P. Garner. 1999. “Growth Monitoring in Children.” Cochrane Database of Systematic Reviews 1999 (4): CD001443. doi:10.1002/14651858.CD001443. Pasricha, S., E. Hayes, K. Kalumba, and B. Biggs. 2013. “Effect of Daily Iron Supplementation on Health in Children Aged 4–23 Months: A Systematic Review and Meta-analysis of Randomized Controlled Trials.” Lancet Global Health 1 (2): e77–86. doi:10.1016/S2214- 109X(13)70046-9. Patel, A. B., M. Mamtani, N. Badhoniya, and H. Kulkarni. 2011. “What Zinc Supplementation Does and Does Not Achieve in Diarrhea Prevention: A Systematic Review and Meta- analysis.” BMC Infectious Diseases 11: 122. doi:10.1186/1471-2334-11-122. Pega, F., S. Y. Liu, S. Walter, R. Pabayo, R. Saith, and S. K. Lhachimi. 2017. “Unconditional Cash Transfers for Reducing Poverty and Vulnerabilities: Effect on Use of Health Services and Health Outcomes in Low- and Middle-Income Countries.” Cochrane Database of Systematic Reviews 2017 (11): CD011135. doi:10.1002/14651858.CD011135.pub2. Peña-Rosas, J. P., P. Mithra, B. Unnikrishnan, N. Kumar, L. M. De-Regil, N. S. Nair, M. N. Garcia‐ Casal, and J. A. Solon. 2019. “Fortification of Rice with Vitamins and Minerals for 154 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Addressing Micronutrient Malnutrition.” Cochrane Database of Systematic Reviews 2019 (10): CD009902. doi:10.1002/14651858.CD009902.pub2. Petry, N., I. Olofin, E. Boy, M. D. Angel, and F. Rohner. 2016. “The Effect of Low Dose Iron and Zinc Intake on Child Micronutrient Status and Development during the First 1,000 Days of Life: A Systematic Review and Meta-analysis.” Nutrients 8 (12): 773. doi:10.3390/nu8120773. Pilkington, G., S. Panday, M. N. Khatib, E. Kotas, R. A. Hill, P. Simkhada, and L. Jones. 2017. The Effectiveness of Community Engagement and Participation Approaches in Low and Middle Income Countries: A Review of Systematic Reviews with Particular Reference to the Countries of South Asia. London: EPPI-Centre, Social Science Research Unit, Institute of Education, University of London. Polec, L. A., J. Petkovic, V. Welch, E. Ueffing, E. T. Ghogomu, J. P. Pardo, M. Grabowsky, A. Attaran, G. A. Wells, and P. Tugwell. 2015. “Strategies to Increase the Ownership and Use of Insecticide‐Treated Bednets to Prevent Malaria.” Campbell Systematic Reviews 11 (1): 1–127. doi:10.4073/csr.2015.17. Prost, A., T. Colbourn, N. Seward, K. Azad, A. Coomarasamy, A. Copas, T. A. Houweling, E. Fottrell, A. Kuddus, S. Lewycka, C. MacArthur, D. Manandhar, J. Morrison, C. Mwansambo, N. Nair, B. Nambiar, D. Osrin, C. Pagel, T. Phiri, A. M. Pulkki-Brännström, M. Rosato, J. Skordis-Worrall, N. Saville, N. S. More, B. Shrestha, P. Tripathy, A. Wilson, and A. Costello. 2013. “Women’s Groups Practising Participatory Learning and Action to Improve Maternal and Newborn Health in Low-Resource Settings: A Systematic Review and Meta-analysis.” Lancet 381 (9879): 1736–46. doi:10.1016/S0140-6736(13)60685-6. Pryce, J., M. Richardson, and C. Lengeler. 2018. “Insecticide-Treated Nets for Preventing Malaria.” Cochrane Database of Systematic Reviews 2018 (11): CD000363. doi:10.1002/14651858.CD000363.pub3. Radeva-Petrova, D., K. Kayentao, F. O. ter Kuile, D. Sinclair, and P. Garner. 2014. “Drugs for Preventing Malaria in Pregnant Women in Endemic Areas: Any Drug Regimen versus Placebo or No Treatment.” Cochrane Database of Systematic Reviews 2014 (10): CD000169. doi:10.1002/14651858.CD000169.pub3. Rahman, A., J. Fisher, P. Bower, S. Luchters, T. Tran, M. T. Yasamy, S. Saxena, and W. Waheed. 2013. “Interventions for Common Perinatal Mental Disorders in Women in Low- and Middle-Income Countries: A Systematic Review and Meta-analysis.” Bulletin of the World Health Organization 91 (8): 593–601I. doi:10.2471/BLT.12.109819. Ramakrishnan, U., P. Nguyen, and R. Martorell. 2009. “Effects of Micronutrients on Growth of Children under 5 y of Age: Meta-analyses of Single and Multiple Nutrient Interventions.” American Journal of Clinical Nutrition 89 (1): 191–203. doi:10.3945/ajcn.2008.26862. 155 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Ramírez-Luzuriaga, M. J., L. M. Larson, V. Mannar, and R. Martorell. 2018. “Impact of Double- Fortified Salt with Iron and Iodine on Hemoglobin, Anemia, and Iron Deficiency Anemia: A Systematic Review and Meta-analysis.” Advances in Nutrition 9 (3): 207–18. doi:10.1093/advances/nmy008. Roberts, J. L., and A. D. Stein. 2017. “The Impact of Nutritional Interventions beyond the First 2 Years of Life on Linear Growth: A Systematic Review and Meta-analysis.” Advances in Nutrition 8 (2): 323–36. doi:10.3945/an.116.013938.s. Ruel, M. T., and H. Alderman. 2013. “Nutrition-Sensitive Interventions and Programmes: How Can They Help to Accelerate Progress in Improving Maternal and Child Nutrition?” Lancet 382 (9891): 536–51. doi:10.1016/S0140-6736(13)60843-0. Ruel, M. T., A. R. Quisumbing, and M. Balagamwala. 2018. “Nutrition-Sensitive Agriculture: What Have We Learned So Far?” Global Food Security 17: 128–53. doi:10.1016/j.gfs.2018.01.002. Sachdev, H. P. S., T. Gera, and P. Nestel. 2005. “Effect of Iron Supplementation on Mental and Motor Development in Children: Systematic Review of Randomised Controlled Trials.” Public Health Nutrition 8 (2): 117–32. doi:10.1079/PHN2004677. Sachdev, H. P. S., T. Gera, and P. Nestel. 2006. “Effect of Iron Supplementation on Physical Growth in Children: Systematic Review of Randomised Controlled Trials.” Public Health Nutrition 9 (7): 904–20. doi:10.1017/PHN2005918. Saeterdal, I., S. Lewin, A. Austvoll-Dahlgren, C. Glenton, and S. Munabi-Babigumira. 2014. “Interventions Aimed at Communities to Inform and/or Educate about Early Childhood Vaccination.” Cochrane Database of Systematic Reviews 2014 (11): CD010232. doi:10.1002/14651858.CD010232.pub2. Salam, R. A., B. A. Haider, Q. Humayun, and Z. A. Bhutta. 2015. “Effect of Administration of Antihelminthics for Soil-Transmitted Helminths during Pregnancy.” Cochrane Database of Systematic Reviews 2015 (6): CD005547. doi:10.1002/14651858.CD005547.pub3. Salam, R. A., C. MacPhail, J. K. Das, and Z. A. Bhutta. 2013. “Effectiveness of Micronutrient Powders (MNP) in Women and Children.” BMC Public Health 13 (Suppl 3): S22. doi:10.1186/1471-2458-13-S3-S22. Saronga, N. J., T. Burrows, C. E. Collins, A. M. Ashman, and M. E. Rollo. 2019. “mHealth Interventions Targeting Pregnancy Intakes in Low and Lower‐Middle Income Countries: Systematic Review.” Maternal and Child Nutrition 15 (2): e12777. doi:10.1111/mcn.12777. Scott, V. K., L. B. Gottschalk, K. Q. Wright, C. Twose, M. A. Bohren, M. E. Schmitt, and N. Ortayli. 2015. “Community Health Workers’ Provision of Family Planning Services in Low- and Middle-Income Countries: A Systematic Review of Effectiveness.” Studies in Family Planning 46 (3): 241–61. doi:10.1111/j.1728-4465.2015.00028.x. 156 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Seguin, M., and M. Niño Zarazúa. 2015. “Non-clinical Interventions for Acute Respiratory Infections and Diarrhoeal Diseases among Young Children in Developing Countries.” Tropical Medicine and International Health 20 (2): 146–69. doi:10.1111/tmi.12423. Sguassero, Y., M. de Onis, A. M. Bonotti, and G. Carroli. 2012. “Community-Based Supplementary Feeding for Promoting the Growth of Children under Five Years of Age in Low and Middle Income Countries.” Cochrane Database of Systematic Reviews 2012 (6): CD005039. doi:10.1002/14651858.CD005039.pub3. Shah, P. S., and A. Ohlsson. 2009. “Effects of Prenatal Multimicronutrient Supplementation on Pregnancy Outcomes: A Meta-analysis.” Canadian Medical Association Journal 180 (12): E99–108. https://doi.org/10.1503/cmaj.081777. Shakya, P., M. K. Kunieda, M. Koyama, S. S. Rai, M. Miyaguchi, S. Dhakal, S. Sandy, B. F. Sunguya, and M. Jimba. 2017. “Effectiveness of Community-Based Peer Support for Mothers to Improve Their Breastfeeding Practices: A Systematic Review and Meta- analysis. PLoS ONE 12 (5): e0177434. doi:10.1371/journal.pone.0177434. Sibley, L. M., T. A. Sipe, C. M. Brown, M. M. Diallo, K. McNatt, and N. Habarta. 2007. “Traditional Birth Attendant Training for Improving Health Behaviours and Pregnancy Outcomes.” Cochrane Database of Systematic Reviews 2007 (3): CD005460. doi:10.1002/14651858.CD005460.pub2. Soares, M. M., M. A. Silva, P. P. C. Garcia, L. S. Silva, G. D. Costa, R. M. A. Araújo, and R. M. M. Cotta. 2019. “Effect of Vitamin A Supplementation: A Systematic Review.” Ciência & Saúde Coletiva 24 (3): 827–38. doi:10.1590/1413-81232018243.07112017. Sondaal, S. F. V., J. L. Browne, M. Amoakoh-Coleman, A. Borgstein, A. Miltenburg, M. Verwijs, and K. Klipstein-Grobusch. 2016. “Assessing the Effect of mHealth Interventions in Improving Maternal and Neonatal Care in Low- and Middle-Income Countries: A Systematic Review.” PloS One 11 (5): e0154664. doi:10.1371/journal.pone.0154664. Soubeiga, D., L. Gauvin, M. A. Hatem, and M. Johri. 2014. “Birth Preparedness and Complication Readiness Interventions to Reduce Maternal and Neonatal Mortality in Developing Countries: Systematic Review and Meta-analysis.” BMC Pregnancy and Childbirth 14 (1): 129. doi:10.1186/1471-2393-14-129. Steinert, J. I., J. Zenker, U. Filipiak, A. Movsisyan, L. D. Cluver, and Y. Shenderovich. 2018. “Do Saving Promotion Interventions Increase Household Savings, Consumption, and Investments in Sub-Saharan Africa? A Systematic Review and Meta-analysis.” World Development 104: 238–56. doi:10.1016/j.worlddev.2017.11.018. Stevens, B., P. Buettner, K. Watt, A. Clough, J. Brimblecombe, and J. Judd. 2015. “The Effect of Balanced Protein Energy Supplementation in Undernourished Pregnant Women and Child Physical Growth in Low- and Middle-Income Countries: A Systematic Review and Meta-analysis.” Maternal and Child Nutrition 11 (4): 415–32. doi:10.1111/mcn.12183. 157 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Stewart, R., L. Langer, N. Rebelo Da Silva, E. Muchiri, H. Zaranyika, Y. Erasmus, N. Randall, S. Rafferty, M. Korth, N. Madinga, and T. de Wet. 2015. “The Effects of Training, Innovation and New Technology on African Smallholder Farmers’ Economic Outcomes and Food Security: A Systematic Review.” Campbell Systematic Reviews 11 (1): 1–224. doi:10.4073/csr.2015.16. Stewart, R., C. van Rooyen, M. Korth, A. Chereni, N. Rebelo Da Silva, and T. de Wet. 2012. Do Micro-credit, Micro-savings and Micro-leasing Serve as Effective Financial Inclusion Interventions Enabling Poor People, and Especially Women, to Engage in Meaningful Economic Opportunities in Low- and Middle-Income Countries? A Systematic Review of the Evidence. London: EPPI-Centre, Social Science Research Unit, Institute of Education, University of London. Suchdev, P. S., J. P. Peña-Rosas, and L. M. De-Regil. 2015. “Multiple Micronutrient Powders for Home (Point-of-Use) Fortification of Foods in Pregnant Women.” Cochrane Database of Systematic Reviews 2015 (6): CD011158. doi:10.1002/14651858.CD011158.pub2. Sudfeld, C. R., W. W. Fawzi, and C. Lahariya. 2012. “Peer Support and Exclusive Breastfeeding Duration in Low and Middle-Income Countries: A Systematic Review and Meta- analysis.” PLoS ONE 7 (9): e45143. doi:10.1371/journal.pone.0045143. Sunguya, B. F., K. C. Poudel, L. B. Mlunde, P. Shakya, D. P. Urassa, M. Jimba, and J. Yasuoka. 2013. “Effectiveness of Nutrition Training of Health Workers toward Improving Caregivers’ Feeding Practices for Children Aged Six Months to Two Years: A Systematic Review.” Nutrition Journal 12: 66. doi:10.1186/1475-2891-12-66. Tan, S. Y., and G. J. Melendez-Torres. 2018. “Do Prospective Payment Systems (PPS) Lead to Desirable Providers’ Incentives and Patients’ Outcomes? A Systematic Review of Evidence from Developing Countries.” Health Policy and Planning 33 (1): 137–53. doi:10.1093/heapol/czx151. Taylor-Robinson, D. C., N. Maayan, S. Donegan, M. Chaplin, and P. Garner. 2019. “Public Health Deworming Programmes for Soil-Transmitted Helminths in Children Living in Endemic Areas.” Cochrane Database of Systematic Reviews 2019 (9): CD000371. doi:10.1002/14651858.CD000371.pub7. Taylor-Robinson, D. C., N. Maayan, K. Soares-Weiser, S. Donegan, and P. Garner. 2015. “Deworming Drugs for Soil-Transmitted Intestinal Worms in Children: Effects on Nutritional Indicators, Haemoglobin, and School Performance.” Cochrane Database of Systematic Reviews 2015 (7): CD000371. doi:10.1002/14651858.CD000371.pub6. Tenório, M. B., R. C. Ferreira, F. A. Moura, N. B. Bueno, M. O. F. Goulart, and A. C. M. Oliveira. 2018. “Oral Antioxidant Therapy for Prevention and Treatment of Preeclampsia: Meta- analysis of Randomized Controlled Trials.” Nutrition, Metabolism and Cardiovascular Diseases 28 (9): 865–76. doi:10.1016/j.numecd.2018.06.002. 158 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Thakur, M., P. A. W. Nuyts, E. A. Boudewijns, J. Flores Kim, T. Faber, G. R. Babu, O. C. P. van Schayck, and J. V. Been. 2018. “Impact of Improved Cookstoves on Women’s and Child Health in Low and Middle Income Countries: A Systematic Review and Meta-analysis. Thorax 73 (11): 1026–40. doi:10.1136/thoraxjnl-2017-210952. Thinkhamrop, J., G. J. Hofmeyr, O. Adetoro, P. Lumbiganon, and E. Ota. 2015. “Antibiotic Prophylaxis during the Second and Third Trimester to Reduce Adverse Pregnancy Outcomes and Morbidity.” Cochrane Database of Systematic Reviews 2015 (6): CD002250. doi:10.1002/14651858.CD002250.pub3. Thompson, J., and B. -A. Biggs, and S. -R. Pasricha. 2019. “Effects of Daily Iron Supplementation in 2- to 5-Year-Old Children: Systematic Review and Meta-analysis.” Pediatrics 131 (4): 739–53. doi:10.1542/peds.2012-2256. Till, S. R., D. Everetts, and D. M. Haas. 2015. “Incentives for Increasing Prenatal Care Use by Women in Order to Improve Maternal and Neonatal Outcomes.” Cochrane Database of Systematic Reviews 2015 (12): CD009916. doi:10.1002/14651858.CD009916.pub2. Ton, G., S. Desiere, W. Vellema, S. Weituschat, and M. D’Haese. 2017. “The Effectiveness of Contract Farming for Raising Income of Smallholder Farmers in Low‐ and Middle‐ Income Countries: A Systematic Review.” Campbell Systematic Reviews 13 (1): 1–131. doi:10.4073/csr.2017.13. Tripathi, A., S. K. Kabra, H. P. S. Sachdev, and R. Lodha. 2016. “Home Visits by Community Health Workers to Improve Identification of Serious Illness and Care Seeking in Newborns and Young Infants from Low- and Middle-Income Countries.” Journal of Perinatology 36: S74–82. doi:10.1038/jp.2016.34. Vaessen, J., A. Rivas, M. Duvendack, R. P. Jones, F. Leeuw, G. van Gils, R. Lukach, N. Holvoet, J. Bastiaensen, J. G. Hombrados, and H. Waddington. 2014. “The Effects of Microcredit on Women’s Control over Household Spending in Developing Countries: A Systematic Review and Meta‐analysis.” Campbell Systematic Reviews 10 (1): 1–205. doi:10.4073/csr.2014.8. van Rooyen, C., R. Stewart, and T. de Wet. 2012. “The Impact of Microfinance in Sub-Saharan Africa: A Systematic Review of the Evidence.” World Development 40 (11): 2249–62. doi:10.1016/j.worlddev.2012.03.012. Venkataramanan, V., J. Crocker, A. Karon, and J. Bartram. 2018. “Community-Led Total Sanitation: A Mixed-Methods Systematic Review of Evidence and Its Quality.” Environmental Health Perspectives 126 (2): 026001. doi:10.1289/EHP1965. Visser, J., M. H. McLachlan, N. Maayan, and P. Garner. 2018. “Community-Based Supplementary Feeding for Food Insecure, Vulnerable and Malnourished Populations—An Overview of Systematic Reviews.” Cochrane Database of Systematic Reviews 2018 (11): CD010578. doi:10.1002/14651858.CD010578.pub2. 159 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Vucic, V., C. Berti, C. Vollhardt, K. Fekete, I. Cetin, B. Koletzko, M. Gurinovic, and P. van’t Veer. 2010. “Effect of Iron Intervention on Growth during Gestation, Infancy, Childhood, and Adolescence: A Systematic Review with Meta-analysis.” Nutrition Reviews 71 (6): 386–401. doi:10.1111/nure.12037. Waddington, H., B. Snilstveit, J. Hombrados, M. Vojtkova, D. Phillips, P. Davies, and H. White. 2014. “Farmer Field Schools for Improving Farming Practices and Farmer Outcomes: A Systematic Review.” Campbell Systematic Reviews 10 (1): i–335. doi:10.4073/CSR.2014.6. Waddington, H., B. Snilstveit, H. White, and L. Fewtrell. 2009. “Water, Sanitation and Hygiene Interventions to Combat Childhood Diarrhoea in Developing Countries.” 3ie Synthetic Review 001. Delhi: International Initiative for Impact Evaluation. Walker, C. L. F., and R. E. Black. 2010. “Zinc for the Treatment of Diarrhoea: Effect on Diarrhoea Morbidity, Mortality and Incidence of Future Episodes.” International Journal of Epidemiology 39 (Suppl 1): i63–69. doi:10.1093/ije/dyq023. Wallace, A., V. Dietz, and K. L. Cairns. 2009. “Integration of Immunization Services with Other Health Interventions in the Developing World: What Works and Why? Systematic Literature Review.” Tropical Medicine and International Health 14 (1): 11–19. doi:10.1111/j.1365-3156.2008.02196.x. Watson, J. A., J. H. J. Ensink, M. Ramos, P. Benelli, E. Holdsworth, R. Dreibelbis, and O. Cumming. 2017. “Does Targeting Children with Hygiene Promotion Messages Work? The Effect of Handwashing Promotion Targeted at Children, on Diarrhoea, Soil- Transmitted Helminth Infections and Behaviour Change, in Low- and Middle-Income Countries.” Tropical Medicine and International Health 22 (5): 526–38. doi:10.1111/tmi.12861. Webel, A. R., J. Okonsky, J. Trompeta, and W. L. Holzemer. 2010. “A Systematic Review of the Effectiveness of Peer-Based Interventions on Health-Related Behaviors in Adults.” American Journal of Public Health 100 (2): 247–53. doi:10.2105/AJPH.2008.149419. Wekesah, F. M., C. E. Mbada, A. S. Muula, C. W. Kabiru, S. K. Muthuri, and C. O. Izugbara. 2016. "Effective Non-drug Interventions for Improving Outcomes and Quality of Maternal Health Care in Sub-Saharan Africa: A Systematic Review.” Systematic Reviews 5: 137. doi:10.1186/s13643-016-0305-6. Welch, V. A., E. Ghogomu, A. Hossain, S. Awasthi, Z. A. Bhutta, C. Cumberbatch, R. Fletcher, J. McGowan, S. Krishnaratne, E. Kristjansson, S. Sohani, S. Suresh, P. Tugwell, H. White, and G. A. Wells. 2017. “Mass Deworming to Improve Developmental Health and Wellbeing of Children in Low-Income and Middle-Income Countries: A Systematic Review and Network Meta-analysis.” Lancet Global Health 5 (1): e40–50. doi:10.1016/S2214-109X(16)30242-X. Welch, V. A., E. Ghogomu, A. Hossain, A. Riddle, M. Gaffey, P. Arora, O. Dewidar, R. Salaam, S. Cousens, R. Black, T. D. Hollingsworth, S. Horton, P. Tugwell, D. Bundy, M. C. Castro, A. 160 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Eliott, H. Friis, H. T. Le, C. Liu, E. K. Rousham, F. Rohner, C. King, E. Sartono, T. Supali, P. Steinmann, E. Webb, F. Wieringa, P. Winnichagoon, M. Yazdanbakhsh, Z. A. Bhutta, G. Wells. 2019. “Mass Deworming for Improving Health and Cognition of Children in Endemic Helminth Areas: A Systematic Review and Individual Participant Data Network Meta‐analysis.” Campbell Systematic Reviews 15 (4): e1058. doi:10.1002/cl2.1058. Wolf, J., P. R. Hunter, M. C. Freeman, O. Cumming, T. Clasen, J. Bartram, J. P. T. Higgins, R. Johnston, K. Medlicott, S. Boisson, and A. Prüss-Ustün. 2018. “Impact of Drinking Water, Sanitation and Handwashing with Soap on Childhood Diarrhoeal Disease: Updated Meta-analysis and Meta-regression.” Tropical Medicine and International Health 23 (5): 508– 25. doi:10.1111/tmi.13051. Wolf, J., A. Prüss-Ustün, O. Cumming, J. Bartram, S. Bonjour, S. Cairncross, T. Clasen, J. M. Colford Jr, V. Curtis, J. De France, L. Fewtrell, M. C. Freeman, B. Gordon, P. R. Hunter, A. Jeandron, R. B. Johnston, D. Mäusezahl, C. Mathers, M. Neira, and J. P. T. Higgins. 2014. “Assessing the Impact of Drinking Water and Sanitation on Diarrhoeal Disease in Low- and Middle-Income Settings: Systematic Review and Meta-regression.” Tropical Medicine and International Health 19 (8): 928–42. doi:10.1111/tmi.12331. World Bank. 2015. Impacts of Interventions during Early Childhood on Later Outcomes: A Systematic Review. Independent Evaluation Group. Washington, DC: World Bank. Yakoob, M. Y., E. Theodoratou, A. Jabeen, A. Imdad, T. P. Eisele, J. Ferguson, A. Jhass, I. Rudan, H. Campbell, R. E. Black, and Z. A. Bhutta. 2011. “Preventive Zinc Supplementation in Developing Countries: Impact on Mortality and Morbidity Due to Diarrhea, Pneumonia and Malaria.” BMC Public Health 11 (Suppl 3): S23. doi:10.1186/1471-2458-11-S3-S23. Yoong, J., L. Rabinovich, and S. Diepeveen. 2012. The Impact of Economic Resource Transfers to Women versus Men: A Systematic Review. Technical report. London: EPPI-Centre, Social Science Research Unit, Institute of Education, University of London. Zeng, W., G. Li, H. Ahn, H. T. H. Nguyen, D. S. Shepard, and D. Nair.. 2017. “Cost-Effectiveness of Health Systems Strengthening Interventions in Improving Maternal and Child Health in Low- and Middle-Income Countries: A Systematic Review. Health Policy and Planning 33 (2): 283–97. doi:10.1093/heapol/czx172. Zhou, S. J., A. J. Anderson, R. A. Gibson, and M. Makrides. 2013. “Effect of Iodine Supplementation in Pregnancy on Child Development and Other Clinical Outcomes: A Systematic Review of Randomized Controlled Trials.” American Journal of Clinical Nutrition 98 (5): 1241–54. doi:10.3945/ajcn.113.065854. 161 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Table B.2. Inclusion Keywords Outcomes of Interest Search Keywords Child undernutrition and development child undernutrition OR undernutrition Birthweight birthweight OR low birthweight Linear growth stunt* OR height-for-age z score* OR length-for-age z score* OR linear growth OR child growth OR infant growth OR foetal growth OR fetal growth OR length OR height Cognitive development Cognitive development OR psychosocial development OR motor development OR attachment Micronutrient status Haemoglobin OR hemoglobin OR serum ferritin OR serum vitamin D or plasma zinc OR plasma folate OR iodine OR serum retinol Breastfeeding practices breastfeed* Breastfeeding initiation delayed initiation OR early initiation OR timely initiation Exclusive breastfeeding exclusive breastfeed* OR mixed feed* OR infant formula OR formula feed* OR pre-lacteal Breastfeeding duration early cessation OR breastfeed* duration OR continued breastfeed* OR any breastfeeding Complementary feeding practices complementary feed* OR complementary food* OR infant feed* OR child feed* Diet nutrient intake* OR micronutrient* OR diet* OR excessive intake* OR diet* divers* OR food intake*OR food group* OR food quality OR diet quality OR animal-source food OR meat consumption OR antinutrient* OR phytate* OR energy intake* OR macronutrient* Weaning weaning OR weaning food* OR introduction of solid food* OR appropriate food* OR appropriate feed* OR food consistency OR meal frequency OR feed* amount OR feed* quantity Responsive feeding responsive feed* OR responsive care OR feeding during illness Childcare practices child care OR child caregiv* OR caregiv* Parenting psychosocial care OR Psychosocial stimulation OR parent* style OR child development OR cognitive stimulation OR cognitive development OR father* OR paternal care OR father engagement OR male engagement Care seeking uptake OR utilisation OR use OR immunization OR vaccination OR growth monitoring* OR care seeking Appropriate caregivers alternative caregiv* OR secondary caregiv* OR childcare OR daycare OR day care 162 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Food and water safety food safety OR water safety OR water quality OR clean water OR safe water OR drink* water Fecal contamination faecal* OR fecal* OR excreta Hygiene practices hand wash* OR hand-wash* OR hygiene OR soap Food preparation and storage food storage OR food preparation Harmful bacteria and toxins listeria OR listeriosis OR toxoplasmosis OR aflatoxin OR mercury OR pesticid* Child health and disease child* illness OR child* disease OR child* infectio* Enteric infection diarrhoea* OR diarrhea* OR enteric OR enteropathy OR helmin* Respiratory infection respiratory infection* OR respiratory tract infection* OR cough OR pneumonia Fever fever Micronutrient deficiency anaemia OR anemia OR deficienc* Malaria malaria Appetite appetite Inflammation, air pollution inflammation OR air pollution Maternal factors Intergeneration* transmission Nutrition maternal undernutrition OR maternal underweight OR thin* OR maternal height OR maternal stature Diet, nutrient intake women* diet OR prenatal diet* OR antenatal diet* OR maternal diet* OR maternal nutrient intake* OR food taboo* OR diet* restriction Health, infection, deficiency women* infectio* OR prenatal infectio* OR antenatal infectio* OR maternal infectio* OR maternal health OR deficienc* Early pregnancy adolescent pregnancy OR teen pregnancy OR teen mother* OR early age OR adolescent* OR teen* OR child marriage Mental health depression OR stress OR distress OR anxiety OR mental health OR self-esteem Women’s status, work, empowerment, gender women* autonomy OR women* empowerment OR women* decision-making OR autonomy OR empowerment OR decision-making IUGR, preterm, SGA, gestational weight gain intrauterine growth restriction OR low birthweight OR pre-term OR preterm Birth spacing birth interval OR birth spacing OR family planning OR contraception Hypertension pre-eclampsia OR eclampsia OR high blood pressure OR hypertensi* Knowledge, intention education level OR literacy OR maternal knowledge OR caregiver knowledge OR education* OR training 163 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio Utilization of services antenatal care OR prenatal care OR postnatal care OR postpartum care OR vaccination Home environment home environment OR home safety Sanitation sanitation OR latrine OR faeces OR feces OR faecal OR fecal OR WASH OR disposal OR fly population Water supply water supply OR water access Food and nutrition security food security OR food insecurity OR household food supply OR food environment OR food system* OR genetic* modifi* OR biofortifi* OR fortifi*OR household diet* diversity Intra-household food allocation intra-household OR food allocation Domestic abuse domestic violence OR gender-based violence OR intimate partner violence Household income and costs Household income OR out-of-pocket Source: Independent Evaluation Group. Note: IUGR= intrauterine growth restriction; SGA = small for gestational age. * Table B.3. Exclusion Keywords Domain Exclusion Keywords Contexts and populations biracial OR crisis OR critically ill OR disaster* OR elderly OR foster OR hospitalized OR hospitalization OR “in low birthweight” OR nosocomial OR older adult* OR older people OR parenter* OR in preterm OR seriously ill OR terminally ill OR trauma* OR travel* OR in very low birthweight Conditions acne OR allerg* OR anthrax OR arthritis OR asthma OR atop* OR autism OR autoimmune OR bone OR bowel OR cancer OR cardiac OR cardiovascular OR cardiology OR celiac OR cerebral* OR caesarean OR cholesterol OR cleft OR colic OR colitis OR crohn* OR cushing* OR cystic* OR dental OR dermat* OR dysmenorrhea OR ebola OR epileps* OR haemorrhage OR headache OR hearing loss OR hepatitis OR hirsutism OR HIV OR hypoglycemia OR infert* OR kawasaki OR kidney OR laryngitis OR leprosy OR leukaemia OR lung OR lymphoma OR macular OR melanoma OR Ménière* OR migraine OR miscarriage OR myocardial OR pancrea* OR parkinson* OR non- communicable disease* OR pain OR palsy OR psoriasis OR pulmonary OR reflux OR sclerosis OR seizure OR sepsis OR sexual dys* OR sickle OR speech OR spina bifida OR spinal OR spine OR stillbirth OR strep OR thyroid* OR tuberculosis OR ulcer* OR urinary OR west nile OR zika Mental health, behaviors, and cognitive issues ADHD OR alzheimer* OR anorexia OR borderline personality disorder OR bulimia OR cannabis OR child 164 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio abuse OR child maltreatment OR dementia OR gang OR obsessive compulsive OR opioid OR schizophrenia OR smoking Medical procedures, treatments, and diagnoses abortion OR catheter* OR dialysis OR fertility treatment OR in-vitro OR IVF OR surgery OR ventilator Source: Independent Evaluation Group. Notes for Figures B.2–B.19 1Portfolio review categories for nutrition and dietary support include the following nutrition- specific interventions: micronutrient supplementation for children; provision of supplementary feeding for children; provision of supplementary micronutrient foods for children; provision of supplementary energy and protein supplement (macronutrients) for women; micronutrient supplementation for women; and SBCC on nutrition and health practices. 2 In this instance of supplementary feeding, birthweight pertains to the next generation. 3Portfolio review categories for child disease prevention and treatment interventions include the following: deworming (such as soil-transmitted helminthiasis); other integrated management of childhood illness for children; malaria prevention (promotion on the use of ITNs with or without the provision of ITNs; treatment of diarrhea (oral rehydration therapy or solution); and treatment of moderate or severe malnutrition of children (such as ready-to-use supplementary food). Portfolio review categories for health and family planning services interventions include the following: supporting demand for family planning and contraception (provision of contraceptives, emergency contraception for birth spacing, and adolescent pregnancy); supporting demand for health care services use, such as fee removal; and supporting maternal mental health. 4Portfolio review categories for WASH approaches include the following: hygiene interventions, such as SBCC promotion and counseling on handwashing with soap and support for a healthy home environment (for example, reduction of indoor air pollution, provision of cookstoves, and fly control); sanitation interventions, such as the provision, construction and promotion of latrines; and water interventions, such as water treatment at point-of-use (for example, chlorination and filtration and solar disinfection), water treatment at source (for example, protected wells, communal tap stands, or chlorination and filtration of community sources), and support for piped water. 5Portfolio review categories for agriculture interventions include the following: support for small-scale animal protein development and livestock production (dairy development, animal husbandry, and poultry development); support for small-scale fisheries and aquaculture; support for development of home garden (household and community own production); support for food safety (food preparation and storage); support for food production, diversification, agriculture machinery and technology, contract farming, food safety (pesticides); fortification and biofortification of food products; cash cropping, that is production of coffee, sugarcane, or other crops intended for sale rather than home consumption, goal income earned from growing cash crops; and support for commercialization of food products (marketed production of crops and animal products by smallholder farmers). 165 Appendix B Systematic Review Map and Relevance of the World Bank Nutrition Portfolio 6Portfolio review categories for early childhood development interventions include support for child play spaces (stimulation environment) and school feeding programs. Social safety nets include cash transfers for families with children or in-kind transfers for families with children (school uniforms). 7Portfolio review categories for institutional strengthening offer support to improve nutrition service delivery and supply and policy, financing, and coordination (nutrition financing and budgeting). Notes 1Nutrition-specific interventions or programs address the immediate determinants of fetal and child nutrition and development—adequate food and nutrient intake; feeding, caregiving, and parenting practices; and low burden of infectious diseases. Nutrition-sensitive interventions or programs address the underlying determinants of fetal and child nutrition and development— food security; adequate caregiving resources at the maternal, household and community levels; and access to health services and a safe and hygienic environment—and incorporate specific nutrition goals and action (Ruel and Alderman 2013). 2Keywords lists are available at the end of the appendix. Context-sensitive keywords are related to document type, study type, context (such as developed countries), health outcomes (such as obesity), population (such as migrant or twin), intervention (such as vaccine or clinical), and condition (such as colic). 3Income-level status is based on Bank Group lending status in 2018. Countries that changed from lower-middle-income status to upper-middle-income status during the evaluation timeline are included (for example, Guatemala). 4For example, institutional delivery is included in the literature review due to its link to early initiation of breastfeeding, whereas skilled birth attendance is not. 5 As shown in SRM Figures B.2–B.19: Positive indicates that the pooled effect (for meta-analyses) or all underlying studies (for narrative syntheses) of the intervention are found to have a positive effect on the outcome of interest. No effect indicates that the intervention is neither significantly positive nor significantly negative on the outcome of interest. Inconsistent indicates that for a narrative synthesis the evidence of a particular intervention on a specific outcome shows a mix of positive and no effects across the underlying studies. Negative indicates that the intervention is found to have a negative effect on the outcome of interest. Given the direction of the evidence, the dark- and medium-green legends indicate that the evidence of an intervention on a particular outcome is found to be positive in more than three systematic reviews or in up to three systematic reviews, respectively. Similarly, the dark-red legend indicates that the evidence of an intervention on a particular outcome is found to be negative in more than two systematic reviews. The light-green legend indicates that the pool of evidence of a particular intervention on a specific outcome shows a mix of positive effects, no effect, or a combination of both (inconsistent) in narrative synthesis. No evidence indicates that there are no systematic reviews identified in the review. 166 Appendix C. Behavior Change Process Map A process map is a practical tool that conveys the relationships and sequencing among inputs, outputs, outcomes, and longer-term impacts across different groups of actors who have roles for achieving development objectives. Mapping the processes through which the World Bank’s interventions in the nutrition portfolio help facilitate and support behavior change provides guidance that can be applied both for evaluating the contributions of the World Bank in addressing undernutrition and for planning engagements that are more likely to support sustainable behavior change in the future. A multitiered approach is used to understand the behavior change sequence and to analyze the extent to which the World Bank has supported interventions within the nutrition portfolio at stages along this sequence. This exercise starts with a structured literature review to identify and categorize behavior change concepts and evidence of how interventions have supported behavior change processes. These findings are used to develop process maps for benchmarking behavior change in projects. The resulting process maps reflect basic results chains by types of actors that can contribute to improvements in nutrition determinants and be adapted to a country context. Structured Literature Review The structured literature review was conducted to understand the incremental sequences of actions that can lead to sustained behaviors to improve nutrition determinants (access to food, caregiving resources, health services, and water, sanitation, and hygiene [WASH]). The review includes qualitative studies, such as qualitative systematic review and empirical studies, on behavior change interventions. The search protocol (i) uses a list of keywords to search databases (PubMed, Econlit); (ii) uses a snowball sampling approach to identify other relevant sources cited in references; and (iii) identifies references recommended in consultations. Five categories of keywords are used to search for articles for the review: keywords related to undernutrition and stunted growth; keywords related to nutrition determinants (breastfeeding, dietary, diversity, and so on); keywords related to types of actors; keywords related to the type of study; and keywords to limit the search to countries or regions. The inclusion criteria for the structured review are that the study provides evidence of effectiveness (that is, confirming that an intervention had facilitated a behavior change related to a nutrition determinant), the intervention and target population(s) are in a low- or middle-income country, the study is published within the past 10 years (2009 or later), and the intervention is designed to address undernutrition rather than obesity. All the reviewed studies are published in English. The initial list of 151 publications was 167 Appendix C Behavior Change Process Map reviewed manually to exclude any studies that did not report on outcomes using transparent, objective, and consistent indicators with some evidence of a results chain (that is, a description of a causal or contributing link between an intervention and improved outcomes). This process yielded a final list of 57 sources that provides relevant qualitative details related to the pathways and processes needed for behavior change. The literature review also reviewed existing behavior change frameworks to identify key concepts from existing behavior change experiences (box C.1). These concepts are used to frame the synthesis of the evidence from the literature using qualitative modeling, as described in the next section. Process Maps to Analyze Behavior Change Box C.1 Key Behavior Change Concepts Actions by change agents. The review of the studies identifies evidence on accelerator actions that may influence one or more behaviors to improve nutrition determinants, as well as related actions that may support caregivers to carry out accelerator behaviors. These changes are identified for the sphere of actors that can influence the mother or caregiver. This draws on the Communication for Development (C4D) framework, the Framework for Scaling Up Infant and Young Child Feeding (IYCF), and The Behavioral Change Framework (UNICEF 2018; Alive and Thrive 2016; USAID 2015). Capabilities and barriers to motivating behaviors. The review of the studies identifies evidence on actions to influence capabilities, and systemic barriers that, if addressed, may motivate behaviors to improve nutrition determinants. This approach is informed by the World Bank’s Mind, Behavior, and Development (eMBeD) group’s work to identify barriers or biases to behaviors, the CrI2SP Framework, the COM-B model, and the Integrated Behavior Model for Water, Sanitation, and Hygiene (IBM-WASH) (Flanagan and Tanner 2016; World Bank 2015; Michie 2011; Dreibelbis et al. 2013). In a country context, the specific barriers are diagnosed to inform intervention design. Measuring progress toward institutional change. The review of the studies identifies evidence to articulate the incremental sequence from activities and outputs to sustained behavior modification. This draws on the Institutional Change Assessment Method and the COM-B theory of change model also informed this process (Roberts and Torkos 2017; Michie et al. 2011). Tracing the sequence from inputs and interventions to outputs and longer-term outcomes provides a basis for understanding how to achieve sustainable behavior change. The first step in this exercise was to synthesize the evidence from the literature as it related to the nutrition determinants (access to food, caregiving resources, health services, and WASH), and types of actors (figure C.1). Synthesizing the evidence by actor provides an understanding of the interactions among change agents and mothers and primary caregiver beneficiaries to improve nutrition determinants. Policy actors are outside the scope of the analysis, and primary caregivers have served dual roles in the analysis, sometimes acting as change agents empowered to shift behaviors and sometimes serving as beneficiaries. Evidence related to determinants of food and 168 Appendix C Behavior Change Process Map caregiving (food and care) is merged given the overlap between feeding and caregiving behaviors in the literature. Figure C.1. Actors for Behavior Change Interventions to Reduce Child Undernutrition Source: Independent Evaluation Group. Evidence regarding the sequence of changes by actor groups presented in the literature was used to draft results chains or “process maps,” by determinant and actor. Each study typically provides evidence supporting only selected steps along the results chain; these examples were collectively analyzed to understand changes in actions across actors. The evidence on behavior change processes from the literature is synthesized in table C.1 by determinant, actor, and levels of the results chain. These progressive changes reflect the translation of information into action assuming that there is access to and control of other resources. Examples of indicators tracking progress for changes in behavior by different actors were also extracted during the literature review for each part of the results chain. These indicators are not exhaustive, but instead, illustrate actual measurement approaches used in the reviewed literature. The measurement of knowledge transfer and use relies heavily on qualitative approaches (interviews and focus groups) given the limited availability of existing timely data from administrative sources (management information systems) or household surveys. A notable contribution of the process maps is the delineation of levels along a results chain (engage-learn-apply-sustain), which lead from initial inputs and outputs all the way to sustained behavior change that could be expected to persist after interventions are completed. Behavior change processes are rarely linear and require interactions 169 Appendix C Behavior Change Process Map across actors and complex reinforcement (or looping) among different types of outputs and outcomes; nonetheless, the definition of clear levels of progress toward behavior change could facilitate effective planning and evaluation (figure C.2). Figure C.2. Tracing Evidence of Behavior Change in Actors Source: Independent Evaluation Group. The synthesized literature is used to develop process maps for key groups of actors that highlight behavior changes for supporting nutrition determinants. Although this progression is complex—with non-linear interactions across types of actors and many contextual variations in how actions are carried out in countries—the articulation of results chains is useful to outline inputs, outputs and outcomes, which can facilitate measurement, as well as the understanding of more complex, contextualized theories of change. An aggregate process map is provided in figure C.3, with results chains that are synthesized across actor groups (mothers and caregivers, family and household, and community and service providers) to understand how factors across the determinants can play a role in inhibiting or advancing achievements. Figures C.4–C.7 expand the process map for a more detailed understanding of the incremental progression for each type of actor. Together, the classification of actors, synthesis of literature, and indicators for measuring incremental changes, and the process map diagrams serve as tools to benchmark nutrition interventions facilitating behavior change. The maps could help inform project planning and evaluation within each area of nutrition determinant. 170 Appendix C Behavior Change Process Map Table C.1. Synthesis Tables Tracing Behavior Change Levels by Determinant and Actor 171 Appendix C Behavior Change Process Map Sources: Agrawal et al. 2012; Baker, Hajeebhoy, and Abrha 2013; Balogun et al. 2015; Bazzano et al. 2017; Belachew et al. 2012; Benson 2015; Black et al. 2015; Darrouzet-Nardi et al. 2016; Favara 2018; Fitzsommons et al. 2012; Gelli et al. 2018; Goyal and Sekher 2016; Haroon et al. 2013; Haselow, Stormer, andPries 2016; Hilmye et al. 2011; Hirani and Roozina 2012; Imdad and Bhutta 2011; Kavle et al. 2017; Kumar, Harris, and Rawat 2015; Malapit et al. 2015; Mduduzi et al. 2015; Nguyen et al. 2018; Noack and Pouw 2015; Numeri et al. 2018; Reerink et al. 2017; Sanghvi et al. 2013; Shi et al. 2010; Yourkavitch et al. 2017. Note: EBF = exclusive breastfeeding; IYCF = infant and young child feeding; NGO = nongovernmental organization. 172 Appendix C Behavior Change Process Map Sources: Alzua et al. 2015; Bauza, Routray, and Clasen 2019; Cumming andCairncross 2016; Davis et al. 2011; Islam et al. 2013; Luby et al. 2018; Newson et al. 2013; Pattanayak et al. 2009; Unicomb et al. 2018; Venkataramanan, Crocker, and Bartram 2018; Watson et al. 2017; Wodnik et al. 2018; Wood and Kols 2012; WHO, USAID, and UNICEF 2015. 173 Appendix C Behavior Change Process Map 174 Appendix C Behavior Change Process Map 175 Appendix C Behavior Change Process Map Sources: Agrawal et al. 2012; CORE Group 2009; Dewey and Mayers 2011; Manda-Taylor et al. 2017; Mutanda, Waiswa, and Namutamba 2016; Okuga et al. 2017; Pell et al. 2013; Rahman et al. 2018; Taleb et al. 2015; Watterson, Walsh, and Madeka 2015. Note: ANC = antenatal care; NGO = nongovernmental organization; ORT = oral rehydration therapy; PNC = postnatal care. 176 Appendix C Behavior Change Process Map Figure C.3. Behavior Change Process Map by Actor Type Source: Independent Evaluation Group. 177 Appendix C Behavior Change Process Map Figure C.4. Behavior Change Process Map for Mothers or Primary Caregivers Source: Independent Evaluation Group. 178 Appendix C Behavior Change Process Map Figure C.5. Behavior Change Process Map for Family/Household Source: Independent Evaluation Group. 179 Appendix C Behavior Change Process Map Figure C.6. Behavior Change Process Map for Community Actors Source: Independent Evaluation Group. 180 Appendix C Behavior Change Process Map Figure C.7. Behavior Change Process Map for Service Providers/Government and Nongovernment Source: Independent Evaluation Group. References General Guidance for Conceptual Approach Alive and Thrive. 2016. “Changing Behaviors, Improving Lives: Scaling Up Nutrition.” https://www.aliveandthrive.org/wp-content/uploads/2018/07/Overview-brief-final.pdf Dreibelbis, R., P. J. Winch, E. Leontsini, K. R. S. Hulland, P. K. Ram, L. Unicomb, and S. P. Luby. 2013. “The Integrated Behavioral Model for Water, Sanitation, and Hygiene: a systematic 181 Appendix C Behavior Change Process Map review of behavioral models and a framework for designing and evaluating behavior change interventions in infrastructure-restricted settings.” BMC Public Health. 2013(13): 1015. Published 2013 Oct 26. Flanagan, Ann Elizabeth, and Jeffery Tanner. 2016. A framework for evaluating behavior change in international development operations (Vol. 2) (English). IEG Working Paper 2016/No. 2. Washington, DC: World Bank Group. Mayne, J. 2015. “Useful Theory of Change Models.” Canadian Journal of Program Evaluation 30(2): 119–142. Michie, Susan, Maartje M. van Stralen, and Robert West. 2011. “The Behavior Change Wheel: A New Method for Characterizing and Designing Behavior Change Interventions.” Implementation Science 6: 42. http://doi.org/10.1186/1748-5908-6-42. Mosler, HJ. 2012. “A systematic approach to behaviour change interventions for the water and sanitation sector in developing countries: a conceptual model, a review and a guideline.” Int J Environ Health Res. 22(5): 431–49. Roberts, D., and A. Torkos. 2017. Results of the World Bank’s RAS program in Romania, 2012–15. Washington, DC: World Bank Group. United Nations Children’s Fund (UNICEF). 2018. Communication for Development (C4D) Strategic Framework 2018–21. UNICEF Regional Office for South Asia. Kathmandu. United States Agency for International Development (USAID). 2015. The Behavior Change Framework: A template for accelerating the impact of behavior change in USAID-supported MCH programs in 24 priority countries. Washington, DC World Bank. 2015. World Development 2015: Mind, Society, and Behavior. Washington, DC: World Bank Group. Food and Care Agrawal, P. K., S. Agrawal., S. Ahmed, G. L. Darmstadt, E. K. Williams, H. E. Rosen, and A. H. Baqui. 2012. “Effect of knowledge of community health workers on essential newborn health care: a study from rural India.” Health policy and planning 27(2): 115–126. Baker, J., T. Sanghvi, N. Hajeebhoy, and T. H. Abrha. 2013. “Learning from the Design and Implementation of Large-Scale Programs to Improve Infant and Young Child Feeding.” Food and Nutrition Bulletin, 34(3_suppl2), S226–S230. Balogun, O., A. Dagvadorj, K. Anigo, E. Ota, and S. Sasaki. 2015. “Factors influencing breastfeeding exclusivity during the first 6 months of life in developing countries: a quantitative and qualitative systematic review.” Maternal & child nutrition. 11. 10.1111/mcn.12180. 182 Appendix C Behavior Change Process Map Bazzano, A. N., A. Kaji, E. Felker-Kantor, L. A. Bazzano, and K. S. Potts. 2017. “Qualitative Studies of Infant and Young Child Feeding in Lower-Income Countries: A Systematic Review and Synthesis of Dietary Patterns.” Nutrients 9(10): 1140. doi:10.3390/nu9101140. Belachew, T., D. Lindstrom, A. Gebremariam, C. Jira, M. K. Hattori, C. Lachat, … P. Kolsteren. 2012. “Predictors of chronic food insecurity among adolescents in Southwest Ethiopia: a longitudinal study.” BMC public health 12: 604. doi:10.1186/1471-2458-12-604 Benson, Todd. 2015. “Association between Irrigated Farming and Improved Nutrition in Farm Households in Malawi.” Agrekon 54(3): 62–86. Black, M. M., R. Pérez-Escamilla, and S. F. Rao. 2015. “Integrating nutrition and child development interventions: scientific basis, evidence of impact, and implementation considerations.” Advances in Nutrition (Bethesda, Md.) 6(6): 852–859. doi:10.3945/an.115.010348. Darrouzet-Nardi, A., L. Miller, N. Joshi, S. Mahato, M. Lohani, and B. Rogers. 2016. “Child Dietary Quality in Rural Nepal: Effectiveness of a Community-Level Development Intervention.” Food Policy 61(0): 185–97. Favara, M. 2018. “Maternal Group Participation and Child Nutritional Status in Peru.” Review of Development Economics 22(2): 459–83. Fitzsimons, E., B. Malde, A. Mesnard, and M. Vera-Hernandez. 2012. “Household Responses to Information on Child Nutrition: Experimental Evidence from Malawi.” CEPR Discussion Papers, CEPR Discussion Papers: 8915. (OR also published by Institute for Fiscal Studies, IFS Working Papers: W–12/07, 2012.) Gelli, A., A. Margolies, M. Santacroce, N. Roschnik, A. Twalibu, M. Katundu, … M. Ruel. 2018. “Using a Community-Based Early Childhood Development Center as a Platform to Promote Production and Consumption Diversity Increases Children’s Dietary Intake and Reduces Stunting in Malawi: A Cluster-Randomized Trial.” The Journal of nutrition, 148(10): 1587–1597. doi:10.1093/jn/nxy148. Geresomo, Numeri C., Elizabeth Kamau Mbuthia, Joseph W. Matofari, and Agnes M. Mwangwela. 2018. “Targeting caregivers with context specific behavior change training increased uptake of recommended hygiene practices during food preparation and complementary feeding in Dedza district of Central Malawi.” Ecology of Food and Nutrition 57(4): 301–313, DOI: 10.1080/03670244.2018.1492379. Goyal, Jaya, and Madhushree Sekher. 2016. “Accountability, Nutrition and Local Institutions in India.” Development 58(1): 79–87. Haroon, S., J. K. Das, R. A. Salam, A. Imdad, and Z. A. Bhutta. 2013. “Breastfeeding promotion interventions and breastfeeding practices: a systematic review.” BMC public health 13 Suppl 3(Suppl 3), S20. 183 Appendix C Behavior Change Process Map Haselow, N. J., A. Stormer, and A. Pries. 2016. “Evidence-based evolution of an integrated nutrition-focused agriculture approach to address the underlying determinants of stunted growth.” Maternal & child nutrition, 12 Suppl 1(Suppl 1), 155–168. doi:10.1111/mcn.12260. Hilmiye Aksu, Mert Küçük, and Gülergün Düzgün. 2011. “The effect of postnatal breastfeeding education/ support offered at home 3 days after delivery on breastfeeding duration and knowledge: a randomized trial.” The Journal of Maternal-Fetal & Neonatal Medicine 24(2): 354–361. Hirani, Shela, and Rozina Karmaliani. 2012. “Evidence based workplace interventions to promote breastfeeding practices among Pakistani working mothers.” Women and birth: journal of the Australian College of Midwives. 26. 10.1016/j.wombi.2011.12.005. Imdad, A., M. Y. Yakoob, and Z. A. Bhutta. 2011. “Effect of breastfeeding promotion interventions on breastfeeding rates, with special focus on developing countries.” BMC public health 11 Suppl 3(Suppl 3), S24. Kavle, J., E. LaCroix, H. Dau, and C. Engmann. 2017. “Addressing barriers to exclusive breast- feeding in low- and middle-income countries: A systematic review and programmatic implications.” Public Health Nutrition 20(17): 3120–3134. Kim, S. S., T. Roopnaraine, P. H. Nguyen, K. K. Saha, M. I. Bhuiyan, and P. Menon. 2018. “Factors influencing the uptake of a mass media intervention to improve child feeding in Bangladesh.” Maternal & child nutrition 14(3), e12603. doi:10.1111/mcn.12603. Kumar, Neha, Jody Harris, and Rahul Rawat. 2015. “If They Grow It, Will They Eat and Grow? Evidence from Zambia on Agricultural Diversity and Child Undernutrition.” Journal of Development Studies 51(8): 1060–77. Malapit, Hazel Jean L., Suneetha Kadiyala, Agnes R. Quisumbing, Kenda Cunningham, and Parul Tyagi. 2015. “Women’s Empowerment Mitigates the Negative Effects of Low Production Diversity on Maternal and Child Nutrition in Nepal.” Journal of Development Studies, 51(8): 1097–1123. Mbuya, Mduduzi N. N., Purnima Menon, Jean-Pierre Habicht, Gretel H. Pelto, Marie T. Ruel. 2015. “Maternal Knowledge after Nutrition Behavior Change Communication Is Conditional on Both Health Workers’ Knowledge and Knowledge-Sharing Efficacy in Rural Haiti.” The Journal of Nutrition 143(12): 2022–28. Mgongo, M., T. H. Hussein, B. Stray-Pedersen, S. Vangen, S. E. Msuya, and M. Wandel. 2019. “Facilitators and Barriers to Breastfeeding and Exclusive Breastfeeding in Kilimanjaro Region, Tanzania: A Qualitative Study.” International journal of pediatrics, doi:10.1155/2019/8651010. Muraya, K. W., C. Jones, J. A. Berkley, and S. Molyneux. 2017. “If it’s issues to do with nutrition…I can decide…: gendered decision-making in joining community-based child 184 Appendix C Behavior Change Process Map nutrition interventions within rural coastal Kenya.” Health policy and planning 32(suppl_5), v31–v39. doi:10.1093/heapol/czx032. Nguyen, P. H., E. A. Frongillo, T. Sanghvi, G. Wable, Z. Mahmud, L. M. Tran, … P. Menon. 2018. “Engagement of Husbands in a Maternal Nutrition Program Substantially Contributed to Greater Intake of Micronutrient Supplements and Dietary Diversity during Pregnancy: Results of a Cluster-Randomized Program Evaluation in Bangladesh.” The Journal of nutrition 148(8): 1352–1363. doi:10.1093/jn/nxy090. Nguyen, Phuong H., Sunny S. Kim, Sarah C. Keithly, Nemat Hajeebhoy, Lan M. Tran, and Marie T. Ruel. 2014. “Incorporating Elements of Social Franchising in Government Health Services Improves the Quality of Infant and Young Child Feeding Counselling Services at Common Health Centres in Vietnam.” Health Policy and Planning 29(8): 1008–. Noack, Anna-Lisa, and Nicky R. M. Pouw. 2015. “A Blind Spot in Food and Nutrition Security: Where Culture and Social Change Shape the Local Food Plate.” Agriculture and Human Values 32(2): 169–82. Okuga, M., M. Kemigisa, S. Namutamba, G. Namazzi, and P. Waiswa. 2015. “Engaging community health workers in maternal and newborn care in eastern Uganda.” Global health action 8, 23968. doi:10.3402/gha.v8.23968. Picolo, M., I. Barros, M. Joyeux, A. Gottwalt, E. Possolo, B. Sigauque, and J. A. Kavle. 2019. “Rethinking integrated nutrition-health strategies to address micronutrient deficiencies in children under five in Mozambique.” Maternal & child nutrition, 15 Suppl 1(Suppl 1), e12721. doi:10.1111/mcn.12721. Reerink, I., S. M. Namaste, A. Poonawala, C. Nyhus Dhillon, N. Aburto, D. Chaudhery, … R. Rawat. 2017. “Experiences and lessons learned for delivery of micronutrient powders interventions.” Maternal & child nutrition, 13 Suppl, e12495. doi:10.1111/mcn.12495. Sanghvi, T., A. Jimerson, N. Hajeebhoy, M. Zewale, and G. Huong Nguyen. 2013. Tailoring Communication Strategies to Improve Infant and Young Child Feeding Practices in Different Country Settings. doi: 10.1177/15648265130343S204. Shi, L., J. Zhang, Y. Wang, L. Caulfield, and B. Guyer. 2010. “Effectiveness of an educational intervention on complementary feeding practices and growth in rural China: A cluster randomised controlled trial.” Public Health Nutrition 13(4): 556–565. doi:10.1017/S1368980009991364. Touré, O., S. Coulibaly, A. Arby, F. Maiga, and S. Cairncross. 2013. “Piloting an intervention to improve microbiological food safety in peri-urban Mali.” Int J Hyg Environ Health 216(2): 138–45. Yourkavitch, J. M., J. L. Alvey, D. M. Prosnitz, and J. C. Thomas. 2017. “Engaging men to promote and support exclusive breastfeeding: a descriptive review of 28 projects in 20 low- and middle-income countries from 2003 to 2013.” Journal of health, population, and nutrition 36(1): 43. doi:10.1186/s41043-017-0127-8. 185 Appendix C Behavior Change Process Map WASH Alzua M. L., A. J. Pickering, H. Djebbari, F. C. Lopez, J. C. Cardenas, M. A. Lopera, N. Osbert, and M. Coulibaly. 2015. Final report: Impact evaluation of community-led total sanitation (CLTS) in rural Mali. Buenos Aires: Universidad Nacional de La Plata, Facultad de Ciencias Económicas, Centro de Estudios Distributivos Laborales y Sociales (CEDLAS) Bauza, V., H. Reese, P. Routray, and T. Clasen. 2019. “Child Defecation and Feces Disposal Practices and Determinants among Households after a Combined Household-Level Piped Water and Sanitation Intervention in Rural Odisha, India.” The American journal of tropical medicine and hygiene 100(4): 1013–1021. doi:10.4269/ajtmh.18-0840. Cumming, O., and S. Cairncross. (2016). “Can water, sanitation and hygiene help eliminate stunted growth? Current evidence and policy implications.” Maternal & child nutrition 12 Suppl 1(Suppl 1): 91–105. doi:10.1111/mcn.12258. Davis, J., A. J. Pickering, K. Rogers, S. Mamuya, and A. B. Boehm. 2011. “The effects of informational interventions on household water management, hygiene behaviors, stored drinking water quality, and hand contamination in peri-urban Tanzania.” The American journal of tropical medicine and hygiene 84(2): 184–191. doi:10.4269/ajtmh.2011.10-0126. Islam, M. S., Z. H. Mahmud, P. S. Gope, R. U. Zaman, Z. Hossain, M. S. Islam, D. Mondal, M. A. Y. Sharker, K. Islam, H. Jahan, A. Bhuiya, H. P. Endtz, A. Cravioto, V. Curtis, O. Touré, and S. Cairncross. 2013. “Hygiene intervention reduces contamination of weaning food in Bangladesh.” Trop Med Int Health 18(3): 250–8. Luby, S. P., M. Rahman, B. F. Arnold, L. Unicomb, S. Ashraf, P. J. Winch, J. M. Colford Jr. 2018. “Effects of water quality, sanitation, handwashing, and nutritional interventions on diarrhoea and child growth in rural Bangladesh: a cluster randomised controlled trial.” The Lancet: Global health 6(3): e302–e315. doi:10.1016/S2214-109X(17)30490-4. Newson, R. S., R. Lion, R. J. Crawford, V. Curtis, I. Elmadfa, G. I. Feunekes, C. Hicks, M. van Liere, C. F. Lowe, G. W. Meijer, B. V. Pradeep, K. S. Reddy, M. Sidibe, and R. Uauy. 2013. Behaviour change for better health: nutrition, hygiene and sustainability. BMC Public Health. 13(Suppl. 1):S1. doi:10.1186/1471-2458-13-S1-S1. Pattanayak, S. K., J. C. Yang, K. L. Dickinson, C. Poulos, S. R. Patil, R. K. Mallick, J. L. Blitsteinb, and P. Praharajf. 2009. “Shame or Subsidy Revisited: Social Mobilization for Sanitation in Orissa, India.” Bull World Health Organ 87: 580–87. Unicomb, L., F. Begum, E. Leontsini, M. Rahman, S. Ashraf, A. M. Naser, and P. J. Winch. 2018. “WASH Benefits Bangladesh trial: management structure for achieving high coverage in an efficacy trial.” Trials 19(1): 359. doi:10.1186/s13063-018-2709-1. Venkataramanan, V., J. Crocker, A. Karon, and J. Bartram. 2018. “Community-Led Total Sanitation: A Mixed-Methods Systematic Review of Evidence and Its Quality.” Environmental health perspectives 126(2): 026001. doi:10.1289/EHP1965. 186 Appendix C Behavior Change Process Map Watson, J. A., J. H. J. Ensink, M. Ramos, P. Benelli, E. Holdsworth, R. Dreibelbis, and O. Cumming. 2017. “Does targeting children with hygiene promotion messages work? The effect of handwashing promotion targeted at children, on diarrhoea, soil-transmitted helminth infections and behaviour change, in low- and middle-income countries.” Trop Med Int Health May; 22(5): 526–538. Wodnik, B. K., M. C. Freeman, A. S. Ellis, E. Awino Ogutu, A. Webb Girard, and B. A. Caruso. 2018. Development and Application of Novel Caregiver Hygiene Behavior Measures Relating to Food Preparation, Handwashing, and Play Environments in Rural Kenya. International journal of environmental research and public health 15(9): 1994. doi:10.3390/ijerph15091994. Wood, S., J. Foster, and A. Kols. 2012. “Understanding why women adopt and sustain home water treatment: insights from the insights from the Malawi antenatal care program.” Social Science & Medicine 75(4), August 2012: 634–642. World Health Organization, USAID, and United Nations Children’s Fund (UNICEF). 2015. Improving nutrition outcomes with better water, sanitation and hygiene: practical solutions for policies and programmes. World Health Organization. Access to Health Services Agrawal, P. K., S. Agrawal, S. Ahmed, G. L. Darmstadt, E. K. Williams, H. E. Rosen, and A. H. Baqui. 2012. “Effect of knowledge of community health workers on essential newborn health care: a study from rural India.” Health policy and planning 27(2): 115–126. doi:10.1093/heapol/czr018. CORE Group. 2009. Community approaches to child health in Malawi—Applying the C-IMCI Framework. Washington (DC): CORE Group. Dewey, K. G., and D. R. Mayers. 2011. “Early Child Growth: How Do Nutrition and Infection Interact?” Maternal & Child Nutrition 7: 129–42. Manda-Taylor, L., D. Mwale, T. Phiri, A. Walsh, A. Matthews, R. Brugha, V. Mwapasa, and E. Byrne. 2017. “Changing times? Gender roles and relationships in maternal, newborn and child health in Malawi.” BMC Pregnancy Childbirth 17: 321. Mutanda, J. N., P. Waiswa, and S. Namutamba. 2016. “Community-made mobile videos as a mechanism for maternal, newborn and child health education in rural Uganda; a qualitative evaluation.” African health sciences: 16(4), 923–928. doi:10.4314/ahs.v16i4.6. Okuga, M., P. Waiswa, R. Mandu, J. Wachira, C. Hanson, and F. Manzi. 2017. “Illness recognition and care-seeking for maternal and newborn complications in rural eastern Uganda.” Journal of health, population, and nutrition 36(Suppl 1): 47. doi:10.1186/s41043-017-0125-x. Pell, C., A. Meñaca, F. Were, N. A. Afrah, S. Chatio, L. Manda-Taylor, and R. Pool. 2013. “Factors affecting antenatal care attendance: results from qualitative studies in Ghana, Kenya and Malawi.” PloS one 8(1): e53747. 187 Appendix C Behavior Change Process Map Rahman, A. E., J. Perkins, S. Islam, A. B. Siddique, M. Moinuddin, M. R. Anwar, … D. Hoque. 2018. “Knowledge and involvement of husbands in maternal and newborn health in rural Bangladesh.” BMC pregnancy and childbirth 18(1): 247. doi:10.1186/s12884-018-1882-2. Taleb, F., J. Perkins, N. A. Ali, C. Capello, M. Ali, C. Santarelli, and D. M. Hoque. 2015. “Transforming maternal and newborn health social norms and practices to increase utilization of health services in rural Bangladesh: a qualitative review.” BMC pregnancy and childbirth 15: 75. doi:10.1186/s12884-015-0501-8. Watterson, J., J. Walsh, and I. Madeka. 2015. “Using mHealth to Improve Usage of Antenatal Care, Postnatal Care, and Immunization: A Systematic Review of the Literature.” BioMed Research International vol. 2015, Article ID 153402. Zwisler, G., E. Simpson, and M. Moodley. 2013. “Treatment of diarrhea in young children: results from surveys on the perception and use of oral rehydration solutions, antibiotics, and other therapies in India and Kenya.” Journal of Global Health 3 (1): 010403. 188 Appendix D. Nutrition Portfolio The objective of this portfolio review is to systematically assess the World Bank’s nutrition lending portfolio on its relevance, its multidimensional approaches, and its contributions to nutrition results in country clients. This appendix first describes the lending portfolio identification strategy and then presents findings from a detailed portfolio review using descriptive statistics. Portfolio Identification Strategy IEG anchors the portfolio identification strategy on the conceptual framework. The fundamental dimensions of the conceptual framework, that is, nutrition outcomes for mothers and children, immediate and underlying determinants, and nutrition-specific, nutrition-sensitive, and social norms interventions and institutional strengthening support, are used at different stages of the portfolio identification process. The portfolio identification strategy consists of four stages—search, delimitation, inclusion, and verification—to progressively define the nutrition-relevant portfolio for this evaluation (figure D.1). Search Stage The search stage consists of retrieving from the World Bank’s Business Intelligence repository active and closed financing projects that fall under the evaluation period FY08–19 (irrespective of their approval date) and are financed through an IBRD, IDA, and Recipient-Executed Trust Fund (RETF) agreements. Several project features are extracted for about 10,000 projects, including project identification, titles, countries, regions, lead GPs, lending instruments, approval and revised closing years, Operations Policy and Country Services (OPCS) sector and theme codes, and additional financing flags. Project development objective (PDO) and intermediate outcome indicator data are also retrieved from Implementation Status and Results Reports (ISRs) in the World Bank’s Systems, Applications and Products. In addition, IEG uses country-level data on nutrition outcomes, including stunted growth rates for children under five, from the Joint Child Malnutrition estimates (UNICEF, WHO, and World Bank 2019) to focus the nutrition portfolio on countries with high stunted growth rates. Delimitation Stage The delimitation stage consists of selecting relevant Operations Policy and Country Services sector and theme codes as project filters guided by the conceptual framework.1 The list of projects is further refined by restricting the sample to those operations implemented in high stunted growth countries. High stunted growth countries are defined as those having stunted growth rates on or above 20 percent at any point during 189 Appendix D Nutrition Portfolio the evaluation period according to the Joint Child Malnutrition estimates. Eighty-four out the 88 high stunted growth countries have received World Bank support.2 Projects led by the Energy and Transport GPs are removed because they are not relevant, reducing the list to 4,260 projects in 84 countries. Inclusion Stage As a first step IEG defined a list of key nutrition concepts and associated keywords as input for a machine learning exercise to improve the accuracy of the project identification through text analytics. These key concepts and associated keywords are based on each building block of the conceptual framework: nutrition outcomes for mothers and children, immediate determinants (feeding and caring, nutrient intake and diet diversity, and health of mother and child), underlying determinants (access to nutrient-rich food, maternal and child resources, access to health, and water, sanitation, and hygiene [WASH] services), nutrition-specific, nutrition-sensitive, and social norms interventions and institutional strengthening support, and also a list of donors and partners in nutrition. Figure D.1. Portfolio Identification Strategy Nutrition Portfolio Identification Strategy 1. Search 2. Delimitation 3. Inclusion 4. Verification Evaluation Period Limited to… Relevant Title, Manual Sector and Theme Database C PDO, Components verification with FY08-19 codes (4,260 projects; or Indicators Global Practice • Agriculture 84 countries) portfolio Database A • Education based on key (since FY11) by (10,000 + projects) • Health concepts of Health, Ag, Water • Social Protection Conceptual and Nutrition Business • Public admin. Framework: Global Solutions Intelligence, • Water Group Analysis for Office • Private sector • Nutrition-related (49 projects added) and SAP • Public sector outcomes • IBRD, IDA, and • SD, HD and • Immediate and More Additional gender underlying Financing RETF • Project features • Rural determinants of For which parent Nutrition development nutrition project met portfolio (ID, country, commitments, • Environment • Nutrition-specific inclusion criteria (282 parent and sensitive (59 projects added) projects and sector and theme interventions 133 AF, in 64 codes, other) • ISR data (PDO, • Enabling Included countries) environment ‘Good’ Projects indicators, other) • Agents of change, matches etc. (3 projects added) Project data collection and Database D Database B extraction • Text extraction of (152 countries) Countries with (8 new projects components and JME UNICEF/WHO high levels of added and 12 ‘false policy area from Included Stunting rates for childhood stunting positive’ projects project (291 projects – children under five >=20%) excluded) documents in the parents and AF) (88 countries) World Bank’s Country Case image bank Studies and final consultations with management (17 projects added) Source: Independent Evaluation Group. Note: AF = additional financing; Ag = Agriculture and Food; health = Health, Nutrition and Population; HD = human development; ID = project identification number; ISR = Implementation Status and Results Report; Joint Child Malnutrition estimates = JME; SD = sustainable development. 190 Appendix D Nutrition Portfolio Text extraction and machine learning algorithms are done in collaboration with DECGA.3 The structured corpus of text for PDOs are retrieved for 3,366 projects (79 percent) and indicator text is retrieved for 2,523 projects (59 percent).4 The unstructured corpus of text for project components or prior actions policy areas are extracted from Project Appraisal Document and Program Document (PD) sections (67 percent), respectively, using regular expressions language to identify them,5 combined with project summary text that bundled shorter text segments from different document sections, including components, monitoring and evaluation (M&E), project finance, and lessons learned (33 percent) due to document quality limitations (table D.1). The unstructured text approach allows collecting text for all projects at the cost of possibly reducing content quality, to the extent unstructured text may not always contain information about what a project is doing in terms of nutrition. Table D.1. Distribution of Unstructured Text Variable by Source Text Source IPF DPL PforR Total Components only 2,396 9 71 2,476 56% 0% 2% 58% Project Summaries only 1,303 70 14 1,387 31% 2% 0% 33% Combinations of prior actions, — 397 — 397 program document policy area, and — 9% — 9% project summaries Total 3,699 476 85 4,260 87% 11% 2% 100% Source: Independent Evaluation Group portfolio review and analysis. Note: Table percentages are reported. Independent Evaluation Group manually extracted text for 166 of the 2,476 projects with component text. Project summaries bundled shorter text segments from different document sections including abstracts, components, M&E, project finance, and lessons. Prior actions were retrieved either from program document sections or Operations Policy and Country Services Prior Actions Database. A first semisupervized machine learning exercise consists of applying a natural language processing technique called “word2vec” to the text corpus database to obtain nutrition concepts that are similar to the original list of key concepts and suggest new keywords resulting in an improved list.6 Next, a recommendation engine is used to identify projects that are similar to the list of key concepts and associated keywords. Similarity scores are assigned to each project, for each text corpus (PDO, component, and indicators) and disaggregated by dimensions of the conceptual framework. 7 In addition, semantic tagging is used to tabulate the 4,260 projects with the text corpus classified across dimensions so that each project is tagged for specific building block information contained in their text. Saliency scores are calculated by IEG based on these tabulations, defined as the frequency of key concepts within each building block divided 191 Appendix D Nutrition Portfolio by the total number of key concepts in that building block. Like similarity scores, saliency scores are assigned for each text corpus (PDO, component, and indicators). This exercise involves several iterations triggered by the refinement of the list of key concepts and associated keywords, adjustments of proximity (that is, distance between keywords measured in character spaces), and removal of acronyms (figure D.2). Similarity and saliency scores are recalculated in each iteration and used as part of the inclusion criteria. In a second supervised exercise, a different NLP algorithm is applied to determine the share of nutrition content in a project based on a set of preidentified core nutrition projects. A set of 19 core nutrition projects are manually identified by IEG as the input training set for the exercise, and DECGA implemented a Random Forest Classification and Regression Analysis to predict the share of nutrition content in a project: matching scores. Unlike similarity and saliency scores, each of the 4,260 projects has only one matching score based on the combined corpus text of PDOs, components, and indicators. Inclusion criteria are applied to the sample of 4,260 projects on the basis that projects have relevant title, or PDO, or components, or indicators. Relevant project titles have a reference to nutrition or stunted growth. For determining relevant PDO, components and indicators a combination of different thresholds for saliency and similarity scores are used to ensure that the most relevant projects are included in the portfolio. The inclusion stage filters 291 projects as nutrition-relevant. Verification Stage This stage consists of a systematic manual verification of the included projects against relevant nutrition projects identified by the Nutrition Global Solution Group, and the Agriculture and Food, Health, Nutrition, and Population (HNP), and Water GPs that have been shared during the consultations. Additional financing of parent projects that met the inclusion criteria as well as few projects with good project matching score are also added.8 In addition, the data coding and extraction process, explained in the next section, adds eight projects and eliminates 12 false positives that did not include nutrition content. The final nutrition portfolio includes 282 parent projects and 133 additional financing. 192 Appendix D Nutrition Portfolio Figure D.2. Machine Learning Exercises Components, PDOs, indicators: text extraction Emergent key concepts from corpus of text Key concepts Underlying determinants and terms ML algorithms: - Word2vec - Random forest classification and regression Recommendation measures: - Saliency - Similarity - Matching Source: Independent Evaluation Group. Note: ML = machine learning. Portfolio Coding and Analysis The nutrition portfolio is manually reviewed and coded by the evaluation team. A coding template is designed based on the conceptual framework and administered through Survey Monkey. The coding template facilitates the extraction of project information on nutrition challenges, PDOs, interventions (including behavior change and emergency interventions), project beneficiaries, service delivery mechanisms, and project-level factors of success and failure relevant for a project’s nutrition outcomes. Document type, document section, relevant input text, and classification based on a predefined typology are extracted for all the above except for factors of success and failure that lacked such typology and were later analyzed through topic modeling. Project indicators are coded in Excel according to typologies developed from the conceptual framework. Coders estimate each project’s share of nutrition content, making judgments based on their review and flagging any remaining false-positive projects. Each coder is assigned a subset of projects. Training, a piloting phase, and periodic quality assessment and spot check are conducted to ensure coder reliability. The resulting information from this exercise is manually reviewed and cleaned by the core evaluation team, yielding the final input for portfolio analysis. Portfolio data are analyzed in Excel, Stata, and Tableau software. 193 Appendix D Nutrition Portfolio Findings: World Bank Support for Nutrition The World Bank’s nutrition portfolio consists of 282 projects in 64 countries that account for over $22 billion in estimated nutrition commitments, over half of them in Africa.9 The portfolio is led by the HNP GP with 42 percent of projects, followed by the Agriculture and Food (Agriculture) GP (21 percent), and Social Protection and Jobs (SPJ) GP (20 percent). About half of the projects in the portfolio are closed. Active projects account for almost 62 percent of nutrition commitments ($13.5 billion).10 Africa led the regions with more than half of the project portfolio (53 percent) and nutrition commitments (55 percent), followed by South Asia and Latin America and the Caribbean (figures D3 A and B). Over 90 percent of projects (260) and 87 percent of commitments ($19.0 billion) are investment project financing (4 percent of projects were Program-for-Results financing and 4 percent were development project financing). The World Bank focuses its nutrition support in IDA countries, particularly those with slow stunted growth reduction. About 74 percent (209 projects) of the World Bank’s support for nutrition is directed toward IDA countries and almost half focuses on countries with slow or no stunted growth reduction (132 projects) (figures D.3 C and D). IDA support accounted for $14.4 Billion of nutrition estimated commitments (65 percent). Over time, the portfolio has shifted to include more projects from the Agriculture and SPJ GPs. Over half the projects in the nutrition portfolio were approved in FY14–19. This includes projects led by HNP (37 percent), Agriculture (29 percent), and SPJ (19 percent). This is a shift from FY98–08, with most of the projects led by HNP (55 percent). In contrast, both the Education and Water GPs decreased their participation in nutrition over time (figures D.3 E and F). 194 Appendix D Nutrition Portfolio Figure D.3. Nutrition Financing Portfolio a. Portfolio by Global Practice b. Portfolio by Region Sources: Independent Evaluation Group portfolio review and analysis. Note: Nutrition commitments are based on the manual portfolio review exercise. The total committed amount of each project was multiplied by the estimated share of nutrition content in the project to estimate the share of the project commitments (IDA, IBRD, and RETFs) supporting nutrition. N = 282 projects. The total committed amount is $21,785 million ($13,483 active; $8,302 closed). c. Portfolio by country annual average reduction rate in d. Portfolio by IBRD, IDA, and RETF funding stunted growth 120 100 80 No. of projects 60 40 20 0 Slow Medium Fast Increased Stunting reduction rate Low baseline stunting Medium baseline stunting High baseline stunting Sources: Independent Evaluation Group calculations based on Joint Child Malnutrition Estimates (UNICEF, WHO, and World Bank, March 2019 update) and World Bank historical data on income classification. Note: In panel C shows the distribution of the project portfolio across countries by the average annual reduction rate in stunted growth: fast stunted growth reduction (average of > 1.33 percentage points per year); medium stunted growth reduction (average of 0.94–1.32 percentage points/ year); slow stunted growth reduction (average of 0–0.93 percentage points per year); increased stunted growth (stunted growth levels increased); Low baseline stunted growth (20.0 –34.8), Medium baseline stunted growth (34.9–47.8), High baseline stunted growth (48.0–59.3). Data are from 2008 and 2019 or for the closest years available. Six regional projects are excluded. RETF = Recipient-Executed Trust Funds. In panel C, N = 276. In panel D, N = 282. 195 Appendix D Nutrition Portfolio e. Projects by approval period f. Projects by approval period and Global Practice Sources: Independent Evaluation Group portfolio review and analysis. Note: Figures present data by fiscal years. Other GPs include Macroeconomics, Trade and Investment, Social Sustainability and Inclusion, Urban, Disaster Risk Management, Resilience and Land, and Governance. N = 282 projects. The nutrition portfolio is also supported by RETFs. RETF total commitments account for $5.8 billion across 172 parent projects, with about half of the commitments in HNP ($2.8 billion, of which $1.7 billion are from closed projects).11 The most important RETF in terms of number of projects is the Global Financing Facility (GFF), followed by the Japan Social Development Fund, and the Japan Policy and Human Resources Development Fund (PHRD). In volume, however, individual country Multi-Donor Trust Funds (MDTFs) account for two-thirds of RETF commitments (figures D4A and B). Country MDTFs with the largest contributions include the Ethiopia Promoting Basic Services Program Phase III MDTF (EPBS III) with $615 million, Afghanistan Reconstruction Trust Fund with $480 million, and Bangladesh Health Sector Development Program MDTF with $328 million. The Power of Nutrition accounted for 3.5 percent of the 172 parent projects and $96 Million. 196 Appendix D Nutrition Portfolio Figure D.4. Commitment Amounts to Nutrition of Recipient-Executed Trust Funds a. Commitment amounts by Global Practice 7,000 70 6,024 6,000 60 5,000 50 Commitments (US$M) 4,196 No. of projects 4,000 3,630 40 3,000 30 2,349 1,966 2,000 1,683 20 1,364 1,159 1,035 938 1,000 721 10 527 592 391 164 77 78 47 108 48 148 141 20 165 - 0 Active Closed Active Closed Active Closed Active Closed Active Closed Active Closed Health, Nutrition, Social Protection Agriculture and Water Education Other and Population and Jobs Food Nutrition commitments RETF commitments No. of projects No. of RETF projects b. Commitment amounts by trust fund name Sources: Independent Evaluation Group portfolio review and analysis; data from World Bank Client Connection. Note: Total nutrition commitments includes IDA, IBRD, and RETF amounts committed. Other Global Practices include MTI, Social Sustainability and Inclusion, Urban, Disaster Risk Management, Resilience and Land, and Governance. ASP = Adaptive Social Protection; MDTF = multi-donor trust fund; RETF = recipient-executed trust fund. In panel a, N = 282 projects; in panel b, N =172 projects. RETF commitments are estimated from the amount of parent projects. RETF linked with additional financing projects and World Bank managed are not included. The Power of Nutrition Trust Fund includes projects in Côte d’Ivoire (P161770), Madagascar (P160848), Ethiopia (P123531), Rwanda (P162646 and P164845), and Tanzania (P152736); Burkina Faso is excluded because the associated project (P168823) was approved outside the evaluation period. Ethiopia’s project, P123531, is the parent project of an AF project funded by the Power of Nutrition (P175166). 197 Appendix D Nutrition Portfolio Nutrition Interventions and Multidimensionality The evaluation identifies nutrition interventions in projects across the dimensions of the conceptual framework (nutrition-specific, nutrition-sensitive, and social norms interventions, and institutional strengthening support). A total of 1,792 interventions are identified in the 282 projects in the portfolio. In addition, to understand how projects and the country portfolio have supported a mix of interventions in the conceptual framework, the evaluation defines two measures of multidimensionality, one at the project level and one at the country level (table D.2). Table D.2. Multidimensional Projects and Country Portfolios Measure Definition Project-level Project multidimensionality score: The sum of the number of nutrition-specific and Multidimensionality nutrition-sensitive intervention areas present in a project divided by the eight possible number of intervention areas. The score had a mean value of 0.24 (approximately two dimensions out of the possible eight) with a standard deviation of 0.15, a minimum value of 0 and a maximum of 0.75. Country-level Country portfolio multidimensionality score: The sum of the number of nutrition- Multidimensionality specific and nutrition-sensitive intervention areas present in a country’s portfolio divided by the eight possible number of intervention areas. The score had a mean value of 0.52 (approximately four dimensions out of the possible eight) with a standard deviation of 0.25, a minimum value of 0 and maximum of 1. World Bank nutrition interventions emphasize institutional strengthening, followed by nutrition-sensitive and nutrition-specific interventions. GPs mostly support interventions related with their own sectors as well as institutional strengthening. Of the 1,792 nutrition interventions, almost 40 percent are institutional strengthening with a focus on improving nutrition service delivery. This is followed by nutrition-sensitive interventions which mostly address health and family planning services and agriculture and food systems; and nutrition-specific interventions which focus on supporting diet and breastfeeding, and child disease prevention and treatment (figures D.5A and B). Over time, support for institutional strengthening has persisted, and nutrition-sensitive interventions have increased, while support for nutrition-specific interventions remains relatively constant. The World Bank continues to emphasize support for nutrition service delivery over other types of institutional strengthening, and attention to health and family planning services has reduced in favor of other nutrition-sensitive interventions (notably social safety nets and agriculture and food systems). Recent investments in nutrition-specific interventions have increased support for dietary diversity and breastfeeding relative to child disease prevention and treatment. World Bank support for adolescent health is limited throughout (figure D.5C). The progression of World Bank support for nutrition interventions holds across regions. The increase in nutrition-sensitive interventions during the evaluation period is notable in AFR, SAR 198 Appendix D Nutrition Portfolio and EAP; support for social norms is highest in EAP, but numbers are too small to comment on a trend for social norms interventions (figure D.5D). Figure D.5. Nutrition Interventions a. Share of nutrition interventions, by area b. Interventions per project, by Global Practice c. Share of interventions, by area and approval period d. Share of interventions, by area, region and approval period 199 Appendix D Nutrition Portfolio Source: Independent Evaluation Group portfolio Note: Other GPs include Macroeconomics, Trade and Investment, Social Sustainability and Inclusion, Urban, Disaster Risk Management, Resilience and Land, and Governance. N = 1,792 interventions Project multidimensionality has increased over time. On average, projects have integrated two nutrition intervention areas per project out of the possible eight areas (equivalent to an average project multidimensionality score of 24 percent). Among GPs, Education and SPJ have higher project multidimensionality in that the projects integrate a range of nutrition-specific and nutrition-sensitive interventions. Education often combines ECD interventions with nutrition-specific interventions (diet and breastfeeding, and child disease prevention and treatment) and WASH interventions, while SPJ combines social safety nets support with a range of interventions from health, agriculture and food systems, WASH, and ECD. Overall, the average number of nutrition interventions included in a project increased between 1998–08 and 2014–19 (figure D.6C). The multidimensionality of countries’ portfolios varies, with some countries having a greater investment in multidimensionality. Countries have invested, on average, in four of the following eight areas: diet and breastfeeding, child disease prevention and treatment, adolescent health, health and family planning services, social safety nets, agriculture and food systems, WASH approaches, and ECD. In some countries, such as Guatemala, Indonesia, Pakistan, and Senegal, the country portfolio has a high level of multidimensionality (figure D.7). This suggests a more comprehensive support from the World Bank to address nutrition determinants in these countries. The multidimensionality of the country portfolio is highest in countries with a lower GDP per capita and Human Capital Index, which would be consistent with the need in these countries to address disadvantaged nutrition determinants in the country context. Other country portfolios such as Liberia, Nigeria, and Sierra Leone, have low multidimensionality, raising concerns about the extent that the World Bank is supporting nutrition determinants in the country context. The country portfolios of fragile and conflict-affected situations (FCS) countries on average have a slightly lower multidimensionality than non-FCS countries (figure D.6D). This is likely due to the implementation challenges in FCS contexts. Over time, however, country portfolio 200 Appendix D Nutrition Portfolio multidimensionality has remained relatively constant with projects in the portfolio supporting an average of about five different types of nutrition intervention areas per approval period (1998–08, 2009–13, and 2014–19). The country portfolios vary in the extent that interventions are implemented by projects across GPs, or whether the country has a multidimensional project integrating multiple interventions and sector ministries. About half of the countries have both multidimensional portfolios and medium-to-high support for institutional strengthening. The evaluation examines institutional strengthening across countries, based on the finding that it is important to improve nutrition determinants. Countries such as Indonesia, Panama, Nicaragua, and Rwanda have considerable investments in a mix of interventions in the portfolio, and institutional strengthening. This suggests strong World Bank support for improving the nutrition determinants. Other countries, such as Afghanistan, the Republic of Yemen, and Zambia, have a limited mix of interventions in the portfolio, and low institutional strengthening. This suggest a need for deeper attention to the nutrition support in these countries, and for addressing needed challenges (figure D.7). Figure D.6. Multidimensionality a. Project multidimensionality by number of dimensions and Global Practice 201 Appendix D Nutrition Portfolio b. Project multidimensionality score by Global Practice c. Project multidimensionality score by approval period 202 Appendix D Nutrition Portfolio d. Country multidimensionality score by FCS status e. Country multidimensionality score and Human Capital Index and GDP per capita Source: Independent Evaluation Group portfolio review and analysis. Note: Color scale on the chart increases with the level of country portfolio multidimensionality. 203 Appendix D Nutrition Portfolio Figure D.7. Country Multidimensionality, Institutional Strengthening, and Commitments Source: Independent Evaluation Group portfolio review and analysis. Note: The Pearson correlation between the country portfolio multidimensionality score and the share of institutional strengthening interventions is −0.17 and is statistically significant at the 10 percent level. the Arab Republic of Egypt and Armenia do not have a multidimensionality index because they only have institutional strengthening interventions. Levels of institutional strengthening support were coded for each project in the portfolio. Level of institutional strengthening in a country is defined as the average share of institutional strengthening share of total support coded within the country’s projects. The data across countries is divided into terciles to classify countries as having low, medium, or high institutional support. N = 282 projects. 204 Appendix D Nutrition Portfolio Nutrition Results: World Bank Objectives and Measurement of Nutrition Outcomes and its Determinants The World Bank’s nutrition portfolio overwhelmingly focuses on improving the underlying nutrition determinants and institutional strengthening. Of the 282 projects, 78 (28 percent) are core nutrition projects (have the words “nutri” or “stunt” in their titles or PDOs and have a high share of nutrition content in the top two quintiles); the remainder are mainly sectoral projects that refer to other areas of the conceptual framework and include nutrition interventions in their components. Most projects (81 percent) have PDOs focused on improving underlying determinants of nutrition. The World Bank has sought to improve immediate determinants of nutrition in 18 percent of projects and to improve higher-level outcomes, such as stunted growth and underweight, in 13 percent of projects (figure D.8A). The focus on improving nutrition determinants is consistent with the timeline of projects. Higher-level outcomes are unlikely to be achieved through one project, while a series of projects may contribute to higher-level outcomes. Figure D.8. PDOs and Results Measurement a. Objectives targeted by nutrition projects b. Measurement of nutrition results in projects by c. Measurement of nutrition results in projects by Global Practice approval period Sources: Independent Evaluation Group; and calculations in panel B based on data from project ISRs. Note: In panel a, one project can target multiple objectives. Panel b shows the average measurement of indicators by Global Practice; panel c shows the average measurement of indicators by approval period. The numerator is the number 205 Appendix D Nutrition Portfolio of indicators in a project that measured results (outcome, intermediate outcomes, outputs) for relevant dimensions of the conceptual framework. The denominator is the total number of dimensions relevant to the project’s interventions. N = 282 projects. There were gaps in measuring nutrition results. The evaluation looks at the extent to which indicators in a project results framework measure the project’s intended contribution to nutrition through its interventions. On average projects measure about 60 percent of their nutrition activities, and measurement has slightly improved over time. Among GPs, Water, HNP, and Agriculture are the best at measuring results, while Education and other GPs often did not track their results (figure D.8B). Nutrition- sensitive interventions are the most frequently measured, especially health and family planning services (93 percent), social safety nets (76 percent), and agriculture and food systems (72 percent), while support for ECD and social norms are seldomly measured (table D.3). Table D.3. Average Measurement by Area and Examples of Project Indicators Average Building block of measurement conceptual in indicators framework Example indicators Not applicable Undernutrition Proportion of underweight children (W/A <2SD) < 3 years old; Percentage of children under 2 with weight-for-age <−2Z in project areas; Percentage of children 6–59 months who are stunted. Not applicable Micronutrient Decreased percent of anemic pregnant women; Proportion of deficiencies children participating in the program with anemia. 54%, n=126 Diet and Percentage of children age 6–8 months receiving solid or semisolid breastfeeding food and breastmilk; Percentage of delivered women having received full micronutrient supplementation. 46%, n = 116 Child disease Percentage of children between 12–59 months receiving deworming prevention and tablets; Number of children under 5 with confirmed malaria who treatment received antimalarial treatment. 7%, n=14 Adolescent health Percentage decrease of pregnancy among adolescent women; First time adolescent girls acceptant of modern contraceptives; Number of female adolescents receiving iron–folic acid supplements. 93%, n= 118 Health and family Women 15–49 and children (<5) using the basic package of planning reproductive health and nutrition services (Number, Custom); Children 0–24 months who benefit from a package of nutrition and child stimulation services; Children 0–11 months fully immunized. 76%, n= 74 Social safety nets Households benefiting from the emergency cash transfers program; Beneficiaries of Safety Nets programs—Cash-for-work, food-for- work and public works (number); Children 0–5 benefiting from cash transfers. 72%, n= 115 Agriculture and Households practicing integrated homestead farming; Proportion of food systems targeted hammer mills fortifying maize flour at least once in the past month; Quantity of salt adequately iodized by small producers. 206 Appendix D Nutrition Portfolio 53%, n= 86 WASH approaches Households that have installed appropriate hand washing points; Number of people benefiting from improved access to safe water; Latrines built or renovated for improved sanitation services (number). 6%, n = 32 Early childhood Percentage of children ages 3–5 in targeted villages with an overall development child development score above 0.6; The number of tasks children are able to complete. These tasks cover the domains of gross motor, fine motor, language, cognitive and socio emotional development. 32%, n = 41 Social norms Married women of reproductive age who usually make their own decision regarding health care; Women ages 15 to 49 years having benefited from functional literacy training with a focus on nutrition and stimulation through the project; Percentage of beneficiary households selecting a female household member as cash transfer recipient; Percentage of women 15–49 years using modern contraceptive methods. 41%, n = 102 Policy, financing Multisectoral coordination and accountability plan and results and coordination dashboard ratified; Validation of manual of harmonized package of nutrition services; Industry guidelines for sugar, salt, fat, fortification developed. 52%, n = 190 Improving nutrition Government and nongovernment providers fully trained and service delivery equipped in delivery of basic health and nutrition services in targeted communities; Percentage of community health and nutrition workers achieving satisfactory score on the community service delivery indicator score; States with nutrition intervention mapping system developed and updated at least annually. 53%, n = 125 Stakeholder Model Mothers trained in community nutrition; Percentage of engagement and children 0–23 months of age participating in CBN activities in target ownership areas; Number of community health development committees who submitted quarterly activity reports. Source: Independent Evaluation Group. Note: No measurement score is calculated for nutrition outcome indicators. Nutrition Results: Contributions of the World Bank Most indicators in project results frameworks measure underlying nutrition determinants. IEG identifies and classifies nutrition-related outcomes, intermediate outcomes, and output indicators for the nutrition portfolio. A total of 2,571 nutrition- related indicators have been coded for the 282 projects during data collection and extraction. The evaluation team classified them according to the dimensions of the conceptual framework. The bulk of the indicators (60 percent) measure underlying nutrition determinants, mostly health and family planning services through supply-side health service provision and use/uptake of health services by mothers and children. Health services include basic packages of reproductive health and nutrition, antenatal and postnatal service uptake by mothers, immunization of children, and disease prevention for mothers (such as intermittent preventive treatment doses for preventing malaria), among others. Indicators of institutional strengthening mostly measure the 207 Appendix D Nutrition Portfolio improvement of nutrition service delivery, for example training of community health and nutrition workers on nutrition service delivery at the community level, or nutrition interventions at the subnational level. Indicators of immediate determinants often measure children receiving breastmilk and micronutrient supplementation of mothers and children (figure D.9). Figure D.9. Distribution of Nutrition Indicators in the Portfolio, by Area Source: Independent Evaluation Group portfolio review and analysis. Note: N = 2,571 indicators in 282 projects. IEG evaluates the achievement of outcomes, intermediate outcomes, and outputs in closed projects. Among the 282 nutrition projects, 135 are closed (48 percent), of which 131 have available information on development outcome and intermediate or output level indicators. 12 The overall achievement rate of nutrition indicators is good (70 percent), yet the breakdown by the dimensions of the conceptual framework highlights important differences. For instance, the World Bank is making important contributions to institutional strengthening (79 percent of indicators achieved) and nutrition determinants (68 percent of indicators achieved), while improvements in nutrition outcomes and social norms have been harder to achieve (62 percent and 53 percent, respectively). At a more disaggregated level, the most successful areas are agriculture and food, and improving nutrition service delivery, both with an 81 percent achievement rate. Adolescent health and ECD have high achievement rates, but these rates are based on very small samples. (figures D.10A and B) Project performance is improving over time except for immediate nutrition determinants. Whereas the achievement rates of institutional strengthening and 208 Appendix D Nutrition Portfolio underlying determinants has increased in recent years from 79 to 90 percent and from 66 to 75 percent, respectively, the achievement rate for immediate determinants has dropped from 67 to 62 percent (figure D.10C). Figure D.10. Overall Nutrition Indicator Achievement a. Achievement rates of nutrition indicators by area of conceptual framework 100 90 79 80 70 70 68 68 Achievement rate (percent) 62 60 53 50 40 30 20 10 - Nutrition Immediate Underlying Social norms Institutional Overall (n = outcomes (n = determinants determinants (n = 21) strengthening 1,030) 36) (n = 151) (n = 608) (n = 214) b. Disaggregated achievement rates of nutrition indicators 209 Appendix D Nutrition Portfolio c. Achievement rates of nutrition indicators by approval period Source: Independent Evaluation Group portfolio review and analysis. Note: N = 1,030 indicators in 131 projects The Agriculture GP outperformed most other GPs in overall achievement, the achievement of nutrition determinants and outcomes, and cross-sector achievements, while Sub-Saharan Africa and East Asia and Pacific had the highest achievements across regions. IEG assesses overall achievement rates, and achievement rates for nutrition outcomes and determinants, and cross-sector support (figure D.11). Cross-sector support captures achievements toward immediate and underlying determinants that are supported by interventions in sectors that differed from the project’s leading GP. For example, in a project led by the SPJ, this measure excludes the achievement of social safety net indicators, and rather looks at achievements related to health services, WASH, and other areas of the conceptual framework. The Agriculture GP shows high achievement rates of 79 percent, 78 percent, and 73 percent in these three groups of indicators, respectively (nutrition outcomes, determinants, and cross-sector support); higher than other GPs, except for cross-sector support, where HNP and other GPs have higher performance. Africa and East Asia and Pacific have consistent achievement rates above 70 percent irrespective of the measure and their achievement rates are above other regions; and Latin America and the Caribbean consistently underperformed. Overall achievement rates have increased steadily from 68 to 69 to 75 percent between the 1998– 08, 2009–13, and 2014–19 approval periods, and nutrition determinants and outcomes have had a similar increased achievement. Cross-sector achievement has improved from 64 to 71 percent between 1998–08 and 2009–13, but then dropped to 67 percent in 2014– 19. This achievement rate is important given the increasing emphasis on projects in GPs supporting a range of nutrition interventions that are traditionally implemented by other sectors. 210 Appendix D Nutrition Portfolio Figure D.11. Nutrition Indicator Achievement by Global Practice and Region a. All indicators b. Nutrition determinants and outcomes c. Cross-sector determinants Source: Independent Evaluation Group portfolio review and analysis. Note: I y-axes always show achievement rates. Figures are based on indicators coded for 131 closed projects 211 Appendix D Nutrition Portfolio Explanatory Factors and Lessons of Portfolio Performance: Successes and Failures IEG identifies factors of success and failure behind the achievement of nutrition results. Identification of factors is based on relevant text from ICRs, ICRRs, and PPARs.13 Factors are flagged if they are considered relevant for nutrition-related outcomes in a project and have been classified by direction (success/ failure). A total of 562 factors are identified for 117 of the 135 closed projects based on this definition, where multiple factors could be identified for a single project. Of the 64 countries, 46 are included in this analysis in addition to regional projects, as per the availability of factors data. Table D.4 presents a description of each of the 10 factor topics. Table D.4. Definition of Factor Topics Factor Topic Definition Country Refers to country contextual conditions including the local political economy and context governance, fragility and conflict, and economic and natural disaster shocks. Strengthening Refers to the specific role of World Bank activities in improving institutional capacity in of government government agencies (or lack thereof) in project implementation and achievement of objectives. Country Refers to government commitment and level of institutional capacity for supporting project ownership and activities. For example, government commitment and capacity for coordinating adequate institutional financing; government use of World Bank projects for boosting initiatives and reforms arrangements related with nutrition; the availability and commitment of a skilled workforce for project implementation; and the capacity of line ministries and executing agencies to coordinate action and service delivery. Use of Refers to the extent to which lessons drawn from previous projects, country diagnostics, diagnostics to IEG evaluative documents or other World Bank analytical work were incorporated in project inform project design and implementation. design and implementation Project design Refers to whether projects had a well-defined scope with realistic objectives given contextual factors and to the influence of their stand-alone or programmatic nature on project implementation. For example, whether a project with objectives involving coordination between multiple sectors had realistic expectations about the feasibility of such coordination between implementing ministries and agencies. M&E Refers to the extent to which the strength of M&E frameworks (or lack thereof) affected the implementation of a project and its ability to reach objectives, including having realistic nutrition-related indicators to measure progress as well as baseline and attainable and measurable targets. World Bank Refers to internal World Bank processes affecting project implementation, including systems and adequacy of financing, timeliness of disbursements, procurement, quality of supervision, performance and quality of team composition. 212 Appendix D Nutrition Portfolio Community- Refers to the strength of community engagement and ownership in implementing nutrition based interventions; for example, the strength of community participation through user groups in implementation delivering nutrition-sensitive services like water supply and sanitation; the extent of support for capacity building in communities for selecting and managing local subprojects; the strength of community leadership; and the collaboration between communities and local partners in delivering social services; among others. Innovations Refers to (i) the existence (or lack thereof) of adaptative and innovative changes for and adaptation improving project implementation when needed, including those that make use of a group’s comparative advantage. For example, the flexibility to transfer management and implementation of service delivery to subnational or nongovernment actors when their capacity is greater compared with the central government. And (ii) adding new elements to project design that are expected to improve outcomes, for example behavior change activities to raise awareness about nutrition practices. Stakeholder Refers to the role that collaboration with stakeholders including donor partners (or its engagement absence), had in project implementation and achievement of objectives. and coordination Source: The most frequent success/failure factors are (i) project design, (ii) community-based implementation, (iii) country ownership and institutional arrangements, and (iv) M&E. These topics account for 60 percent of the total factors identified, irrespective of factor direction. Factors (i), (ii), and (iii) are also the most frequent success factors, accounting for about half of all project success factors; and factors (i), (iii), (iv) together with country context, are the most frequent failure factors, accounting for 72 percent of them (figure D.12). 213 Appendix D Nutrition Portfolio Figure D.12. Distribution of Factor Topics and Factor Direction Source: Independent Evaluation Group portfolio review and analysis. Note: N = 562 factors: 336 success factors and 226 failure factors. Project design, country ownership and institutional arrangements, and community- based implementation stand out as success factors in countries with high project achievement. Project design, for example, is a success factor for a project with an achievement rate of 91 percent in Nepal (a medium stunted growth reduction country). The project sought to enhance food and nutritional security of targeted communities in selected locations. Its design has a multisectoral approach that successfully integrated agricultural development, food security, nutrition, and public health as part of fostering Nutritionally Sensitive Agriculture systems, by bringing together several well- coordinated technical ministries and specialized entities to operate under the same project umbrella. Country ownership and institutional arrangements is a success factor for a project in Peru (a fast stunted growth reduction country) with an achievement rate of 73 percent. The project sought to increase demand for nutrition services by strengthening the operational effectiveness of a CCT program (Juntos); and, to improve coverage and quality of supply of basic preventive health and nutrition services in the communities covered under the Program (Articulated Nutrition Program [PAN]), including Juntos. In Peru, the Ministry of Economy and Finance has played an important role in achieving synergies to formalize commitment for better results and greater accountability on nutrition outcomes by including the PAN among the programs to be monitored under the performance-based budgeting pilots. According to the ICR and ICRR, the success of PAN has rested on three pillars: use of result-based budgeting; a 214 Appendix D Nutrition Portfolio unified approach with no one entity having “ownership” of nutrition (a shared priority under shared responsibility); and specialized training for public servants. Finally, community-based implementation is a success factor for a project with a 75 percent achievement rate in Indonesia (a slow stunted growth reduction country). The project sought to empower local communities in low-income, rural subdistricts in project provinces to increase use of health and education services. It has revitalized community health posts (posyandu), which are critical for the achievement of health and nutrition outcomes. Instead of creating new institutions, the project has enabled communities to allocate portions of their block grants to fund interventions that incentivize participation at the posyandu, such as providing nutritional supplements to mothers who attended, funding subsidies for pre‐ and postnatal care, and remunerating posyandu volunteers. As a result, community participation in posyandu activities has improved significantly and has sustained throughout project implementation. Overall, project design, community-based implementation, and country ownership are the most frequent success factors identified in countries with good project performance irrespective of their pace in stunted growth reduction (figure D.13A). Additional countries that have similar success factors and high achievement are Benin, Djibouti, India, Lao PDR, Madagascar, Rwanda, and Senegal. Further, these factors are often absent in countries with projects that have lower achievement. 14 Conversely, weak project design, lack of country ownership, difficult country contexts and low-quality M&E frameworks are frequent failure factors in low performing projects spread across different stunted growth reduction rate countries. Just as good project design and strong government commitment are seen in high achieving countries, weak project design and the absence of country commitment are seen frequently in countries with lower achievement. In addition, problems related with country context and M&E are frequent. For example, M&E issues are seen in a project with only 43 percent achievement rate in Pakistan (a slow stunted growth reduction country). The project sought to increase the coverage, in project areas, of interventions that are known to improve the nutritional status of children under two years of age, of pregnant and of lactating women. The delayed start of the project has had a negative impact on the monitoring and supervision activities. Data are not collected for the indicators intended to measure changes in knowledge about nutrition among communities or health workers, and the midterm review mission has identified issues in data quality with incomplete reporting and inconsistencies between data gathered at facilities and communities. The project also reports a lack of adequate field level staff and support mechanisms, so even routine information could not be collected. M&E issues are also found for other projects in countries like Ethiopia, Guatemala, and Malawi. Finally, country context is often an obstacle for the achievement of nutrition results. This factor 215 Appendix D Nutrition Portfolio captured exogenous conditions like fragility and conflict or natural disasters and how this affected projects. For example, the deteriorated security situation in the Republic of Yemen has resulted in the halting of operations in March 2015; or the extremely dry climate conditions in Timor-Leste due to El Niño phenomenon since late 2015 has resulted in late planting seasons and reduced ability of families to feed their children, negatively affecting the results of the Community-Driven Nutrition Improvement Project. The distribution of failure factor topics is shown in figure D.13B. Figure D.13. Success/failure factors by project performance and stunted growth reduction rate of countries a. Success factors b. Failure factors Source: Independent Evaluation Group portfolio review and analysis. Note: Bubble size increases with the share of total success factors or total failure factors, whichever the case may be. Success factor shares range from 0.3 percent to 7 percent; failure factors range from 0.4 percent to 6 percent. Good nutrition indicator achievement captures countries with average achievement rates above the country median of 0.67; Low nutrition indicator achievement captures countries with average achievement rates on or below 0.67. Fast stunted growth 216 Appendix D Nutrition Portfolio reduction (average of > 1.33 percentage points/ year); medium stunted growth reduction (average of 0.94– 1.32 percentage points/ year); slow stunted growth reduction (average of −1.125–0.93 percentage points/ year—it includes countries with increasing stunted growth rates); figures are based on 332 success factors and 226 failure factors in 111 closed projects, excluding regional projects. Factors of Success and Failure and Multidimensionality Community-based implementation, project design, and country ownership, followed by innovations and adaptations, are the most frequent success factor in countries with multidimensional portfolios (more than six nutrition dimensions in the portfolio) and high achievement of nutrition results (achievement rates above two-thirds) (figure D.14A). Community-based implementation is consistently the most frequent success factor. It accounts for 16 percent of factors in countries with high portfolio multidimensionality, 8 percent in those with medium portfolio multidimensionality (3–6 dimensions) and 30 percent in those with low portfolio multidimensionality (0–2 dimensions). The country case studies found that supporting community-based programs is one way to support a multidimensional package of nutrition interventions to benefit rural communities. In Malawi, for example, community-based program support led by one project is the main nutrition support in the country portfolio. Importantly, failure in countries is often not addressing these success factors (that is, community-based implementation, project design and country ownership, innovations and adaptations). M&E quality is also important with 18 percent, 12 percent and 8 percent of failure factors in countries with low, medium and high portfolio multidimensionality, respectively, and weaker project performance. Lastly, community- based implementation is rarely a failure factor, suggesting that having community-based support to nutrition in a country portfolio often improved performance (figure D.14B). At project level, success and failure factors are similar, but World Bank systems and performance have more importance to support project performance (figures D.14C and D). Factors of Success and Failure and Relevance of Support (Needs and Literature) The three top success factors (that is, community-based implementation, project design, and country ownership) are also consistent for well performing projects in countries with a high percent of interventions matching the country needs, strong alignment of their portfolio with the evidence in the literature, and for core nutrition projects with nutrition addressed in their title or PDO. Community-based implementation remains the most frequent success factor in countries with a high percent of interventions matching their country’s nutrition needs (high: > 50 percent), with 13 percent of factors in that category identified for projects that performed well (figure D.15A); and for projects with interventions aligned to the evidence in the literature (figure D.16A); and for core 217 Appendix D Nutrition Portfolio nutrition projects (23 percent) (figure D.17A). M&E again is the most frequent failure factor (figures D.15B, D.16B, and D.17B). In addition, weaker country ownership and institutional arrangements are a notable negative factor in countries with a low percent of matching of interventions to needs. Figure D.14. Success and Failure Factors and Multidimensionality15 a. Success factors—country portfolio b. Failure factors—country portfolio multidimensionality multidimensionality c. Success factors—project multidimensionality d. Failure factors—project multidimensionality Source: Independent Evaluation Group portfolio review and analysis. Note: High country portfolio multidimensionality [0.750,1.000]; medium country portfolio multidimensionality [0.375, 0.75); low country portfolio multidimensionality [0.000,375). High project multidimensionality (0.500, 1.000]; medium project multidimensionality (0.250, 0.500]; low project multidimensionality [0.000, 0.250]. Good nutrition indicator achievement 218 Appendix D Nutrition Portfolio captures countries with average achievement rates above the country median of 0.67; Low nutrition indicator achievement captures countries with average achievement rates on or below 0.67. N = 562 factors: 336 success factors and 226 failure factors. Figure D.15. Success and Failure Factors and Address of Country Needs a. Success factors b. Failure factors Source: Independent Evaluation Group portfolio review and analysis. Note: A country need is defined as any country-level indicator of a nutrition determinant falling in the bottom 50 percent (see appendix F). The percent match is the extent to which the nutrition portfolio matched interventions to the needs of the countries. High matching percent (>50 percent match); low matching percent (<=50 percent match). Figures are based on 381 factors: 223 success factors and 158 failure factors, coded for 93 projects in 33 countries. Both data on success/failure factors and needs was available for only 33 countries. Figure D.16. Success and Failure Factors by Share of Intervention with Positive Evidence 16 a. Success factors b. Failure factors Source: Independent Evaluation Group portfolio review and analysis. Note: Low share of intervention outcomes with positive evidence in the literature [0.0000, 0.0860], medium [0.0862, 0.1354], high [0.1379, 0.2800]. N=560 factors: 336 success factors and 224 failure factors. Two factors for projects in the Arab Republic of Egypt are excluded because the project only had institutional strengthening support, which is not mapped in the systematic review map. 219 Appendix D Nutrition Portfolio Figure D.17. Success and Failure Factors for Core Nutrition Projects a. Success factors b. Failure factors Source: Independent Evaluation Group portfolio review and analysis. Note: A core nutrition project has the words “nutri” or “stunt” in its title, PDO or both, and has a nutrition content share equal or above the top two quintiles of the distribution (top 40 percent). N=562 factors: 336 success factors and 226 failure factors. Notes 1Selected sector codes are New codes—AH, AL, AI, AB, AT, AF, AK, AZ, EC, EP, ES, ET, EW, EL, EF, EZ, HG, HQ, HF, SA, SG, YA, BC, BH, BG, BZ, WA, WB, WC, WF and WZ; Old codes—AB, AZ, BH, EC, JA and JB, WA, WZ. Selected theme codes were New codes—20, 40, 50, 60, 70, 80, 24, 241, 242, 43, 434, 437, 52, 521, 523, 53, 531, 532, 533, 63, 632, 635, 636, 637, 67, 671, 672, 71, 711, 712, 716, 72, 721, 723, 724, 82, 822, 823, 85, 851; Old codes—26, 54, 57, 58, 59, 60, 62, 63, 68, 69, 70, 71, 73, 75, 77, 78 and 79. 2Countries with no relevant projects include Democratic People’s Republic of Korea, Equatorial Guinea, Malaysia, and Nauru. 3 The Development Data Group of the Development Economics Vice Presidency. 4Missing text does not affect portfolio identification because the final inclusion criteria are based on relevant components, PDOs, or indicator text; and there is component text for all projects. 220 Appendix D Nutrition Portfolio 5Regular expressions is a language embedded inside Python software that allows specifying rules for a set of strings (text) that need to be matched, in this case document section titles. 6 The word2vec algorithm uses a neural network model to learn word associations from a large corpus of text, which in our case is one of the three text variables. Word2vec is one of several NLP word embedding techniques, where words or phrases from a vocabulary are mapped to vectors of real numbers. Once trained, the model can detect synonymous words or suggest additional words for a given sentence. There are two main learning algorithms in word2vec: continuous bag-of-words (CBOW) and continuous Skip-gram. Both algorithms learn the representation of a word that is useful for prediction of other words in the sentence. The CBOW architecture predicts the current word based on the context (surrounding words), and the Skip-gram predicts surrounding words (Mikolov, Corrado, Chen, and Dean 2013). 7Cosine similarity is used to compute similarity scores, defined as the dot product of two nonzero vectors. One of the vector’s elements are the key term frequencies (tf) from a specific conceptual framework building block weighted with their inverse project frequencies (ipf), a measure of how much information a key term provides (that is, it measures if a key term is common or rare across all projects). The weighted output is called the term-frequencies-inverse- project-frequencies or tpif, and tfipf=tf×ipf. The other vector’s elements are the tfipf in one of the ⃗⃗⃗⃗⃗⃗⃗⃗ Vk1 ∙V ⃗⃗⃗⃗⃗⃗⃗⃗ k2 three text variables (such as PDOs). Formally, cosine similarity is defined as C(p|ki) = ⃗⃗⃗⃗⃗⃗⃗⃗ ⃗⃗⃗⃗⃗⃗⃗⃗ , |V k1 ||Vk2 | where p represents the project; k1 is one of the three text variables and k2 one of the 12 building Vk1 and ⃗⃗⃗⃗⃗⃗ blocks from the CF; ⃗⃗⃗⃗⃗⃗ Vk2 represent each of the corresponding vectors with tfipf as their Vk1 ∙ ⃗⃗⃗⃗⃗⃗ elements; ⃗⃗⃗⃗⃗⃗ 96 Vk2 = ∑i=1 Vik1 Vik2 = ∑96 96 i=1(tf × ipf)ik1 (tf × ipf)ik2 = ∑i=1(tfipf)ik1 (tfipf)ik2 where i takes on values between 1 and 96 based on the 96 nutrition key terms defined by IEG; tf is the N frequency of key terms in a project from a specific building block; ipf(t, D) = log , where 1+|{pϵP:tϵp}| N is the total number of projects and |{pϵP: tϵp}| is the number of projects where the term t 2 2 ⃗⃗⃗⃗⃗⃗ appears. Finally, |V ⃗⃗⃗⃗⃗⃗ 96 96 k1 ||Vk2 | = √∑i=1 Vik √∑i=1 Vik . The cosine similarity ranges between −1 and 1 1 1 where higher values imply higher similarity between the vectors. 8After the identification process, engagement with country operations and further consultations highlighted 17 additional projects that are included in the evaluation portfolio. 9 List of countries in nutrition portfolio: AFR: 1. Benin, 2. Burkina Faso, 3. Burundi, 4. Cameroon, 5. Central African Republic, 6. Chad, 7. Comoros, 8. Congo Republic, 9. Congo Democratic Republic, 10. Côte d’Ivoire, 11. Eritrea, 12. Ethiopia, 13. The Gambia, 14. Ghana, 15. Guinea, 16. Guinea-Bissau, 17. Kenya, 18. Lesotho, 19. Liberia, 20. Madagascar, 21. Malawi, 22. Mali, 23. Mauritania, 24. Mozambique, 25. Niger, 26. Nigeria, 27. Rwanda, 28. Senegal, 29. Sierra Leone, 30. South Sudan, 31. Tanzania, 32. Togo, 33. Uganda, 34. Zambia, 35. Zimbabwe; LAC: 36. Belize, 37. Bolivia, 38. Ecuador, 39. El Salvador, 40. Guatemala, 41. Haiti, 42. Honduras, 43. Nicaragua, 44. Panama, 45. Peru; EAP: 46. Cambodia, 47. Indonesia, 48. Lao PDR, 49. Marshall Islands, 50. Philippines, 51. Timor-Leste, 52. Vietnam; SAR: 53. Afghanistan, 54. Bangladesh, 55. Bhutan, 56. India, 57. Nepal, 58. Pakistan; ECA: 59. Armenia, 60. Kyrgyz Republic, 61. Tajikistan; 221 Appendix D Nutrition Portfolio MENA: 62. Djibouti, 63. Egypt, and 64. Yemen. In addition to the 64 countries above, a country category for regional projects is used in the analysis: 65. Regional projects. 10 Portfolio data were retrieved on November 10, 2019. 11The estimated commitment for RETFs include the entire amount of the RETF, while in some cases nutrition interventions may have been limited to one component. 12The status of some closed projects is updated during the portfolio review. Of the four closed projects for which indicators are not available, one is cancelled (P143608), and three are small grants without available indicator data (P121690, P132751, and P150974). 13Factor topics are identified through unsupervised hierarchical clustering machine learning algorithms by Oxford Analytics and Endeavour, who partnered with IEG on a pilot exercise to apply machine learning methods in thematic evaluations. The nutrition evaluation’s portfolio is the focus of the exercise and topic modeling of factors of success and failures is one of the main tasks performed. The team subsequently refined the final list of 10 topics through manual review. 14 Examples in text are from projects P128905, P117310, P132585, P131850, and P145491. 15High country portfolio multidimensionality [0.750,1.000]; medium country portfolio multidimensionality [0.375, 0.75); low country portfolio multidimensionality [0.000,375). High project multidimensionality (0.500, 1.000]; medium project multidimensionality (0.250, 0.500]; low project multidimensionality [0.000, 0.250]. Good nutrition indicator achievement captures countries with average achievement rates above the country median of 0.67; Low nutrition indicator achievement captures countries with average achievement rates on or below 0.67. N = 562 factors: 336 success factors and 226 failure factors. 16Low share of intervention outcomes with positive evidence in the literature [0.0000, 0.0860], medium [0.0862, 0.1354], high [0.1379, 0.2800]. N = 560 factors: 336 success factors and 224 failure factors. Two factors for projects in Egypt are excluded because the project only had institutional strengthening interventions, which were not mapped in the systematic review map analysis. 222 Appendix D Nutrition Portfolio References Mikolov, T., K. Chen, G. Corrado, and J. Dean, J. (2013). “Efficient Estimation of Word Representations in Vector Space.” In International Conference on Learning Representations (ICLR) Workshop Papers. UNICEF, WHO (World Health Organization), and World Bank. 2019. Levels and Trends in Child Malnutrition: Key Findings of the 2019 Edition of the Joint Child Malnutrition Estimates. Geneva: WHO. 223 Appendix E. Behavior Change Portfolio Analysis The portfolio of projects supporting nutrition is analyzed to identify findings related to the World Bank’s engagement in and effectiveness of behavior change interventions. In total, 236 projects are identified with at least one behavior change intervention (83 percent of the portfolio). Coding is conducted to identify behavior change interventions in projects, relevant indicators, and target actors of the interventions. This coding yields 673 behavior change interventions and 822 behavior change indicators, about 38 percent of interventions and indicators in the total portfolio. The main limitation of the analysis is that projects with few behavior change interventions often lack indicators and descriptive details. The interventions are mapped against the dimensions of the conceptual framework (access to food and care, health services, water, sanitation, and hygiene [WASH], social norms, and institutional strengthening)1, and the indicators are mapped to the results chain (engage-learn-apply-sustain). Descriptive data analysis is then conducted in SAS and Tableau. Behavior Change Interventions Behavior change interventions equally cover nutrition-specific (32 percent) and nutrition-sensitive areas (35 percent) (figure E.1), while the intensity of interventions varies across projects. Most of the projects (64 percent) include behavior change interventions in at least two areas of the conceptual framework: 36 percent have interventions in one area; 27 percent in two or three areas; 8 percent in four areas; and 3 percent in all five areas. In terms of the dimensions of the conceptual framework, interventions in areas of food and care (38 percent), institutional strengthening (27 percent), and health services (21 percent) are more widespread in the portfolio, whereas fewer interventions support WASH (9 percent) and social norms (6 percent). Figure E.1. Behavior Change Interventions in the Portfolio Source: Independent Evaluation Group. Note: A project is coded as having an intervention in the behavior change category if it had at least one relevant intervention. Boxes show percent of interventions. N = 673 interventions. Interventions by Global Practice, Region, and Time By GP, projects in Health, Nutrition, and Population (HNP) (92 percent), Social Protection and Jobs (SPJ) (91 percent), and Education (91 percent) equally integrated behavior change interventions, whereas behavior change in projects is less frequent in Water (74 percent), and Agriculture and Food (Agriculture) and other GPs2 (67 percent). 224 Appendix E Behavior Change Portfolio Analysis HNP has the largest number of projects integrating behavior change since the nutrition lending portfolio predominates in the health sector (HNP 53 percent, SPJ 20 percent, Agriculture 13 percent, Education 4 percent, Water 4 percent, and other 6 percent). Across regions, the share of projects with behavior change interventions varies notably. Regions with a high proportion of projects integrating behavior change interventions are East Asia and Pacific (91 percent), South Asia (89 percent), Latin America and the Caribbean (89 percent), and Africa (81 percent), whereas behavior change in projects is less frequent in Middle East and North Africa (71 percent) and Europe and Central Asia (67 percent). The interventions in each GP are multidimensional in that they spread across different behavior change areas, with the largest proportion of interventions in food and care, and institutional strengthening (often training activities) (figure E.2). In HNP and SPJ, health services is also an important behavior change area. In Water, most behavior change interventions focus on WASH. By region, the share of interventions by behavior change area is similar, with most interventions in food and care (about 40 percent), followed by institutional strengthening (about 25 percent), health services (21 percent), WASH (about 9 percent), and social norms (about 5 percent). However, Europe and Central Asia stands out with more interventions in food and care (57 percent), and limited institutional strengthening (7 percent). Moreover, Latin America and the Caribbean has a greater emphasis on social norms (10 percent) than other regions. The share of behavior change interventions in the portfolio has remained similar over time, while the total number of interventions has increased, with more nutrition projects, and there has been an increasing focus on food and care. About 85 percent of projects have at least one behavior change intervention, and behavior change interventions are about 38 percent of total nutrition interventions. There has been a shift to focus on behavior change interventions in the food and care area since 2008, while the share of behavior change interventions in health services and institutional strengthening has decreased (figure E.3). 225 Appendix E Behavior Change Portfolio Analysis Figure E.2. Behavior Change Interventions by Practice and Area Source: Independent Evaluation Group. Note: Other GPs include Macroeconomics, Trade and Investment, Social Sustainability and Inclusion, Urban, Disaster Risk Management, Resilience and Land, and Governance. Figure E.3. Behavior Change Interventions by Project Approval Fiscal Year and Area Source: Independent Evaluation Group. Actors Engaged in Behavior Change Most of the behavior change interventions target mothers and caregivers and communities, while the targeting of actors varies by intervention area, with service providers the main target of institutional strengthening (55 percent). Among the interventions that are implemented, 50 percent target mothers/caregivers, 29 percent target households, 53 percent target communities, 30 percent target service providers, and 7 percent target adolescents. Food and care interventions often target mothers and caregivers (53 percent) or households (55 percent). Social norms interventions mainly focus on adolescent behaviors (28 percent) (figure E.4). 226 Appendix E Behavior Change Portfolio Analysis Figure E.4. Actors Engaged in Behavior Change by Area Source: Independent Evaluation Group. Note: One intervention can engage multiple types of actors. Behavior Change Measurement The World Bank’s operations most often measure behavior change at the level of practice, while engagement, learning, and sustained behavior changes are less often measured (table E.1). In terms of actors, projects most often measure behavior change related improvements among mothers and caregivers (62 percent), given most interventions target these actors. Interventions for other actors are often not measured: service providers, 17 percent; households, 13 percent; and communities, 8 percent. Overall, there is weak measurement of the progression along the results chain to sustain behavior change. Sustained changes are most often measured in areas of food and care and health services and less often in institutional strengthening (figure E.5). The measurement of institutional strengthening is often limited to the engage (51 percent) and learn levels (61 percent), whereas health services often do not measure learning- related changes (4 percent). WASH behavior changes are often less measured. Table E.1 Examples of Indicators by Results Chain Level of Behavior Change Framework Level Example actions Example indicators Engage • Attending community • Number of women attending community events (23%, n=187) awareness event • Number of mothers of targeted children • Joining community participating in monthly information and mobilization session education session in intervention areas (such as for CLTS • Number of community households attending campaign) triggering session • Participating in training • Children 36–59 months with adult member on nurturing care engaged in at least four learning activities in past three days 227 Appendix E Behavior Change Portfolio Analysis Learn • Attending training of • Proportion of women participating in the (14%, n = women’s groups on the program with sufficient knowledge about 114) preparation of nutritious childcare, food consumption, and home hygiene foods • Proportion of parents able to correctly name at • Receiving family least three key actions to improve child nutrition planning counseling • Percent of participating girls and women (10–19) • Receiving livelihood and with improved knowledge on RMNCHN skill training • Proportion of primary health care workers able to correctly answer questions on management of common childhood illnesses Apply • Following breastfeeding • Percent of newborns put to breast within the (41%, n=336) and complementary first hour feeding guidelines • Health facility use rates • Adhering to community • Percent of children fed a diverse diet (at least health worker four food groups) recommendations • Percent of female farmers using improved • Applying livelihood and methods skill training Sustained • Consistently applying • Percent of children ages 0–24 months fed in behavior and breastfeeding and accordance with all three IYCF practices institutional complementary feeding • Diet diversity score of mother change guidelines • Prenatal care (four or more visits) (23%, n= 185) • Consistently applying • Percent of communities free of open defecation livelihood and skill training • Average percentage increase in crop production • Community-level change in applying WASH principles Source: Independent Evaluation Group. Note: RMCHN = Reproductive, Maternal, Newborn, Child Health and Nutrition; PHC = Primary Health Care; IYCF = Infant and Young Child Feeding. 228 Appendix E Behavior Change Portfolio Analysis Figure E.5. Indicators by Behavior Change Area and Results Chain Level Source: Independent Evaluation Group Achievement Rate of Behavior Change Moving from the engage level to sustained behavior changes appears to be challenging, as is achieving behavior changes for mother and caregiver beneficiaries compared with other groups, such as service providers and community. Achievement rates of indicators by level are as follows (figure E.5): engage (84 percent); learn (79 percent); apply (61 percent); and sustained behavior change (69 percent). Further analysis of the achievement rates of indicators by actor shows that interventions supporting service providers and communities have the highest achievement rate (81 percent), followed by households (73 percent), and mothers and caregivers (66 percent). The achievement rates of indicators also vary by GP (Agriculture 75 percent, HNP 72 percent, SPJ 64 percent, Water 60 percent, and other sectors, including Education 80 percent). Figure E.6. Indicator Achievement Rate by Results Chain Level Source: Independent Evaluation Group. Note: total number of indicators in closed projects is n = 347. 229 Appendix E Behavior Change Portfolio Analysis Across intervention areas, the progression from practice to sustained behavior change may be challenging (figure E.6). However, in projects where behavior changes are measured, there is some success at achieving indicators. Health services, food and care and institutional strengthening interventions are most often tracked. The engage and learn level achievements may be easier in all intervention areas. The data on social norms and WASH, however, is based on too few results to make an assessment. Figure E.7. Indicator Achievement by Area and Results Chain Level Source: Independent Evaluation Group. Note: WASH = water, sanitation and hygiene. Food and care: engage (n = 6), learn (n = 14), apply (n = 46), and sustained behavior change (n = 24). Health: engage (n = 34), learn (n = 2), apply (n = 67), and sustained behavior change (n = 54). Institutional strengthening: engage (n = 39), learn (n = 30), apply (n = 7), and sustained behavior change (n = 0). Social norms has 12 indicators at the apply level and other cells are <1. WASH has 4 indicators at apply level and 5 indicators at sustained behavior change, and other cells are <1. Notes 1Food and care include breastfeeding, child feeding, and stimulation; social safety nets; early childhood development; dietary support; and agriculture and food systems. Health services include adolescent health, child disease prevention and treatment, and health and family planning service. 2Other GPs include Macroeconomics, Trade and Investment, Social Sustainability and Inclusion, Urban, Disaster Risk Management, Resilience and Land, and Governance. 230 Appendix F. Heat Map of Country Needs Objective The objective of the heat map analysis is twofold: (i) to understand the situation of countries in the evaluation portfolio in relation to nutrition outcomes, their determinants, and their empirical links based on the conceptual framework; and (ii) to assess the extent to which the World Bank’s interventions, which have been supported through lending, align well with the country needs. The latter aims to respond to evaluation question one on the relevance of World Bank interventions (see appendix A for methodology). Data and Methodology The analysis uses key dimensions of the conceptual framework—nutrition outcomes, determinants (including access to nutrient-rich food, maternal and child caregiving, water, sanitation and hygiene [WASH], and health services), and social norms and behaviors—to guide the data collection and assess country situations.1 The selection of indicators (table F.1) for each building block in all included countries2 aligns with Skoufias et al. (2019), who analyze child undernutrition and nutrition determinants in African countries. Child nutrition outcomes: Outcome indicators relate to the Global Nutrition Targets 2025.3 They are re-expressed to reflect a positive outcome, such as no stunted growth (children under five years not stunted); no anemia (children under five years not anemic); no low birthweight (live births with weight more than 2,500 grams); no underweight (children under five years not underweight); and no wasting (children under five years not wasted). Access to food and care: Access to food and care can influence diet diversity and maternal knowledge and behaviors to care for and feed children. Given the strong link between the access to nutritious food and maternal and child caregiving determinants, these two dimensions are combined. Indicators of food insecurity and mother’s diet diversity are not available and thus are not included in the analysis. Access to WASH: These environmental indicators reflect the sanitary and hygienic conditions in the child’s household and community. Key indicators are access to drinking water, access to sanitation, and the disposal of a child’s stool. Access to health: These indicators capture maternal and child access to, and use of, skilled medical care for illness and preventive care. Skoufias et al. (2018) includes four indicators: use of antenatal care (ANC) services, births assisted by a health care 231 Appendix F Heat Map of Country Needs professional, postnatal checkups, age-appropriate immunization status, and mosquito nets. This analysis did not include mosquito nets to ensure comparability, since they are not relevant across all countries. The analysis did add provision of iron tablets during ANC, vitamin A supplementation, and distance barriers to health facilities to reflect the quality of care and access. Social norms: Social norms can provide an understanding of gender roles, such as those related to decision-making in relation to the care of children, and social and cultural practices that may influence the nutrition status of children and pregnant and lactating women. Key indicators capture aspects of gender roles, sociocultural practices, and women’s empowerment. Table F.1 Indicators Used for Heat Map Analysis Building block of conceptual framework Area assessed Data Child undernutrition outcomes No stunted growth: percentage of children under age 5 Nutrition outcome UNICEF NOT falling below −2 standard deviations (moderate and 2008–18 severe) from the median height-for-age of the reference population No anemia: percentage of children under age 5 whose Nutrition outcome UNICEF hemoglobin level is NOT less than 110 grams per liter at sea 2016 level No low birthweight: percentage of live births that weighed Nutrition outcome UNICEF NO less than 2,500 grams (5.51 pounds) 2015 No wasted: percentage of children NOT wasted (below −2 Nutrition outcome DHS standard deviations of weight-for-height according to the 2000–18 WHO standard) No underweight: percentage of children NOT underweight Nutrition outcome DHS (below −2 standard deviations of weight-for-age according 2000–18 to the WHO standard) Access to food and care Minimum Dietary Diversity of children age 6–23 months Child feeding UNICEF 2008–18 Households consuming iodized salt Access to nutrient-rich food DHS 2008–18 Exclusive breastfeeding of infants age 0–5 months Caring behavior UNICEF 2004–18 Care seeking for diarrhea: children under age 5 with Health-seeking behavior DHS diarrhea for whom advice or treatment was sought from a 2008–18 health facility or provider Financial inclusion: women (age 15+) who reported having Women’s knowledge and access World an account at a bank or another type of financial institution to resources Bank or personally using a mobile money service in the past 12 2014 months 232 Appendix F Heat Map of Country Needs WASH Access to water (at least basic): access to drinking water Access to safe water WHO/U (improved and available) NICEF 2017 No open defecation Access to community-level WHO/U sanitation NICEF 2017 Access to basic handwashing facility with water and soap Access to handwashing facilities WHO/U NICEF 2017 Access to health services DPT3: infants who received the third dose of DTP- Child health UNICEF containing vaccine (12–23 months old) 2018 Skilled birth attendant: deliveries attended by skilled health Safe delivery, newborn care UNICEF personnel 2010–18 Women (age 15–49) who received PNC within two days Access to quality services UNICEF after birth 2010–18 Distance not barrier: women not reporting distance to Access to services DHS health facility as a problem in accessing health care 2001–18 Women (age 15–49) who attended at least four ANC visits Healthy pregnancy UNICEF during pregnancy by any provider 2007–18 Iron tablets during ANC: women with a live birth in the Access to quality services DHS three years preceding the survey who received iron tablets 2001–18 or syrup during ANC Vitamin A supplementation: children age 6–59 months who Access to quality services DHS received vitamin A supplements in the 6 months preceding 2008–18 the survey Social norms Women’s decision power: women who said that they alone Gender roles in the household DHS or jointly have the final say in all three main decisions 2001–18 (health care; making large purchases; visits to family, relatives, and friends) No high-risk births of mothers age <18 Sociocultural practices DHS 2000–18 Women currently using any modern method of Sociocultural practices DHS contraception 2000–18 Literacy: women who are literate Women’s empowerment DHS 2001–18 Source: UNICEF, WHO, and World Bank 2019; USAID 2020; and World Bank 2017. Note: DHS = Demographic and Health Survey; DPT3 = third dose of DPT (diphtheria, pertussis, and tetanus) vaccine; UNESCO = United Nations Educational, Scientific and Cultural Organisation; UNICEF = United Nations Children’s Fund; WHO = World Health Organization. Level of indicators: For each building block, the baseline and current levels of the indicators are categorized for each country at the national level as lowest (i), low (ii), 233 Appendix F Heat Map of Country Needs medium (iii), and high (iv), with high being the best situation of the indicator. The thresholds for the categories are based on the literature to the extent possible,4 or calculated in quartiles based on the distribution of the data in the included countries for the available years. Data for the baseline levels are from 2008, and the current levels are from 2018 or the closest year available. Trend of indicators: Similarly, indicator trends are calculated at the national level over the 10-year period, assuming a constant annual growth rate. The trend is categorized in quartiles, from high decrease (i), modest decrease (ii), modest increase (iii), to high increase (iv). As described above, thresholds are based on the global average from the literature or the average for the data in the evaluation countries. For indicators with only one data point available the trend is not calculated. Composite scores: Using principal component analysis, composite measures are also constructed for each nutrition determinant (that is, access to food and care, WASH, and health services), social norms, and undernutrition outcomes at the baseline level, the current level, and their trends over the 10-year period. Due to data availability the composite scores are calculated for countries with complete data, or no more than 50 percent of the values for the indicators missing. Remaining missing values are replaced with the regional average for the indicator.5 Based on the composite score for each building block, countries are categorized into quartiles at baseline and current levels from lowest <25% (i), low 25–<50% (ii), medium 50–<75% (iii), to high >75% (iv). For trends, the scale is from no increase or decrease (i), low increase (ii), medium increase (iii), to high increase (iv). The most desired situation is to observe a high increase, even if the current situation is categorized as low, compared with other countries. Overall composite of determinants: The composite scores of access to food and care, WASH, health services, and social norms are combined to construct an overall composite that summarizes the situation of nutrition determinants in a country. The overall composite is the unweighted average of the baseline or current levels, or the trend of the four composite scores. The overall composite is calculated for countries with composite scores available for at least two of the determinants, replacing remaining missing values with the regional average, where relevant.6 Using the estimated data mentioned above, the following analyses are performed. • First, a heat map is constructed combining the composite scores for nutrition outcomes and each of the four determinants to summarize country situation at baseline and its trend over the 10-year period (table F.2). 234 Appendix F Heat Map of Country Needs • Second, Pearson correlation analyses are conducted between the levels and trends of nutrition outcomes and their determinants in the evaluation countries to empirically test their links in the conceptual framework (figures F.1 and F.2). • Third, using portfolio review data, nutrition-related interventions supported by lending operations are mapped into the determinants areas to assess whether country needs have been matched by World Bank interventions. A country need is defined as any individual indicator of a determinant falling below their corresponding threshold established by the literature to the extent possible, or falling in the bottom 50 percent of the distribution at the beginning of the evaluation period. The percent match of World Bank interventions with needs is calculated for the portfolio and by country (table F.2). Links between Nutrition Determinants, Outcomes, Country Needs, and the World Bank Portfolio The countries’ conditions in nutrition determinants matter for achieving better nutrition outcomes. Cross-country correlation analysis confirms the conceptual framework links between nutrition determinants and outcomes. Countries that are better off in terms of food and care, access to WASH and health services, and social norms indicators tend to have better nutrition outcomes at the beginning and at the end of the evaluation period. Correlations between the overall composite of determinants and outcomes are strong and positive, ranging from 0.54 to 0.57 for baseline and current levels. The association between overall determinants and specific nutrition outcomes, such as anemia, is particularly strong (ranging from 0.72 to 0.77), albeit data on anemia is only available for current levels (figure F.1). The link between health determinants and outcomes is strongest across all nutrition outcomes. The correlation coefficients of health determinants and overall outcomes is 0.65, followed by social norms (0.50), WASH (0.40), and food and care (0.29) at baseline. Similar patterns emerge at current levels. This result reinforces the importance of having interventions in health synergized with multidimensional interventions across determinants to improve outcomes. Countries with relatively disadvantaged determinants are slowly converging in outcomes, and thus there is potential for them to catch up over time. When comparing initial conditions to current outcomes, once again the association is positive and significant (0.56 for overall composite measures). Yet the magnitude of changes in outcomes is smaller among countries with higher baseline levels of determinants as indicated by the negative correlation (−0.46). This negative association holds for individual determinants related to access to WASH and health services and social 235 Appendix F Heat Map of Country Needs norms. Figure F.1 plots countries’ overall determinants and nutrition outcomes showing their positive association over time (Panel A) and negative association with respect to outcome changes (Panel B). These results are encouraging and suggest the inequality in undernutrition outcomes among countries could decrease over time as determinants improve. However, once outcomes and determinants improve in a country, last mile improvements to benefit vulnerable populations may be slower. The World Bank’s lending support that tackles the determinants to reduce undernutrition largely aligns with country needs. Table F.2 shows baseline levels and trends for overall nutrition outcomes, determinants by area of concern, number of supported interventions by area, and the matching score by country. Overall, about 79 percent of countries in the sample implemented interventions that address nutrition determinants needs with financial support from the World Bank, suggesting that the World Bank supported the right interventions. Matching of country needs varies across determinants. For instance, identified needs related to food and care are addressed by appropriate interventions in 95 percent of the cases, and the World Bank support in access to health services was highly relevant to needs (90 percent). In contrast, needs related to other areas, such as access to WASH and social norms, are often not addressed by interventions (64 percent and 52 percent, respectively). Since the World Bank’s nutrition portfolio has increasingly supported multidimensional interventions at project level beyond their respective technical sectors (see Annex D), these findings should not be interpreted as an assessment of the relevance of the nutrition portfolio for each GP. The World Bank can do better in increasing its alignment of the portfolio interventions with country needs in areas where the association with country nutrition outcomes is the strongest. Heat map analysis shows that at the national level, the alignment of the portfolio is particularly high in access to health care, which shows the strongest association with country nutrition outcomes, respectively. However, alignment falls short in addressing social norms needs given its relatively importance for nutrition outcomes. The World Bank should increase its emphasis on social norms tackling women empowerment and early pregnancy, which currently accounts for only 6 percent of the portfolio across all GPs and sectors. 236 Appendix F Heat Map of Country Needs Table F.2. Comparing Undernutrition Needs and World Bank Interventions in Countries 237 Appendix F Heat Map of Country Needs 238 Appendix F Heat Map of Country Needs Table F.3 Correlation Analysis of Outcomes and Determinants Source: Independent Evaluation Group. Note: The cells present the Pearson correlation coefficient, p-value (p), and the number of observations (N)., which varies due to the differences in the number of missing values of indicators by country. LBW = low birthweight. 239 Appendix F Heat Map of Country Needs Figure F.1. Undernutrition Determinants and Nutrition Outcomes a. Overall Composite of Four Determinants (baseline) and Composite Measure of Outcomes (current) b. Overall Composite of Four Determinants (baseline) and Composite Measure of Outcomes (trend) Source: Independent Evaluation Group. Note: Each scatter plot presents the Pearson correlation coefficient by country. Panels A and B correspond to figure 2; Health in Base level of determinants (2008) and Base level of outcomes (2008), Base level of composite determinants (2008) and Current level of outcomes (2018), and Base level of composite determinants (2008) and Trend of outcomes, respectively. 240 Appendix F Heat Map of Country Needs Notes 1Available indicators for the enabling environment do not fully capture the complexities embedded under the institutional strengthening building block and thus are not included in the analysis. Such indicators include the country’s voice and accountability score, government’s effectiveness of public services score, enabling environment score of Scaling Up Nutrition, and percent of districts with community programs that include infant and young child feeding and counseling. 2Indicator data are gathered for 64 countries with high stunted growth rates; that is above 20 percent either at the beginning or at the end of the evaluation period. 3 See https://www.who.int/nutrition/global-target-2025/en/. 4Global guidance is used to define thresholds to assess indicators as follows: stunted growth (UNICEF 2020); anemia (de Benoist B et al. 2008); low birthweight (UNICEF 2019b); wasted (WHO 2020c); underweight (Abarca-Gómez et al. 2017); MDD (Development Initiatives 2018); iodized salt (Tran et al. 2016); breastfeeding (UNICEF and WHO 2017, Cai et al. 2012); financial inclusion (Clement 2018); access to water (WHO 2019a and 2019b); open defecation (WHO 2019c); DPT3 (WHO 2018; Peck et al. 2018); skilled birth attendance (WHO 2020a); postnatal care (Maternal Health Task Force 2020); antenatal care (WHO 2020b); Iron tablet (Ba et al. 2019); first birth before 18 (UNICEF 2019); modern contraceptive use (United Nations 2019); and literacy (UNESCO Institute for Statistics 2010, 2015). For the indicators of care seeking for diarrhea, handwashing, distance not barrier, and vitamin A supplementation, the threshold is based on the distribution of the data in the evaluation countries. 5The composite scores are calculated for 60 countries for child undernutrition outcomes, 55 countries for access to food and care, 62 countries for access to WASH, 58 countries for access to health, and 50 countries for social norms (total countries = 64). The missing values are replaced with the regional average for 6 percent of outcomes, 14 percent of countries for food and care, 3 percent of countries for WASH, 9 percent of countries for health, and 27 percent of countries for social norms. Due to missing data across countries, the correlation analyses are based on different samples with different sizes, which impose limits to the full comparability of correlation coefficients across determinants and outcomes. 6Countries excluded from the analysis are Bolivia, Central African Republic, Djibouti, Ecuador, Marshall Islands, and South Sudan. 241 Appendix F Heat Map of Country Needs References Abarca-Gómez, L., Z. Abdeen, Z. Hamid, N. Abu-Rmeileh, B. Acosta-Cazares, C. Acuin, R. Adams, W. Aekplakorn, K. Afsana, C. Aguilar-Salinas, and C. Agyemang. 2017. “Worldwide trends in body-mass index, underweight, overweight, and obesity from 1975 to 2016: a pooled analysis of 2416 population-based measurement studies in 128· 9 million children, adolescents, and adults.” The Lancet 390(10113): 2627–2642. Ba, D, P. Ssentongo, K. Kjerulff, M. Na, G. Liu, X. Gao, and P. Du, 2019. “Adherence to iron supplementation in 22 Sub-Saharan African countries and associated factors among pregnant women: a large population-based study.” Current Developments in Nutrition 3(12): nzz120. Cai, X, T. Wardlaw, and D. Brown. 2012. “Global trends in exclusive breastfeeding.” International breastfeeding journal 7(1): 12. Clement, M. 2018. “Gender gap in bank account ownership hasn’t sifted in seven years” Women’s Advancement Daily. https://www.newsdeeply.com/womensadvancement/articles/2018/04/18/gold-barred-in- rural-kenya-women-banned-from-mining-their-own-land. De Benoist, B., M. Cogswell, I. Egli, and E. McLean. 2008. Worldwide prevalence of anaemia 1993– 05; WHO Global Database of anaemia. Development Initiatives. 2018. 2018 Global Nutrition Report: Shining a light to spur action on nutrition. Bristol, UK: Development Initiatives. Chapter 4 Maternal Health Task Force. 2020. Postnatal care https://www.mhtf.org/topics/postnatal-care/ Peck M, M. Gacic-Dobo, M. Diallo, Y. Nedelec, S. Sodha, A. Wallace. 2018. “Global Routine Vaccination Coverage.” MMWR Morb Mortal Wkly Rep 2019(68): 937–942. DOI: http://dx.doi.org/10.15585/mmwr.mm6842a1. Skoufias, E., K. Vinha, and R. Sato. 2019. All Hands on Deck: Reducing Stunting through Multisectoral Efforts in Sub-Saharan Africa. World Bank Publications, 2019. Tran, T, B. Hetzel, and J. Fisher. 2016. “Access to iodized salt in 11 low-and lower-middle-income countries: 2000 and 2010.” Bulletin of the World Health Organization 94(2): 122. United Nations. 2019. Global progress in satisfying the need for family planning. https://www.un.org/en/development/desa/population/publications/pdf/popfacts/PopFacts _2019-3.pdf. UNESCO (The United Nations Educational, Scientific and Cultural Organization) Institute for Statistics (UIS). 2010. Adult and Youth Literacy: Global Trends in Gender Parity. http://uis.unesco.org/sites/default/files/documents/fs3-adult-and-youth-literacy-global- trends-in-gender-parity-2010-en.pdf. UNESCO. 2015. Literacy rate available at http://uis.unesco.org/. 242 Appendix F Heat Map of Country Needs UNICEF (United Nations Children’s Fund), Division of Data Research and Policy. 2019. Global UNICEF Global Databases: https://data.unicef.org/dv_index/ UNICEF. 2019a. Datasets. https://data.unicef.org/resources/resource-type/datasets/. UNICEF. 2019b. Early Childbearing, available at https://data.unicef.org/topic/child- health/adolescent-health/. UNICEF. 2019c. Low birthweight, available at https://data.unicef.org/topic/nutrition/low- birthweight/. UNICEF. 2020. Malnutrition, available at https://data.unicef.org/topic/nutrition/malnutrition/. UNICEF (United Nations Children’s Fund), WHO (World Health Organization). 2017. Tracking progress for breastfeeding policies and programmes: global breastfeeding scorecard 2017. Geneva, Switzerland: World Health Organization. UNICEF (United Nations Children’s Fund), WHO (World Health Organization), and World Bank. 2019. Levels and Trends in Child Malnutrition: Key Findings of the 2019 Edition of the Joint Child Malnutrition Estimates. Geneva: WHO. USAID (United States Agency for International Development). 2020. DHS Program STATcompiler. https://www.statcompiler.com/en/. WHO (World Health Organization). 2018. Diphtheria-tetanus-pertussis (DTP3) immunization coverage, https://www.who.int/gho/immunization/dtp3/en/. WHO. 2019a. Drinking Water. https://www.who.int/news-room/fact-sheets/detail/drinking-water WHO. 2019b. Drinking Water. https://www.who.int/water_sanitation_health/monitoring/water.pdf WHO. 2019c. Sanitation. https://www.who.int/news-room/fact-sheets/detail/sanitation WHO. 2020a. Skilled attendants at birth. https://www.who.int/gho/maternal_health/skilled_care/skilled_birth_attendance_text/en/ WHO. 2020b. Antenatal care. https://www.who.int/gho/maternal_health/reproductive_health/antenatal_care/en/ WHO. 2020c. Global Database on Child Growth and Malnutrition—Description. https://www.who.int/nutgrowthdb/about/introduction/en/index2.html WHO (World Health Organization) and UNICEF (United Nations Children’s Fund). 2019. Joint Monitoring Programme for Water Supply, Sanitation and Hygiene: https://washdata.org/data/household!/. World Bank. 2017. Global Findex database: https://globalfindex.worldbank.org/. World Bank. 2019. Worldwide Governance Indicators: https://info.worldbank.org/governance/wgi/. 243 Appendix G. Case Studies This appendix presents the main findings and evidence that have been collected for the eight case study countries—Ethiopia, Indonesia, Madagascar, Malawi, Mozambique, Nicaragua, Niger, Rwanda—selected for this evaluation. Methodology Selection of cases. The evaluation includes a case-based analysis of the World Bank’s nutrition portfolio in 8 countries, selected from the 65 countries covered by the evaluation. The inclusion criteria for the countries are (i) countries with at least one closed and Independent Evaluation Group (IEG)–evaluated project with a nutrition focus in the title or project development objective; (ii) countries with support for institutional strengthening and behavior change interventions related to nutrition; and (iii) countries with projects in at least three Global Practices (GPs). Other considerations are the availability of impact evaluation (IE) evidence on interventions in the country; whether the country has a low Human Capital Index in the bottom or third quartile compared with other countries; the extent that the country’s experience is already documented; and the coverage of countries in different Regions. Criteria used to vary the selection of countries are the average annual change in stunted growth rates during the evaluation period (slow, medium, and fast, based on the quartiles of the data across countries) and the overall project performance based on the achievement rates of nutrition indicators in the portfolio. These criteria result in a list of 15 eligible countries, which have been discussed with operational counterparts to finalize the country selection. Methods and data collection. The data collection in each country follows a case study protocol organized in relation to the conceptual framework and evaluation questions that looked at the relevance, multidimensionality, and results of World Bank support. The case study covers all active and closed lending projects and knowledge work in the country portfolio that supported nutrition-related interventions during the 10-year evaluation period (fiscal years 08–19). Evidence sources that are triangulated for each country include the following: • A portfolio review of relevant lending projects and analytical work, including a review of Project Appraisal Documents, program documents, Concept Notes, Implementation Completion and Results Reports, and knowledge work. • Semistructured interviews with World Bank staff (task teams, country management, and experts engaged in nutrition support in each country), and stakeholders from government and civil society (national and subnational), and 244 Appendix G Case Studies development partners. Interviewees have been selected based on an analysis of actors involved in projects in the country’s portfolio. • Semistructured with beneficiaries of interventions at the subnational levels, including local leaders and community agents. Snowball sampling has been used to identify and interview beneficiaries from a purposeful sample of Regions and communities that have been supported by projects. • Secondary data on nutrition-related indicators from the heat map analysis of country context and needs (appendix F). • Evidence from existing IEs and IEG evaluations. • World Bank Country Partnership Frameworks and Strategies from the evaluation period. • Each government’s national development plan and or nutrition strategy and plan. Evidence sources are triangulated to assess the contribution of the World Bank to nutrition improvements in each country. A country-specific theory of change is developed to assess how the nutrition interventions in the country program have contributed to the dimensions of the conceptual framework. This includes (i) the assessment of nutrition-related interventions that are supported by World Bank projects in the portfolio against the conceptual framework, including target populations and geographies of interventions, and roles of other partners involved in supporting interventions; (ii) the assessment of the alignment of nutrition-related interventions in the portfolio against country context and needs; (iii) the identification of achievements of the World Bank support against outcomes, intermediate outcomes, and outputs in the conceptual framework; and (iv) mapping how behavior changes are supported by project interventions to contribute to improvements in nutrition determinants. Evidence of behavior changes is assessed using the behavior change process maps that have been developed for the evaluation (appendix C). Case studies draw on the full portfolio of projects in a country, expanding the global nutrition portfolio. The global nutrition portfolio identifies about 90 percent of the projects in the countries with nutrition interventions. Some projects with nutrition interventions integrated in components are missed in the portfolio identification if nutrition activities are not detailed in project documents. These additional projects are identified and reviewed at the country portfolio level and part of the case study assessment. Moreover, the case studies portfolio also considers newly approved or pipeline projects, which are not part of the global portfolio. 245 Appendix G Case Studies Most data collection has been done remotely because of travel restrictions related to coronavirus pandemic. In each country, the IEG team worked with national consultants to facilitate country stakeholder interviews. Integral to the case study data collection in Madagascar and Malawi are work on an IEG Project Performance Assessment Report, field missions, and extensive in country interviews that were conducted before these travel restrictions. Findings Table G.1 summarizes the case study findings for each country. Multidimensionality of the Country Portfolio All countries have had a continuum of support to nutrition interventions during the evaluation period. This includes projects implemented by different GPs across sectors (such as Health, Agriculture, Social Protection, Education, Water, and Macroeconomics) at different time frames. Nutrition interventions are embedded in project components. By design, the case studies are selected from countries that have received World Bank support to nutrition from different sectors. This selection criterion is important for understanding the different types of interventions supported by projects across GPs and sectors. In most countries, interventions remain at an early stage of implementation or require further support to bring their contributions to fruition. The interventions supported by projects in the country portfolios are described below and mapped to the conceptual framework (figures G.1–8). • Community-based programs. In seven of the eight countries (Ethiopia, Indonesia, Madagascar, Malawi, Mozambique, Nicaragua, and Rwanda), community-based programs are a principal component of nutrition support. These ongoing programs provide a unique platform to deliver a package of nutrition-specific and nutrition-sensitive interventions, such as household visits and community discussions. A key aspect is social and behavior change communication (SBCC) messaging. Community-based programs strengthened networks of frontline workers, such as care groups, community health workers (CHWs), and other volunteers, to deliver interventions for malnutrition prevention and treatment, pregnancy care, and other services. These interventions included infant and young child feeding, growth monitoring and promotion, cooking demonstrations, home-based care of childhood illness, parenting education, and referrals to health facilities. Improving the quality and coverage of services and monitoring and evaluation of nutrition interventions have been common challenges across countries. 246 Appendix G Case Studies • Health services. Six of the eight countries (Ethiopia, Madagascar, Mozambique, Nicaragua, Niger, and Rwanda) include nutrition as a core part of support to improve basic health services. This support includes improving policy and health facility services (such as pregnancy care, family planning, nutrition counseling, and iron deficiency supplementation) and the link between facility services and community-level programs to increase service demand. Interventions in maternal dietary and adolescent health services are more recent areas of support. Some countries, such as Madagascar, also have emphasized health and nutrition services in schools. Donor division of labor in some countries, such as Malawi and Rwanda, have limited World Bank investments in health services during the evaluation period. • Agriculture and food approaches. These approaches were a main area of nutrition-sensitive support in two countries (Malawi and Rwanda) and less developed in others (Madagascar). Agriculture interventions often support smallholder farmers to improve their productivity and seasonal access to food crops. Interventions include support to water and land management, livestock and poultry, agriculture inputs (such as drought-resistant seeds of key food crops), appropriate technologies to boost productivity, and safe food storage. A critical area of support in Malawi, Mozambique, Nicaragua, and Rwanda has been biofortification and the promotion of protein-rich crops, such as legumes. Home gardens in Ethiopia, Madagascar, Malawi, Niger, and Rwanda are promoted to improve subsistence access to fruits and vegetables, and livestock and poultry for meat, milk, and eggs. In Niger, however, the scale of the support has been limited. Home gardens have often been supported by agriculture extension workers, nongovernmental organizations, or nutrition programs. In Malawi and Rwanda, agriculture diversification has supported access to a diversity of food crops for sustainable land management, but in most countries, productivity increases focus on a few staple crops. • Early childhood development (ECD). This support includes child stimulation, parenting education, and early childhood development (ECD) programs, which are increasingly being delivered or piloted in communities across all eight countries. • Social safety nets. Safety nets are an important nutrition-sensitive intervention in many of the countries (Ethiopia, Madagascar, Mozambique, Nicaragua, Niger, and Rwanda). These interventions include community block grants or cash transfers and accompanying measures, such as SBCC, incentives to use basic health services, and parenting education. Other interventions include livelihoods 247 Appendix G Case Studies skills building, income generation support, social funds in communities, and voluntary saving and credit support (for example, in Madagascar, Malawi, Niger, and Rwanda), intended to support household food security and resilience to shock. Safety nets are often integrated with health, education, and agriculture interventions. • Water, sanitation, and hygiene approaches. Four countries have had strong investment in interventions to improve access to rural water, sanitation, and hygiene (WASH) development: Ethiopia, Malawi, Nicaragua, and Rwanda. Interventions include piped water, water treatment, WASH in schools and health facilities, and sanitation infrastructure, such as toilets and hand washing facilities. Collaboration with other donors to promote open defecation–free communities through latrines and campaigns is part of this support in some countries. WASH has also been mainstreamed in SBCC activities in community- based programs. In other countries, such as Madagascar, Mozambique, and Niger, WASH support has been weak or limited to SBCC promotion or integrated activities delivered by other sectors, such as social protection or health. • Social norms support. All the case study countries have had some support aimed to empower women’s engagement in agriculture (Madagascar, Rwanda, and recently Niger and Ethiopia) and promote life skills, family planning, and or girl’s education (Ethiopia, Niger, and Rwanda). Across most countries, however, this support is limited or is a more recent development, which requires further expansion or investment. Nicaragua stands out for its consistent support to develop and implement strategies on sexual and reproductive health rights since 2011. Niger likewise stands out for its emphasis on investments in women and girls through education, health, and social protection, and more recently in agriculture. Alignment of Interventions with Country Situations In most countries, the World Bank’s portfolio has aligned with a need to improve nutrition determinants and institutionalize interventions in national and subnational programs. Challenges across countries include the limited timeline during which nutrition interventions have been implemented and the pilot scale of some interventions, with many interventions in an early stage requiring more years of implementation learning to contribute strongly to results. Moreover, countries differ in their approaches and investments to align to multisectoral efforts to coordinate nutrition, and only some countries coordinate support across World Bank projects in different GPs. 248 Appendix G Case Studies Countries have lacked consistent support to address all relevant dimensions of nutrition. Although the World Bank portfolio has included projects across Agriculture, Social Protection, Health, and other areas, interventions are at various stages of development. For example, most countries have lacked consistent and sustained support to improve dietary intake and diet diversity of mother and children. Often, a challenge for community-based programs is long-term support to institutionalize interventions of frontline services. In agriculture, the expansion of home gardens, biofortification, high- nutrient crops, and diverse cropping practices are promising for results, although still at an early stage. Health intervention packages continue to be strengthened across most of the countries to support maternal and child health. WASH requires additional attention, especially in disadvantaged rural areas, and more explicit links to nutrition. Support to improve access to maternal and childcare resources (such as fee exemptions, safety nets for households with children, parenting education, and childcare) is in an early phase, but shows promise. Social norms is an area where the World Bank has had limited project support, except for Nicaragua and Niger. Interventions had weak intraportfolio alignment. Projects in different GPs are implemented in different geographical areas and for different target groups, and they often lack coordination to integrate or converge actions or build on respective achievements to improve nutrition outcomes in the same communities, such as by addressing food, care, WASH, and health. Health interventions often target women and children in rural communities with low nutrition indicators, and coverage of remote areas continues to be a challenge. Safety net and ECD interventions increasingly are coordinated with health interventions by focusing on lower-income households in the same communities, such as in Nicaragua. However, agriculture interventions target farmers and geographies important to the food supply or at risk of natural disasters, and WASH interventions are often in towns, neglecting rural households. Thus, agriculture and WASH approaches often do not benefit the same communities as those supported by social sectors such as Health, Social Protection, and Education. Yet, coordination to integrate or converge the implementation of interventions of different sectors in the same geographical areas are emerging in countries such as Indonesia and Rwanda. A key challenge is responding to tension that nutrition is not the only priority in these areas and may, at times, conflict with other priorities, such agriculture productivity. The challenge remains as to how to integrate nutrition interventions in a way that can maximize the role of each sector and the combined synergies of multiple sectors. In some countries this challenge has been addressed by integrating WASH, agriculture, and other interventions in community-based programs supported by Health or Social Protection projects (such as in Madagascar and Malawi). However, this often does not address the need for supply-side support to improve access to water and sanitation in remote areas. 249 Appendix G Case Studies Countries differ in their alignment with multisectoral nutrition agendas. In all eight countries, nutrition is a priority of the World Bank’s Country Partnership Framework or Country Partnership Strategy. In countries such as Ethiopia, Indonesia, and Rwanda, the World Bank’s strategy also has more recently aligned with the country’s multisectoral nutrition coordination and the global agenda on nutrition. Other countries, such as Niger and Nicaragua, have lacked this alignment. In some countries, such as Madagascar, Malawi, and Mozambique, the World Bank has previously supported multisectoral coordination efforts, but in recent years has mainly focused on the development of community-based nutrition (CBN) and ECD interventions or shifted its focus to the health sector. Remaining challenges are the weak institutional capacity and lack of accountable mechanisms to effectively integrate interventions to improve nutrition across sectors. Country portfolios have had good alignment with national or regional programs to expand interventions. Institutionalizing nutrition interventions supported by investment program financing has been achieved through alignment with national programs. Examples of this alignment include support to a health extension program in Ethiopia, an agriculture village kitchen garden demonstration program in Rwanda, a CHW network in Mozambique, and the development of community care group structures in Malawi. Moreover, in Niger and Nicaragua, the expansion of social norms interventions for women and girls aligns to regional and national programs. Development policy financing and Program-for-Results financing support in Indonesia and Rwanda has been strategically aligned with national programs to help institutionalize interventions, such as performance-based financing, inclusion of nutrition indicators to the performance contracts of subnational leaders, and new interventions in ECD. Policy Dialogue, Knowledge Generation, and Convening Countries differ in how they leverage knowledge activities for learning. All countries offer a blend of knowledge activities that complement the portfolio of projects supporting nutrition. However, some countries (Ethiopia, Indonesia, Madagascar, and Rwanda) have better leveraged a mix of knowledge activities or have had a more consistent flow of learning, such as that from evaluations, knowledge sharing, diagnostics, and other activities, to help strengthen the nutrition results of projects and the country’s program. Evaluations have supported evidence-based learning to design effective interventions. Countries differ in how consistently they used evaluation evidence to adaptively improve nutrition interventions supported by operations. In Madagascar, IEs provided more than a decade of learning to strengthen the rollout of interventions. In some countries (Madagascar and Rwanda) partnership with the Development Economics Vice 250 Appendix G Case Studies Presidency of the World Bank is supporting increased attention to learning. In Ethiopia, Madagascar, Malawi, Nicaragua, and Rwanda, IE evidence has also been important for learning about CBN programs. In Rwanda, IEs have supported efforts to develop performance-based financing in health facilities and at the community level, and agriculture interventions; a recent IE will inform efforts to improve high-impact health services in facilities and nutrition interventions at the community level. In social protection, some countries (Madagascar, Malawi, Nicaragua, Niger, and Rwanda) have used IEs to improve the design of interventions, specifically the links between cash transfers and behavior nudges to improve the demand for child health services, parenting behaviors programs, or child feeding practices. IEs are also supporting learning on ECD programs in some countries (Ethiopia, Madagascar, Niger, and Rwanda). In Madagascar, the focus has been on the effects of adding ECD interventions to CBN programs. These evaluations provide learning to integrate nutrition interventions across social sectors. However, similar attention to designing nutrition- sensitive interventions is lacking in the agriculture and WASH sectors. Even where the World Bank has worked in a small geographical area, evaluations have been important to facilitate the mainstreaming of interventions or experiences leveraged from World Bank support, that is, for more widespread institutionalization in the country’s own program. Diagnostics provide evidence for country programs and policy. In Ethiopia, studies have generated evidence on the cost-effectiveness of nutrition interventions, inequalities in maternal and child health, and the small-area estimation of child undernutrition to improve nutrition policy and programming. In Rwanda, a nutrition situation analysis, the mapping of nutrition-specific and nutrition-sensitive interventions, and a nutrition public expenditure review have supported the government in developing its multisector strategy and identifying needs to improve nutrition financing for multisectoral coordination. In Niger and several other countries (such as Madagascar), diagnostic evidence has been important to inform the development of social protection systems. Leadership and convening activities support policy commitment and action. In Rwanda, a visit from the World Bank president has helped catalyze high-level leadership on and commitment to nutrition. In Indonesia and Rwanda, the engagement of government actors at all levels in nutrition strategy and planning has also been important for leadership building. In Malawi, the Scaling Up Nutrition Forum supported by the World Bank became important for learning across districts implementing nutrition plans, collaboration among partners and stakeholders, and dialogue on policies. Through these convening activities, the World Bank has supported the national multisectoral nutrition plan and policy and the development of its coordination structures. In Mozambique, the World Bank has supported the establishment of national multisectoral coordination 251 Appendix G Case Studies structures at the national and provincial levels that are taking an increasing role to promote the nutrition agenda and strengthen interagency coordination. In Ethiopia, the World Bank has helped convene donors to harmonize nutrition support, such as trust funds, monitoring, and surveillance. Also, in Ethiopia, policy dialogue has supported the development of the national multisectoral nutrition program and coordination structures. In Niger, although multisector coordination of nutrition is lacking, leadership technical assistance has been important to improve service delivery in nutrition. South-South learning has supported new approaches. Engagement in South-South learning across countries, such as in Indonesia, Madagascar, and Rwanda, has facilitated high-level adoption of new approaches. For example, reforms to develop the interoperability of social sector information systems in Rwanda are based on knowledge sharing with Peru. In Madagascar, the World Bank organized a large cross-country learning event to support the development of the social protection policy and program and to demonstrate its feasibility in a low-income, low-capacity environment. Trust funds and partnerships support innovation. Trust funds from Japan have been important in Ethiopia, Malawi, and Niger to pilot interventions for adolescents and nutrition-sensitive agriculture. Financing from the Global Financing Facility is developing health and nutrition services in Ethiopia, Mozambique, and Rwanda, including fiduciary management, information systems, and intervention packages. HarvestPlus has supported nutritional biofortification in Mozambique and Rwanda, and the World Bank has played a role in food fortification strategies and engaging farmers. In Rwanda, the Health Results Innovation Trust Fund has been important in supporting the development of performance-based financing. Also in Rwanda, Global Agriculture and Food Security Program financing has supported the village kitchen program, and the Bill and Melinda Gates Foundation is currently supporting the Mind, Behavior, and Development Unit, evidence-based learning on the national behavior change strategy, which could help in rethinking behavior change interventions for nutrition. In Madagascar, the Knowledge for Change program and the Health Results Innovation Trust Fund have provided critical support for the continuity of IE and other operational learning activities to adaptively improve the CBN program, including for human- centered design learning to improve the effectiveness of interventions. Project Contributions Contribution to Results The World Bank has contributed to improving nutrition determinants in all of the countries. Over the evaluation period, indicators of wasting and underweight have decreased in most of the countries, likely because of investments in growth monitoring 252 Appendix G Case Studies and promotion and treatment of malnutrition. In Malawi and Mozambique, repeat crises likely limited improvements in nutrition indicators. Indicators of stunted growth, low birthweight, and anemia may have decreased, but levels remain high in most of the countries. The pathway to improve outcomes has been through support to nutrition determinants by the World Bank and other donors. The results supported by the World Bank are described and summarized across countries (table G.2). A key challenge has been the weak consistency of support to address needs relevant to low or disadvantaged nutrition determinants in particular country contexts. Moreover, results are yet to be seen in more recent investment areas, such as for parenting behaviors and diet diversity. • Breastfeeding, child feeding, and caregiving. The World Bank contributed to improvements in these areas through support to CBN programs in Ethiopia, Madagascar, Malawi, Mozambique, Nicaragua, and Rwanda. Success factors have included consistent support to improve a package of well-designed interventions over time along with strong support to community volunteers from government at all levels. In Malawi, care groups in communities likely have helped improve the early initiation of breastfeeding, but child feeding did not improve. This is likely due to the duration of the World Bank’s support, which was not adequate to strengthen the care groups. Moreover, the World Bank’s support in Malawi has overlapped with periods of crisis and worsening food insecurity. In Madagascar, community-based programs likely have contributed to the improved quality and quantity of food provided to children under three. In Mozambique, Niger, and Rwanda, evidence of dietary and feeding improvements remain limited, but breastfeeding has increased. Social protection support for families also have supported improvements in the minimum diet diversity of children in Madagascar, but national levels remain low. • Child health and disease. Contributions to child health have been through the expansion of CBN programs, specifically by expanding growth monitoring and promotion, screening, and treatment of malnourished children (for example, in Ethiopia, Madagascar, Niger, and Rwanda). There are also gains in the prevention and treatment of childhood diseases, including diarrhea, deworming, and malaria, supported by health services. • Maternal health. Contributions to maternal health are through support to micronutrient supplementation during pregnancy (for example, in Ethiopia and Madagascar), particularly through the provision of iron–folic acid to pregnant women as part of the minimum package of health services. • Access to health services. Ethiopia, Mozambique, Nicaragua, Niger, and other countries have improved access to health services, such as immunizations, family 253 Appendix G Case Studies planning, institutional delivery, and antenatal and postnatal care, but the quality and coverage of services remains a challenge. In Mozambique, success factors have been mobilizing pregnant women in communities and skill building of health professionals. Similarly, in Nicaragua, success factors have been the use of volunteers (reaching 20,000 families) to identify pregnant women in rural areas and the coordination of interventions across social services (health, social protection, and education). Success factors in Madagascar include fee exemptions and drug vouchers, which made services and medicines available to women and children free of charge. • Access to nutrient-rich food. In some countries (such as Ethiopia) the World Bank contributed through agriculture services that have improved the seasonal availability and access to food. For example, fortified crops are an important contribution of the World Bank in Mozambique, Nicaragua, and Rwanda. A success factor to supporting fortified foods in Mozambique was creating synergies with the health project. In Malawi, the World Bank’s contribution has improved agriculture yields and diversified crops. A success factor has been the use of model villages to promote practices among a cluster of villages. In Rwanda, a success factor has been the extensive and consistent support to farmers groups over many years. However, access to diverse foods and dietary intake remains a key challenge across all the countries, overshadowing the need to increase food availability through productivity of staple food. Safety net interventions and community-based food preservation activities in countries, such as Madagascar, also have supported food security among lower-income households. • Access to WASH. In Ethiopia, Malawi, Mozambique, Nicaragua, and Rwanda, the World Bank has increased access to water and sanitation, such as piped water and latrines. The main success factor has been collaboration with the community to strengthen water management. However, in Madagascar and Niger, the World Bank’s contribution to WASH is modest. In Niger, there have been some improvements in WASH supported through social protection and education projects. The World Bank also has contributed to improved WASH behaviors of households with children through CBN programs implemented by care groups in Malawi. • Maternal and childcare resources. In Madagascar, Malawi, Mozambique, Nicaragua, and Niger, safety nets have contributed to improvements in areas such as income, food consumption, and school enrollment and retention among households (many headed by females). Challenges are the limited coverage of 254 Appendix G Case Studies safety net support and the sustainability of supporting households in graduating and sustaining nutrition and livelihood improvements. Results of ECD interventions are not available in most countries. In Nicaragua, safety net support has helped improve parenting skills, including time that parents spend with children on nutrition. • Social norms. Across countries, projects likely have had some contribution to improve sexual and reproductive health rights and knowledge to delay pregnancy (Ethiopia, Nicaragua, and Niger), gender roles in agriculture (Madagascar, and Rwanda), girls’ enrollment in school (Niger), and family planning usage (Ethiopia, Niger, and Rwanda). Nicaragua’s support to sexual and reproductive health rights likely helped increase contraceptive usage and reduce teen pregnancy and gender-based violence. Agriculture support to engage women has likely been important in improving women’s participation in farming and decisions in households, and resources to care for children, including income, livestock, and other assets. Community-based family planning has likely been important in Rwanda to support contraceptive use. However, across countries, results in social norms are limited. Contributions to Behavior Change Community-based programs have contributed to behavior changes to improve nutrition determinants. The evaluation applied a behavior change model to assess how behavior changes (appendix C) have been supported in countries by frontline workers, community groups, and nongovernmental organizations, among other stakeholders. This model has traced incremental behavior changes along a results chain, leading from initial inputs and outputs all the way to sustained behavior change that could be expected to persist after interventions are completed. Although such processes are rarely linear and require interactions across actors and among different types of outputs and outcomes, the model clarifies how progress must traverse four levels. First, the designated actors (caregivers, health providers, and so on) will gain the awareness and motivation to engage in the change process. In the second level, they will learn what is needed for behavior change and then they will draw on available resources and programs to apply new knowledge and skills in the third level. The final level reflects a sustained change in behavior for improving determinants. Examples of behavior change maps are provided for selected countries (table G.3). The main challenges to mapping of behavior change has been the fragmentation of interventions implemented in communities by different projects and the limited measurement of results related to behavior change. Hence, a limitation of this mapping is the reliance on available evidence from project indicators, studies, and stakeholder 255 Appendix G Case Studies interviews. Evidence suggests that the World Bank has contributed to engaging actors and learning (although not often measured), and in some cases new practices by caregivers, farmers, and health workers, among others, but there is limited evidence that the World Bank has contributed to longer-term sustained changes in the behaviors of actors. In most of the countries, CBN programs are still being strengthened, providing an opportunity to improve evidence and learning regarding behavior change. • Maternal and child caregiving and nutrient-rich food. In countries with CBN intervention packages (including Ethiopia, Madagascar, and Malawi), these programs have supported households with young children in adopting maternal and child caregiving behaviors. For example, caregivers participate in community demonstrations and receive counseling and education through home visits by CHWs and care groups. These activities have improved knowledge to prepare food, maintain home gardens, and practice breastfeeding and complementary feeding of children. ECD programs in communities have supported parents’ adoption of new practices, such as child stimulation. Agriculture extension workers also communicate messages on diversifying crops and home gardens, and farmers increased production of micronutrient-rich crops and the practice of home subsistence gardens. Among countries, Madagascar provides evidence that suggests consistent attention to community programs has supported continued improvements in dietary intake in children in areas where the World Bank worked. • Access to health. Behavior change has been through support to frontline workers to promote and deliver services. For example, in Rwanda, there has been an increase in preventive care visits for children and the use of family planning through community-based visits of CHWs. In Madagascar, there has been improved adherence of frontline workers to guidelines for child growth monitoring and promotion activities. In Ethiopia, health extension workers have delivered health messages, and communities have organized child health days. This has supported the increased use of antenatal care, birth registration, contraceptive services, and vitamin A delivery through routine child health services. In Mozambique, promotion activities by CHWs and community leaders have supported the increased use of pregnancy services and vaccination. In Malawi, care groups have promoted health service use (such as antenatal care and child disease management). The treatment of malnourished children and diarrhea has also increased. Nicaragua and Rwanda provide evidence that suggests sustained improvement in the use of a range of maternal and reproductive health services to contribute to nutrition improvements. 256 Appendix G Case Studies • Access to WASH. In Ethiopia, Madagascar, and Nicaragua, caregivers participate in community conversations and or receive messages on WASH practices delivered by extension workers, community groups, or nongovernmental organizations. In Niger, WASH messages are also delivered to social protection beneficiaries. The World Bank has supported training in WASH communication and demonstration skills. In Ethiopia, the World Bank’s support also has facilitated adoption of improved toilet facilities and water sources among households, and monitoring of water quality among communities who then declare themselves free of open defecation. The involvement of local leaders to mobilize communities to address WASH has also been seen in Malawi and Rwanda. In Rwanda, WASH is being integrated in ECD programs in communities, and this is resulting in the adoption of improved WASH practices among participating households. In addition, there was support to households and communities to adopt improved water treatment and management in both Niger and Rwanda. There is some evidence that the World Bank has supported sustained use of improved drinking water and sanitation in communities in Malawi and Rwanda. • Social norms. Agriculture, health, and livelihood interventions in communities have engaged women and promoted gender roles, but evidence regarding these interventions is limited. 257 Appendix G Case Studies Table G.1. Summary of Findings from the Case Study Countries Policy Dialogue, Knowledge Contributions to Nutrition Multidimensionality Alignment with Country Situation Generation, and Convening Outcomes Ethiopia The World Bank’s portfolio in Ethiopia The World Bank’s support has aligned The convening of donors has harmonized Nutrition outcomes have improved has increasingly transitioned from a with the global nutrition movement. nutrition support. This includes the overall—the magnitude of stunted focus on food security to support a Knowledge from this movement has division of work programming and growth, wasting, underweight, and national multisectoral nutrition program provided a critical push to the country pooled financing modalities, mobilization anemia has decreased, but levels focused on building comprehensive program. of resources (such as trust funds), and remain high. This is supported by nutrition services including nutrition- World Bank country strategies have nutrition monitoring and surveillance. many donors (especially UNICEF) and sensitive and nutrition-specific evolved to sharpen attention toward Diagnostics and evidence on effectiveness likely overall economic growth during interventions. nutrition. The CPS for FY08–11 have informed the nutrition program and the period evaluated. The World Bank, together with other acknowledges the economic costs of policy. For example, the study analytical The World Bank has contributed to donors, has developed CBN services malnutrition. The CPS for FY13–16 work on malnutrition in Ethiopia has improving breastfeeding, child feeding, across the country by strengthening recognizes nutrition determinants and generated evidence on cost-effectiveness and diet diversity. The main success frontline workers. Support includes donor partnerships on nutrition. The of interventions, and analytical work on factor has been the scale-up of CBN nutrition messages; growth CPF for FY18–22 articulates a family planning has provided qualitative services (55.8 million people gained monitoring and management of multisectoral approach to nutrition, research to improve the CBN program. access to the CBN services). malnutrition; agriculture development which is seen in the World Bank’s Analytical work on maternal and child Moreover, the scale-up through the agents to promote nutritious food internal coordination of project teams health inequalities has provided existing government system ensures intake and farming techniques; and and multidimensional design of diagnostic evidence to inform policy. reach and sustainability, and the use support in schools for adolescent projects. Between 2015 and 2018, several studies of community conversation promotes nutrition and WASH facilities. The World Bank’s support has aligned have provided evidence on institutional behavior change. Examples of World Bank support with challenges to improving nutrition capacities to improve WASH programs, The World Bank has contributed to include block grants to local outcomes. However, many challenges policy, and M&E. In 2017, research on the increasing maternal and child health, government, salaries, and training. remain, including to improve care small-area estimation of child and access to health services through Support has also included nutrition- seeking for diarrhea, disparities in undernutrition has provided evidence for the expansion of community-based sensitive interventions in health, social access to health, quality of health care the subnational nutrition program. programs. Specifically, these provide safety nets, women’s empowerment, services, access to safe water and Analytical work on investing in the early increased demand and immunization, agriculture, and WASH. The World sanitation, access to and demand for years in 2018 has informed programming growth monitoring and promotion, Bank has provided support to the dietary diversity, and low maternal to integrate nutrition into non–health screening for malnutrition, development of basic health service knowledge and autonomy. sector operations. micronutrient supplementation, and coverage and policies, such as the The World Bank’s support has aligned Policy dialogue and operations have registration of zinc as an essential with the government’s strategy to expand supported the development of the national 258 Appendix G Case Studies drug. There has also been support to service delivery coverage, with a focus on nutrition program and coordination pregnancy care with an emphasis on CCTs for food-insecure families. The rural areas. Early nutrition support has structures. The challenge remains nutrition. World Bank provides grants for been piloted in a limited number of inadequate leadership and accountability community-driven IGSs and is piloting regions and then expanded nationally, to implement multisectoral actions for support for ECD. It has supported life targeting women and children in the coordination at the local government and The World Bank has contributed to skills and schooling for girls. In access 1,000-day window. More precise community level. improving access to food and care to food, the World Bank has intervention targeting has occurred at Trust funds have supported innovation, through agricultural support services supported irrigation infrastructure and the local government level and in attention to needs, and coordination. Japan and nutrition-sensitive safety net improved inputs for farmers, among agriculture for population groups such trust fund support has piloted support, including increasing access other areas. In WASH, the key support as pastoralists. community-level interventions for to financing through IGS. However, has been to rural water supply and adolescents. Financing from the GFF is food security, inadequate diet, and sanitation development. supporting RMNCH and nutrition food diversity remain challenges. services, including by bringing Newer support in ECD promises to coordinated donor attention and add results in this area. resources to challenges, such as fiduciary The World Bank has contributed to management, information systems, and increasing access to safe water in rural vital registries. areas. Improvements also have occurred in the time and distance required to access water. The main success factor has been collaboration at the community level, reaching more than 6 million people. Indonesia The World Bank has supported CDD The World Bank’s support has aligned Engagement in South-South learning has Nutrition outcomes have improved approaches in social, WASH, and ECD with the country’s challenges. A series of facilitated the high-level adoption of new overall. Between 2018 and 2019, sectors. This support includes water projects (still under approaches. The World Bank supported a across the country, the stunted community block grants to incentivize implementation) have supported the visit of a high-level government growth rate for children under five the use of health care services and for achievement of WASH MDG targets delegation to Peru for the new stunted declined from 30.8 to 27.7 percent, hygiene promotion and construction through programmatic mainstreaming growth strategy to draw on lessons from and wasting decreased from 10.2 to of water and sanitation facilities in the and expansion of a nationwide CDD their experience in stunted growth 7.4 percent. The World Bank has had social and water sectors, respectively. approach. Addressing the low use of reduction. a big role in supporting stunted In ECD, support has been to ECD services by lower-income The engagement of government actors at growth reduction acceleration and integrated ECD services for lower- communities, the World Bank supports all levels in nutrition strategy and planning convergence. The Generasi program investments to improve lower-income has been important for leadership building. is ensuring monthly weight increases children’s overall development and The GFF supports “Stunting Summits” for infants and improved 259 Appendix G Case Studies income communities, teacher training, readiness for further education. In the aiming to secure political commitment underweight and stunted growth of and SBCC and parent counseling. social rural sector, support has target children. The ECD project from district leaders regarding priorities Agriculture support has been marginal included the empowerment of local expanded access to community- and finances catalytic technical assistance during the evaluation period. It includes communities in lower-income rural to the Office of Vice President and based ECD services in rural Indonesia, a pilot supporting credit to women for areas to increase use of health and selected line ministries. which has led to improvements in staple food purchases during the dry education services through a CDD IEs have supported evidence-based learning lower-income children’s social season, food savings accounts for approach. More recently, the World to design and adaptively improve nutrition competence, language, cognitive harvest surpluses, and training on use Bank has supported the strengthening interventions in operations. This includes development, and emotional and provision of food storage media. of delivery systems in social protection, the Generasi program and its impact on maturity. More recently, a series of PforR in SPJ health, and nutrition through a series use of basic health services and on The World Bank has contributed to and health have focused on of PforR instruments. In agriculture, achildren’s nutrition status; the CLTS increasing the use of health services. strengthening the delivery of national trust-funded project tested the cost- project that improved sanitation The IE of the community-based sector programs key to reducing effectiveness of a pilot program of practices and reduced parasitic Generasi program shows that stunted growth (ECD, food assistance, food credits and a storage system in infestations; and the ECD project with community block grants to rural and interpersonal communication). It lower-income communities. However, positive results on children’s cognitive communities are an effective tool to includes support to the adaptation of there is still a need to strengthen development. boost the use of basic health care nutrition-sensitive food assistance aspects of agricultural policy to Technical assistance and diagnostic work services. This includes increases in the program, the consolidation of cast promote vegetable and fruit are strengthening nutrition policies. This number of children under age five transfer delivery systems focused on production through small-scale local includes an assessment of the double treated for moderate or severe acute family planning for reproductive farmers and improve quality of food burden of the malnutrition problem; malnutrition, pregnant women taking health, provision of universal ECD, the availability and incomes among rural analytic and advisory services to iron tables, prenatal and postnatal strengthening of the conditional cash lower-income people. strengthen World Bank support in visits, among others. transfer program delivery system, and The World Bank’s support has aligned monitoring, evaluation, and information The World Bank has contributed to converging village service delivery on with the country’s priorities. This is by system development to enhance design improving access to water and all 1,000‐day households of priority supporting the government’s National and implementation of the PAMSIMAS sanitation. The IE of the large-scale nutrition-specific and nutrition‐ Strategy to Accelerate Stunting Community-Based Drinking Water community sanitation program sensitive interventions. Reduction 2018–22 (StraNas) Supply and Sanitation program; a reports an increase in the rate of coordinated by the Office of the Vice programmatic advisory services and toilet construction and a decrease in President to accelerate stunted growth analytics (IMN-PASA) to support the open defecation, parasitic reduction by addressing the implementation of the multisectoral infestations, and diarrhea prevalence convergence of national, regional, and convergence approach to addressing among young children, likely affected community programs involving 22 malnutrition and child development in by differences in drinking water and ministries in 33 priority nutrition the early years, including a recent PER to hand washing behavior. Between interventions across health, water and assess the level and allocation of stunted 2017 and 2019, the percentage of sanitation, ECD, social protection, and growth –related expenditures. households with children under two 260 Appendix G Case Studies food security for 48 million with access to improved drinking beneficiaries over 514 districts. PforRs water rose from 70 to 72 percent at are strategically aligned with national the national level, and from 65.3 to programs to help institutionalize 69.0 percent in the 100 priority interventions. The Investing in districts. The percentage of Nutrition and Early Years program households with access to improved supports the implementation and sanitation rose from 62.4 to expansion of StraNas to increase 66.6 percent at the national level, and simultaneous use of nutrition from 54.3 percent to 58.0 percent in interventions by 1,000‐day households the 100 priority districts. in priority districts by incentivizing the The World Bank has contributed to government to strengthen maternal and childcare resources. The management capacity and system ECD project has helped increase across sectors and levels and to use enrollment in ECD services by lower- existing resources more effectively. income children, and improve early The World Bank’s country strategy also development scores of children aligned with the country’s multisectoral entering kindergarten or the first nutrition coordination and the global grade of primary school. Community agenda on nutrition. block grants also contributed to Coordination to converge the improved household expenditure implementation of interventions of rates and access to economic and different sectors in the same social services in more than 4,000 geographical areas is emerging. The subdistricts. government has institutionalized a modified version of the project’s village scorecard (from the Generasi program) used by an outreach network of HDWs to improve the convergence of nutrition interventions on priority households. In 2019 the scorecard was rolled out in 160 priority districts with high stunted growth rates. Madagascar The World Bank has transitioned from The World Bank’s support has aligned Evidence learning has supported the Nutrition outcomes have improved supporting a humanitarian response to with the country’s challenges. Its adaptive design and delivery of the overall. The magnitude of stunted 261 Appendix G Case Studies support the comprehensive investments in the social sectors, food community-based programs. IEs growth and anemia has decreased, development of nutrition services, security, women’s empowerment, and (conducted between 1999 and 2016) but levels remain high. Wasting and including nutrition-specific and income generation are all appropriate were pivotal in generating new underweight have more sharply nutrition-sensitive interventions in thefor addressing low social indicators, knowledge and evidence to enhance the declined. The CBN program has been national nutrition plan. This includes high levels of poverty, and inequitable impact of the CBN program. For example, a key pathway to contribute to interventions in health, education, and access to basic information and IE evidence has called for more attention nutrition outcome improvements. social protection, and emergency services. to address chronic malnutrition (in its The World Bank has contributed to interventions in coordination with The World Bank’s recent strategy focuses early years), on which the program had increasing breastfeeding and child partners such as the World Food on nutrition. The CPF for FY17–21 aims had little influence. The Mahay pilot has feeding through consistent support of Programme. to strengthen children’s human accordingly tested new community- the health sector to of the CBN Key support has been to a CBN program development. Delivery of integrated based approaches to reduce chronic program under the leadership of the through a phased approach. The CBM health, nutrition, and social protection malnutrition and improve early child national nutrition coordination unit, program provides a platform to interventions in the regions with the stimulation, and used a human-centered which has improved the quality and integrate nutrition-specific and highest stunted growth rates is design approach to understand quantity of food provided to children nutrition-sensitive interventions (such expected. The previous interim strategy constraints to address to improve under three. Social protection as home gardens, school-based had focused on vulnerability and interventions. Evaluation shows that projects also have supported services, women’s empowerment, resilience, including preservation of intensive counseling works in changing improvements in the minimum diet hand washing and hygiene, and health, education and nutrition, and behaviors and shows the effectiveness of diversity of children, but national counseling) in the same communities. disaster management. lipid-based nutrient supplements in levels remain low. It has strengthened the role of The World Bank’s early support to a young children. This adaptive learning The World Bank has contributed to community groups in nutrition. multisectoral approach aligns with has been supported by trusts funds, child health and health service More recent support includes country priorities. The postcrisis national which were critical in influencing the improvements through the treatment interventions in social protection, health, development plan (2015–19) design of subsequent nutrition of malnourished children. There are and education. Social safety net advocated for a multisectoral approach operations and the country’s CBN gains in the prevention and treatment interventions promote the use of basic to human development. program. of childhood diseases (including services available in localities. Attention to agriculture and WASH, Analytical work has strengthened health diarrhea, parasitic infestations, and Education interventions support specifically rural water and sanitation, has and social protection programs. Together, malaria) supported by the CBN and school-based health and nutrition been consistently weak. Although other a country health status report (2010), a school-based services. The World services and early child stimulation donors have worked on these sectors, health PER, and a report on health service Bank also has invested substantially in through parenting support and there is still a role for the World Bank delivery indicators identified challenges basic health services for mothers and community childcare centers. Health to support WASH and agriculture to improving basic health services. children and their synergies with the interventions support access to activities, which are important for Analytical work on social protection has CBN activities. RMNCH services. contributing to the nutrition agenda. helped promote and develop a viable The World Bank has contributed to Social sector projects have targeted social protection program and improve maternal health improvements through vulnerable geographies and populations. dialogue on effective links between cash the provision of IFA to pregnant 262 Appendix G Case Studies Health projects target women and transfers and behavior “nudges” to fine- women as a part of the minimum children in regions with high poverty tune social safety net interventions. package of health services. and low health and nutrition indicators Pilot learning in projects has generated The World Bank has made some and limited access to services. Social knowledge to expand interventions. For contributions to improving access to safety net interventions, similar to example, health projects piloted and nutrient-rich foods mainly through health, reached mothers in demonstrated the positive impact of fee safety net interventions for food- geographies with high poverty and exemptions, which lifted financial insecure households, the CBN food insecurity but also have constraints and boosted use of critical program’s support of family gardens, considered the complementarity of maternal and child health services. livestock and cooking other interventions to foster synergies Policy dialogue regarding operations has demonstrations, and emergency with available social services. The supported the launch of the multisectoral interventions. There has been limited education project is national in its approach to nutrition. The long-running World Bank support to improve long- scope. Agriculture focused on Community Nutrition II project (P001568) term agriculture productivity and geographies important for the food has been important in establishing and diversity. supply. rendering functional the national The World Bank has contributed nutrition coordination structures and the modestly to WASH. Knowledge on decentralized branches. But subsequent WASH likely has improved since investments in nutrition have not further promotional activities on WASH were developed this coordination capacity. supported by all health and nutrition projects in the portfolio. However, there are no supply-side investments in WASH during the period under review. The World Bank has contributed to improvements in maternal and childcare resources. These include improvements in incomes, consumption, and school enrollment and retention through safety net support to vulnerable households (many headed by females). Malawi The World Bank has supported the The World Bank has aligned with needs The World Bank’s policy dialogue has The magnitude of most nutrition rollout and development of community- related to social protection, agriculture, supported the government to develop its outcomes has improved through based care groups. The nutrition WASH, and social mobilization and nutrition coordination structures and support of multiple donors, particularly 263 Appendix G Case Studies project (P125237) has supported the behavior change; there are gaps in strategies. This includes the national the levels of wasting and development of care groups to deliver support to health and WASH support for nutrition policy and communication underweight. Levels of stunted CBN interventions, such as IYCF rural lower-income people. In social strategy, and the organization of growth and anemia remain high. counseling, WASH promotion, protection and agriculture, multilevel structures to coordinate The World Bank has contributed to cooking demonstrations, home interventions have targeted the lowest- nutrition. The World Bank’s nutrition improvements in breastfeeding and gardens, promotion of care seeking income and most vulnerable project (P125237) supported the National child health, but diet diversity remains a for childhood diseases, and promotion households and areas hit by natural Nutrition Strategic Plan (2007–12) and key challenge. Care groups likely have of health services. The new early years disasters, with an emphasis on rural the National Multi-Sector Nutrition Policy improved early initiation of project (P164771) adds a focus on poverty reduction. The main nutrition (2018–22). Strengthening multisector breastfeeding, but there is no community-based childcare and project (P12523) in the portfolio has coordination and decentralized evidence of improvement in child parenting. Other support has been to supported care groups to reach coordination, including structures of care feeding. The duration of the support emergency food support. households in 50 percent of Malawi’s groups, village development committees, is likely too short to adequately In social protection, the World Bank districts. In WASH, most interventions and nutrition coordinating committees, strengthen the care groups and supported the social CCT program for have been in urban areas, despite the have been key to improving access to overlapped with periods of crisis. In families with children. Elements include need to reach rural areas. The lack of services at the household level. The terms of child health, emergency nutrition promotion and livelihood investment in health services is based challenge has been continuity of this treatment and food support of the skills support through village savings on a desire to limit the number of support to reinforce coordination World Bank and UNICEF likely has and loans. sectors in the country portfolio. structures and financing and planning helped reduce wasting. Care groups In agriculture, support has been to Maternal health especially has lacked that are needed to improve the are also trained to promote growth agriculture extension services to engage support. Although other donors do interoperability of sector programs. monitoring and identify malnourished in nutrition-sensitive activities. This support health, there is a role for the Moreover, there has been limited effort children. includes promotion of drought- World Bank to improve support. to synergize nutrition support with other The World Bank has contributed resistant crops, livestock, poultry and The World Bank strategy has prioritized sectoral investments, such as in health, substantially to improve access to fisheries, crop diversification, nutrition to reduce vulnerabilities. The WASH, and agriculture. nutrient-rich foods. The World Bank improved seeds, legume farming, FY07–10 CAS had a focus on Convening of actors has supported learning funded about 25 percent of Malawi’s fortified crops, integrated homestead vulnerability at the household level to to improve nutrition policy and agriculture budget, supporting farming (including home gardens), HIV/AIDS and malnutrition. The FY13– programming. For example, the World increased crop yields and, in more appropriate technologies, and safe 17 CAS aimed to improve nutrition to Bank has helped develop the country’s recent years, diversified crops for food preparation and storage. enhance human capital and reduce Food and Nutrition Conferences nutrition and climate risk In WASH, there has been support to vulnerabilities. The focus has been one research-policy learning and the annual management. A success factor has water management and supply. This nutrition project, which is now closed SUN Forum for district knowledge been the use of model villages to includes strengthening services at all (P125237). sharing. promote practices among a cluster of levels to manage water and some The coordination of interventions across Knowledge generation has provided some villages. rural infrastructure (piped water, water World Bank projects and sectors to evidence to design the community-based The World Bank has contributed to address nutrition has been lacking. For program. In 2011, a national survey improved access to water and 264 Appendix G Case Studies points, harvesting structures, and example, the planting of home conducted with USAID has provided sanitation. Strengthening water latrines). gardens, small livestock rearing, and evidence to improve IYCF activities of management services, including in nutrition promotion activities are care groups in the CBN program. The communities, has been a success supported by operational projects in World Bank also conducted an IE of the factor to improve local water services. social protection, agriculture, and care group support. Care groups also have improved health sectors without coordinated Analytical work has generated evidence to WASH behaviors of households with implementation or results learning. improve social protection and agriculture children. interventions. For example, studies have The World Bank has contributed to provided evidence to enhance targeting maternal and childcare resources. The of interventions and diversify agriculture IE of CCTs found beneficiary crop support to improve its impact on households have increased food nutrition. consumption by 23 percent. In Trust funds have supported learning on addition, CCTs have helped build the new approaches. For example, the resilience of female-headed adolescent nutrition-sensitive agriculture households to economic shocks. The pilot supported by Japan supports year- challenge of safety nets remains the round production of micronutrient-rich limited coverage. More recent crops at demonstration sites. support on ECD should also contribute results in this area. Mozambique The World Bank has supported nutrition In social sectors, the World Bank’s has Analytical work is generating relevant Limited changes in nutrition outcomes interventions in health facilities and at support increasingly aligned with evidence to enhance nutrition activities. In have occurred in Mozambique during the community level. This is through nutrition needs. Key are the Health education, an evaluation of the ECD the evaluation period. Levels of the progressive rollout of the Service Delivery Project (P099930) and program informed the expansion of stunted growth, wasting, Nutrition Intervention Package the Primary Health Care Strengthening interventions and links between ECD and undernutrition, and LBW remain high, through health facilities and a network Program (P163541). Health support health services. A collaborative study also especially in the northern part of the of CHWs that coordinate care groups focuses on the northern provinces, with informed interventions to delay childbirth country. Diet diversity and and CBN sites. Relevant interventions high rates of stunted growth and among adolescents. In agriculture, an IE micronutrient deficiencies (such as include the promotion of access to nutrition services, of extension networks to support farmers anemia) also remain a significant breastfeeding and IYCF, cooking emphasizing the first 1,000 days. In has provided evidence on the challenge. This is further challenged demonstrations, distribution of social protection, the World Bank has effectiveness of different delivery by repeated crises from natural micronutrient powder and zinc and increasingly targeted women in lower- modalities in boosting farmers’ adoption disasters (frequent droughts and iron tablets, treatment of diarrhea, income households. In education, ECD of sustainable practices, with a focus on floods), which likely have had a growth monitoring and promotion, support responds to needs related to increased productivity by smallholders. In detrimental influence on outcomes promotion of pregnancy care, child caregiving. 2014, an evaluation supported by and limited observed improvement. 265 Appendix G Case Studies immunization, malaria prevention, and In WASH and agriculture, interventions HarvestPlus has provided evidence on The World Bank likely has contributed community-based family planning. and needs are less aligned. In WASH, the nutritional biofortification. The World to improvements in breastfeeding and The World Bank’s support to WASH is World Bank has addressed challenges Bank has played a role in designing the child feeding. This is through limited. It includes water points, of water supply and sanitation, but National Food Fortification Strategy engagement with other donors to desalination, rehabilitation of drinking most of the support is in towns. 2016–21. Moreover, the World Bank has strengthen the rollout of community- water systems, and promotion of open Moreover, explicit links to improving recently conducted a study on nutrition- level services, which have likely defecation–free campaigns and latrine nutrition are lacking in the sector. In smart agriculture (including a country helped improve outcomes. However, construction. agriculture, the World Bank has a few profile for Mozambique [June 2020]). The results will need to be assessed, In agriculture, there has been limited examples of successful support, such as recommendations could support the including evidence of dietary and support to the productivity of crops, for crop fortification, but the support agricultural portfolio from FY21 onward. feeding improvement. food fortification, and food distribution has focused on emergency food In social protection, a social assistance The World Bank has contributed to in emergencies. The World Bank has distribution, with a need for more PER (2012) identified options to improve improving in health service use by supported fortification of staple crops emphasis on sustained access to food social programs. In WASH, a recent mobilizing pregnant women in and providing improved and drought- and food diversification. diagnostic has been conducted to communities, supporting resistant seeds for smallholder There has been limited coordination develop policy options for improvement. immunization, and building the skills farmers. The World Bank also financed across World Bank projects to implement Technical support in the health sector has of health professionals. the distribution of food to populations nutrition support. Mainly though the strengthened nutrition services. Recent Improvements are seen in affected by droughts. health sector, the World Bank has support is to the nutrition department in institutional delivery and In social protection and education, the supported the government’s nutrition the health sector, in partnership with the immunization, but the quality of World Bank has supported social CCTs policies and strategies, such as the GFF to develop nutrition services and services remains a challenge. It also and childcare. The ECD program has Multisectoral Action Plan for Reduction primary care RMNCH services, including has contributed to maternal thus far been rolled out to reach of Chronic Undernutrition 2011–. strengthening CHWs. This is core to the increased IFA supplementation and children between ages three and five Agriculture support is aligned with national program to expand health deworming. in 350 rural communities. It is government production and services, with increasing support to The World Bank has made some intended to influence social norms emergency strategies. addressing the nutrition agenda and contribution to access to food. In and the behaviors of parents and The FY17–21 CPF has a focus on ECD and challenges. agriculture, this is through improved children in five provinces. nutrition to build human capital and on Policy dialogue has supported nutrition inputs (for fortified crops), improved addressing nutrition in value chains for coordination. Early support had focused technologies to enhance smallholder food. However, the recent national beyond health, facilitating the productivity, and food assistance to development plan (2020–24) does not establishment of multisectoral vulnerable groups during make explicit reference to prioritizing coordination at the national and emergencies. the nutrition agenda. The FY12–15 CPS subnational levels. These coordination The World Bank has made some had aimed to mainstream nutrition in structures are taking an increasing role to contributions to access to drinking the portfolio to address challenges of promote the nutrition agenda and water and reduced open defecation. strengthen interagency coordination. This is through desalination, 266 Appendix G Case Studies chronic malnutrition, focusing on the Their challenge is to strengthen the expanding piped water among the CBN program and health. institutional capacities and effectiveness lowest-income households, and of these structures at both national and sanitation promotion. The challenge subnational levels. remains the coverage of rural areas. The World Bank is contributing to maternal and childcare resources. In social protection, the World Bank has coordinated with other donors to increase the coverage of social safety nets to support households at risk of food shocks, and increase female beneficiaries of safety nets. The newer ECD program includes support for parenting education to contribute to the knowledge and feeding practices of caregivers. Nicaragua The World Bank portfolio has included a Diet diversity is first identified as a need IEs have helped improve social programs. Nutrition outcomes have improved mix of nutrition-specific and nutrition- in the FY13–17 CPS, in response to the For example, the Women’s Power, overall—the magnitude of stunted sensitive interventions. In social sectors government’s request for support from Conditional Cash Transfers, and growth, wasting, and LBW decreased, (health, education, social protection), GAFSP. In the FY18–22 CPF, nutrition is Schooling in Nicaragua (2008) have but levels of stunted growth remain these interventions have been addressed by improving health and provided evidence to improve measures above the regional average. coordinated by local government and ECD and providing water supply and for women and children. Similarly, a 2009 The World Bank has contributed to through the community-based sanitation in rural communities (that is, IE of a Conditional Cash Transfer Pilot in improving breastfeeding and child program, but there has been limited Dry Corridor and Caribbean regions). Nicaragua has provided evidence for feeding. Success factors have been coordination of nutrition at the The World Bank support in health, social CCTs on child health and education. consistent support to a package of national level. protection, water, and agriculture has Analytical work has prioritized community services and community volunteers The World Bank has supported an aligned with nutrition needs. In health investments. In Social Protection in 2008 a with strong support from government integrated package of nutrition-specific and social protection sectors, the public expenditure review recommended at all levels. Improvements in diet and nutrition-sensitive interventions for support has been to mothers and the extension of the package of basic diversity are limited since health, safety net support, and education children in geographies that have the health services to communities. It also interventions started in FY15 and for lower-income families with children. lowest health indicators and to emphasized ECD and nutrition programs concentrated in two regions. There is This is through the Family and indigenous and lower-income for lower-income children. These limited evidence on home-based care Community‐based Social Welfare populations. In the water sector, the priorities are reflected in the government of diseases since these interventions Model (MAIFC), and the Family and focus has been on access to WASH in and World Bank strategies. Analytical are still active. 267 Appendix G Case Studies Community Health Model (MOSAFC). rural areas. In agriculture, key support work on agriculture performance and The World Bank has made strong This support has helped local has been provided in coastal areas challenges has prioritized support to the contributions to improving access to authorities to coordinate a package of based on the food supply, but there is Dry Corridor with high nutrition needs health. There are increases in services, including health and nutrition so far no continuity of this support. and has provided evidence to expand institutional deliveries, antenatal care, promotion, prevention and care Most of the World Bank support across support to smallholder famers. postnatal care, family planning, and services, family grants, and a network sectors has not been coordinated to Trust funds have generated knowledge for immunization. Success factors have of volunteer counselors to provide address nutrition needs in the same collaborative approaches. The Nordic trust included the use of volunteers home-based support and parental communities. However, there has been fund support to SRHR has supported (reaching 20,000 families) to identify education. Emphasis is on access to simultaneous support to needs in analyses by the health sector, PAHO, and pregnant women in rural areas and services for vulnerable populations health, social protection, water, and UNFPA and informed a multisectoral the coordination among social and involving the community as a agriculture in Jinotega during FY11–17. strategy on adolescents’ SRHR. Trust services. partner. Madriz, Leon, and Chinandega have funds and partner collaboration are Agricultural support has contributed to Other support has extended to the water received health, social protection, and important in supporting evidence increased productivity and food security and agriculture sectors. In water, water support during FY11 – 17, and gathering. in two regions. Interventions have support has developed supply chains Boaco, Chontales, RACCN, and RACCS helped improve biofortification and for rural water and sanitation services, have received support in the health, diversification of crops. A success including community-based agriculture, and water sectors. factor has been promoting synergies committees, latrines, and piped water. Moreover, the GAFSP support led to with other projects. In agriculture, support has the Caribbean Coast Food Security The World Bank has contributed strengthened farmers’ solidarity project (P148809) in FY15. The project strongly to improving sanitation, access groups, agriculture inputs, provides a unique example of to water, and reduced open defecation. biofortification, and community agriculture leading the coordination of The Rural Water Supply and volunteers to promote nutrition- interventions in two regions with other Sanitation (P106283) project sensitive agriculture, but this support World Bank projects to reach improved access to water for over has been limited to two regions and beneficiaries in lower-income and 69,000 people and sanitation for over one project. hard-to-reach communities. 44,000 people. The World Bank has contributed to parenting improvements. About 18,000 lower-income families have been reached by social safety nets, to improve schooling and parenting skills, including time that parents spend with children and nutrition. The World Bank has contributed to improvements in sexual and 268 Appendix G Case Studies reproductive health for girls. This has been through its support the multisectoral strategy on SRHR. Niger In health, the World Bank has supported The CPF (FY18–22) has prioritized Diagnostics evidence has informed policies. There were limited improvements in the expansion of the basic package of nutrition to improve human capital. This This includes a 2019 assessment of food nutrition outcomes during the services. This has included growth includes an emphasis on nutrition in security policies and programs. In social evaluation period. The magnitude of monitoring and promotion, nutrition early childhood, diet diversity in protection especially, the World Bank has some outcomes such as stunted counseling on IYCF, the referral of agriculture, maternal health, and provided consistent support to policies growth and anemia have decreased, malnourished children for treatment, women’s and girls’ empowerment. In and a framework to better target and but levels of stunted growth, wasting, and home-based care for childhood the FY13–18 CPS, nutrition was coordinate social protection services. A and undernourishment remain illness. It has also supported addressed in a more limited way diagnostic on land management has unacceptably high, above the pregnancy care, immunization, and through rural health and food security. provided an analytical basis for the regional average. family planning services in remote The World Bank has provided support to preparation of programs, given that this The World Bank has made some areas. nutrition interventions through health, is a critical challenge in communities, and contribution to improve breastfeeding In social protection, the World Bank has social protection, education, agriculture, may have important links to addressing and complementary feeding. This is provided extensive support through and urban sectors. Support to child nutrition in Niger. through support in social protection cash transfers. Cash transfers have feeding and parenting is in early Evaluation evidence has supported learning. to lower-income households with included accompanying measures, phases and will need to be expanded This is particularly the case in the use of children and improvements in such as SBCC on breastfeeding, child to meet needs. Overall, needs related IE evidence to develop safety net nutrition counseling in the health feeding and essential family practices, to maternal knowledge and resources programs and school grant programs to sector. The challenge remains the life skills, promotion of delayed are being addressed by social benefit lower-income families and scale of this support, given the extent pregnancy, promotion of water protection projects. Support to women and girls. The SWEDD regional of undernutrition. Moreover, the treatment, IGS activities, skills manage childhood illnesses has project also has supported an IE to learn benefits to improve diet diversity are building, creation of community responded to critical needs to reduce how to address social norms related to yet to be seen. savings groups, and home visits and the burden of malnutrition, malaria, girls’ empowerment. The World Bank has contributed to workshops on ECD and healthy diarrhea, and other illnesses. There has Leadership building has improved health child and maternal health. This has parenting and family relationships. also been a consistent addressing of and nutrition services. The World Bank has mainly been through improved The World Bank has supported girls’ needs to improve access to health provided technical assistance in a treatment of malnutrition, promotion education and WASH in schools. services. Key challenges remain the collaborative leadership program in the of bed net use, and improvements in Collaborative support between social inadequate coverage of remote rural health sector to improve reproductive pregnancy care, including malaria protection, health, and education has populations, improving the quality of health and nutrition service in districts prevention and treatment. However, also supported grants for girls to services, and addressing vulnerable and communities. This included a needs in this area remain vast. attend school. Education has also communities such as refugee stakeholder mapping of actors involved The World Bank has contributed to supported SBCC promotion on WASH, communities. improved access to health services. This 269 Appendix G Case Studies latrines, school-based deworming, Support to WASH and access to nutrient- in nutrition to identify disaggregated has been by expanding the basic vegetable gardens at schools, and rich food could be strengthened. WASH needs in communities. service package. Key improvements literacy training. has been integrated in social Trust funds have supported nutrition- are in assisted birth, antenatal care, In agriculture, there has been limited protection, education, and urban sensitive activities. The PHRD fund has nutrition counseling, and support to intensify nutritional crops and projects, but significant needs remain. supported the development of crops with immunization. However, the milk, livestock, and poultry production. Access to nutrient-rich food is high nutritional value in five rural challenge remains the lack of services This has included improved seeds, addressed through the PHRD fund, but municipalities of Niger (Kao, Bambey, in rural and remote areas. promotion of cowpeas and moringa, interventions are not yet mainstreamed Bangui, Hawandawaki, and Korgom) The World Bank has made some and demonstrations on the in agriculture programs. Support to most affected by malnutrition, especially contributions to improve access to transformation of foods in flour and diet diversity is a newer priority in the among women and children. nutrient-rich food. This is mainly oils. CPF. Reginal support has been critical to develop through improved inputs for Coordination on nutrition across projects women’s and girls’ empowerment activities productivity of staples and livestock. and sectors is lacking. Support to and to build productive assets of women. However, improvement is limited nutrition has been through individual The Sahel Women’s Empowerment and given the pilot scale of the support. sector programs. There was no Demographic Dividend project (P150080) The World Bank has made some evidence of coordinated or strong has supported policy, SBCC (husbands’ contribution to improve access to country leadership. However, the schools), girls’ empowerment (safe WASH. This has been mainly through portfolio is increasingly aligned to spaces), and health service improvements the promotion of hygiene and coordinate efforts on reproductive related to SRHR and nutrition across sanitation practices in communities, health and women and girls’ communities to increase demand for and support to develop latrines and empowerment. There have also been services. This support has addressed water points. some efforts by projects to coordinate critical issues such as early marriage. The The World Bank is contributing to food community-based interventions, which knowledge support of the Sahel Adaptive security and parental knowledge and could provide an entry point for Social Protection Program (P173603) has assets in households. This has been nutrition coordination. provided a package of productive through social protection support activities to cash transfer beneficiaries to and improvements in household build household assets and skills so they productive assets (livestock, and so may move out of poverty and become on). IEs found that productive more resilient to shocks. measures boosted income- generating activities. Beneficiaries displayed higher levels of total consumption and food security. Beneficiaries also experienced improved mental health and social 270 Appendix G Case Studies well-being. A success factor was peer learning support. The World Bank is contributing to improving social norms. This has been through support to girls’ education and increased contraceptive usage. Rwanda The World Bank has supported health The World Bank portfolio is increasingly Convening of leaders has demonstrated Nutrition outcomes have improved services. This included early support to aligned with country challenges. In nutrition commitment and helped pivot the overall—the magnitude of stunted develop PBF services. The current health and water, the World Bank had country to action. For example, a visit from growth, wasting, and LBW has health project (P164845) is further supported access to services until the World Bank president helped catalyze decreased, but levels of stunted strengthening CHWs and high-impact about 2012. Recent health support high-level leadership to act on nutrition. growth remain above the regional services. This includes PBF incentives responds to the need to improve the Engagement of actors at all levels average. Anemia also remains high. to improve nutrition interventions quality of nutrition services and (district, community, and so on), however, The World Bank has contributed to (such as early antenatal care visits, community-level outcomes. Agriculture have been critical to mobilize action on improvements in child health, early postnatal care, IFA supplementation, support has focused on challenges of nutrition. The World Bank supports the breastfeeding, and access to health nutrition counseling, and growth productivity and food security, monitoring of Imihigo, which is a services, but child feeding and maternal monitoring and promotion), and the although the address of food contract between the president and local nutrition remain challenges. This is distribution of fortified blended food. diversification could be strengthened. government leaders on achieving targets mainly though the expansion of PBF The World Bank supports safety nets Social protection addresses challenges for key programs. at the health facility and community and childcare. The World Bank is among lower-income households to Evaluation evidence is supporting learning. level, including nutrition services, providing nutrition grants to families, care for and feed children. Support in For example, IEs on PBF, nutrition- pregnancy care, and family planning. including parenting education and governance has addressed challenges sensitive social protection, and Key achievements have included incentives to use health services. This in accountability of services and agriculture are guiding implementation increased care seeking for children includes a program to employ improved the social registry system for learning. The current health project is and institutional deliveries. Success community workers and roll out reaching lower-income households. using IE learning to strengthen high- factors have been the strengthening childcare in rural areas. Other support The World Bank has targeted relevant impact health services and services of of CHWs and fit of the approach with has been to a voluntary savings policy populations. In health and social CHWs. the highly organized health system in for lower-income households, grants protection, mothers with young Analytical evidence has informed policy. For Rwanda. to carry out IGS, and social funds for children are a special target group, as example, the Rwanda Nutrition Situation The World Bank is contributing to farmers. are districts with high nutrition needs. Analysis and Policy Options program parental caregiving improvements The World Bank has provided extensive WASH support has focused on rural (P162400) has supported the government through support to child grants and support in agriculture. This included areas. Agriculture support has reached in improving the nutrition response. childcare. Initial evidence shows support to improve seeds, biofortified lower-income smallholder farmers in Similarly, a recent nutrition expenditure improved parenting behaviors of crops (such as iron-fortified beans), marshlands and hillsides. A challenge is review (P169988) has identfied needs to mother and father and improved 271 Appendix G Case Studies sustainable land management, water the weak overlap of agriculture support improve nutrition financing and child feeding, but intervention management, and postharvest with other support. accountability for multisectoral coverage remains limited. infrastructure. The kitchen garden The World Bank strategy has consistently coordination. The World Bank has contributed to program is the main initiative to addressed nutrition, but political Knowledge exchange is supporting improved access to food by increasing promote diverse foods. commitment has increased in recent innovation. For example, study visits to arable land and reducing seasonal The World Bank has supported WASH. years. The FY09–12 CAS has prioritized Peru and Indonesia and ongoing vulnerability; improving financial This included access to piped water malnutrition and its determinants partnership across the countries have access; increasing production of and sanitation through PPPs, water (water, health, food access). The FY14– helped reform safety net interventions staple crops and milk, meat, fish, and treatment, and WASH in health 18 CPS has focused on nutrition in and the interoperability of social sector eggs; and expanding kitchen garden facilities and schools. WASH is also agriculture, given that donor division of data systems. practice across communities. A part of SBCC across sectors. labor shifted the portfolio from health. Trusts funds have supported nutrition- success factor has been the The FY21–26 CPF prioritizes child sensitive approaches and partnership. For strengthening of farmers’ groups. nutrition improvements to support the example, the GAFSP financing (P124785) However, projects have not multisector government program. has supported the village kitchen contributed to improved diet Projects in health, social protection, demonstration program. Power of diversity. agriculture, and education are Nutrition and GFF are supporting The World Bank has contributed to coordinating nutrition interventions. learning to converge sectoral improved access to water and Policy operations have scaled up support. interventions across World Bank projects sanitation. This is through the Rural DPLs have supported the expansion of in a selection of districts. The GFF has Water Supply and Sanitation project interventions such as the PBF. A DPL on supported multipartner nutrition (P045182) and later DPLs. There have human capital supports interventions planning, involving multiple sectors and been large improvements in access to such as childcare. levels of government. BMGF is drinking water and sanitation supporting eMBeD (P169525) evidence- facilitates. Nevertheless, water access based learning on the national behavior remains a challenge in rural areas. change strategy. Source: Independent Evaluation Group. Note: BMGF = Bill and Melinda Gates Foundation; CAS = Country Assistance Strategy; CBN = community-based nutrition; CDD = community-driven development; CCT = conditional cash transfer; CHW = community health worker; CPF = Country Partnership Framework; CPS = Country Partnership Strategy; DPF = development policy financing; DPL = development policy loan; ECD = early childhood development; eMBeD = Mind, Behavior, and Development Unit; FY = fiscal year; GAFSP = Global Agricultural and Food Security Program; GFF = Global Financing Facility; HDW = human development worker; IE = impact evaluation; IFA = iron–folic acid; IGS = income generation support; IYCF = infant and young child feeding; LBW= low birthweight; M&E = monitoring and evaluation; MDG = Millennium Development Goal; PAHO = Pan American Health Organization; PBF = performance-based financing; PER = public expenditure review; PHRD = Japan Policy and Human Resources Development; PforR = Program-for-Results; PPP = public-private partnership; RACCN = North Caribbean Coast Autonomous Region; RACCS = South Caribbean Coast Autonomous Region; RMNCH = reproductive, maternal, newborn, and child health; SBCC = social and behavior change communication; SPJ = Social Protection and Jobs; SRHR = sexual and reproductive health rights; SUN = Scaling Up Nutrition; SWEDD = Sahel Women’s Empowerment and Demographic Dividend; UNFPA = United Nations Population Fund; UNICEF = United Nations Children’s Fund; USAID = United States Agency for International Development; WASH = water, sanitation, and hygiene. 272 Appendix G Case Studies Table G.2. Observed Results of Nutrition Support in Case Study Countries Outcome Ethiopia Indonesia Madagascar Malawi Mozambique Nicaragua Niger Rwanda Immediate Determinants Child + dietary diversity n.a. + breastfeeding + breastfeeding + food supplements + breastfeeding + parenting + early feeding + minimum + diet of children + complementary to vulnerable + diet diversity practices breastfeeding and acceptable diet + child stimulation feeding practices + breastfeeding for children < 5 + breastfeeding + parenting caregiving + exclusive and feeding + parenting + complementary behaviors breastfeeding practices behavior feeding + micronutrients Child health + feeding and care + treatment of + treatment of + Children with + treatment of + improved + treatment of + treatment and disease of children with child malnutrition malnutrition and ITNs who slept malnutrition home-based malnutrition of child illness diarrhea diarrhea under them last + deworming care of vector- + bed net use and diarrhea, + treatment of + health of child night borne diseases malnutrition malnutrition + Children with + bed net use diarrhea given increased fluids Maternal + reduced + intake IFA + micronutrients + IFA + IFA during + improved diet + treatment for + healthy health and underweight during pregnancy + health of mother antenatal care diversity for malaria during pregnancy diet pregnancy pregnant pregnancy + IFA women supplementation Underlying Determinants Access to + access to + food storage + food enrichment + Home or + fortified food + quality of food + improved inputs + food nutrient- agricultural support knowledge on backyard gardens products produced and for productivity of consumption rich food services − repayment preparation of planted consumed staples and rates of food nutrient-rich meals + food supply livestock (poultry, + + improved credit using affordable + increased food + productivity small ruminants) productivity, agricultural practices local products access staples, fortified fortified crops, livestock + production crops, + irrigation support + nutritious crops + crop yields, nutrient-rich crops livestock, milk diversification + improved (pilot) + iodized salt + home gardens inputs, storage + improved + agriculture inputs inputs, arable and livestock land, storage 273 Appendix G Case Studies + kitchen gardens Maternal + food security + enrollment + food security of + households with + beneficiaries of + retention of + food security + seasonal and ECED services by vulnerable three meals a day cash transfers children in food security childcare + access to income lower-income preschool + access to income resources support and cash people + household + household food support or cash + safety nets transfers incomes consumption + school transfers for lower- + household children income expenditures and + household assets, + household asset receiving meals + household people access to social resilience depletion productive assets services prevented (livestock, and so + health- + community day on) seeking care + productive assets behaviors + coverage of + community safety nets insurance coverage Access to + access to health + basic nutrition + supply, + caregivers of + vitamin A + health services + health services + health health and community services coverage affordability, uptake children benefiting coverage coverage service services nutrition services of RMNCH services from services in + immunization coverage + provision IFA care groups + use of services + assisted birth + immunization + immunization + antimalarials by pregnant + assisted + ANC, PNC + antenatal care women + antenatal care birth + LLINs + knowledge of + institutional + SBA benefits of RMNCH deliveries + immunization + immunization + antenatal + vitamin A services care and + vitamin A + nutrition services postnatal care + antenatal care coverage + Immunizations − child monthly weighed 274 Appendix G Case Studies Access to + access to improved n.a. + knowledge of + SBCC on WASH + access to clean + water supply + improved + clean water WASH water hygiene, sanitation practices water in rural areas hygiene and services sanitation practices + sanitation + access to latrines + community + piped water + improved facilities and sanitation projects hygiene and + latrines and latrines + functional water + improved sanitation in water points schemes sanitation facilities rural areas + piped water + water management Source: Independent Evaluation Group. Note: Results evidence in the table are from project indicators in results frameworks and evidence from evaluations supported by projects in the country. ANC = antenatal care; ECED = early childhood education and development; IFA = iron–folic acid;; ITN = insecticide-treated net; IYCF = infant and young child feeding; LLIN = long-lasting insecticide-treated net; n.a. = not applicable; PNC = postnatal care; RMNCH = reproductive, maternal, newborn, and child health; SBA = skilled birth attendance; SBCC = social and behavior change communication; WASH = water, sanitation, and hygiene. 275 Appendix G Case Studies Table G.3. World Bank Support to Behavior Change in Select Case Study Countries Maternal and Child Level of Caregiving and Nutrient- Evidence Rich Food Access to Health Access to WASH Madagascar Level 1— • Caregivers attended • Caregivers received • Caregivers at nutrition Engage nutrition sessions in the sensitization on the sites and health centers community and received benefits of accessing received messages on messages on health and nutrition hygiene, sanitation, and breastfeeding, IYCF, and services. the use of drinking cooking demonstrations. • Community nutrition water (hand washing, • Parents testified about agents provided health latrines, environmental their experiences and and nutrition education cleaning). brought food for for the community on culinary demonstrations. various vaccination themes. Level 2— • Caregivers acquired • Mothers realized the • Caregivers improved Learn knowledge on benefits of monitoring their knowledge of breastfeeding, the health and nutrition hygiene, hand washing, complementary feeding, of their children. food hygiene, use of foods rich in • Providers improved their soap, and use of latrines. micronutrients, and knowledge to promote having a diverse diet. nutrition. • Caregivers acquired • NGOs improved their knowledge to cultivate skills to supervise vegetable gardens. community nutrition. Level 3—Apply • Households cultivated • Providers carried out products with high nutrition services as per nutritional value. guidelines at supported • Mothers improved their sites as part of health breastfeeding. services. • Households used • Community nutrition peanuts, beans, and agents ensured the other legume proteins approach of the 1,000 for consumption and days, following the fortified food products, newborn baby into such as sweet potatoes growth monitoring and and yams. counseling and continuing screening children 6 to 59 months. Level 4— • Caregivers improved the Sustained quantity and quality of behavior dietary intake of children change on a sustained basis. 276 Appendix G Case Studies Related Communication by radio, Operational support to Strengthening the technical Interventions community meetings, and community nutrition sites. capacities of community campaigns. Training of caregivers on nutrition agents. Education by community monitoring their children. Provider training on nutrition nutrition agents. Training of health workers, NGOs, WASH approaches. Training in village culinary and community groups to demonstrations with local provide health and nutrition products. education. Demonstration sites for Support to improve the provision vegetable gardens. of nutrition services (training, Fortification and biofortification guidelines, supervision, and of food products. evaluation). Capacity building of community Strengthening of collaboration of actors and evaluation of the implementing actors. program by beneficiaries. Malawi Level 1— • Communities organized • District and subdistrict • Local chiefs were Engage nutrition days to engage officials in nutrition involved in the oversight adolescents. committees participated of their village WASH in activities with care practices and facilities. groups. • SBCC messages sensitize households on what factors contribute to healthy home environment. Level 2— • Caregivers improved their • Local leaders learn how Learn knowledge of feeding to promote and support practices. WASH practices. Level 3— • Care group delivered a • Providers practice new • Households improved hand Apply minimum package of guidelines on how to treat washing practices. community nutrition malnourished children and services through SBCC and children with diarrhea. group education, individual counseling, and home visits. • Caregivers improved their timely introductions of complementary foods to children. Level 4— • Mothers adhere to practice • Households improved sustained of early breastfeeding. access to drinking water behavior sources, latrines, and change sanitation facilities. Related Support to care groups on SBCC Care groups promoted maternal Care groups organized and interventions messaging on breastfeeding and and nutrition services. provided SBCC messaging on complementary feeding, cooking Nutrition training for district and WASH. demonstrations, and promotion subdistrict-level officials on SBCC Training of local leaders in of production of fruits, messaging for care groups. WASH. vegetables, and small livestock. Support for healthy home Support to nutrition days. environment (indoor air Promotion of integrated pollution, cook stoves, fly homestead farming. control). 277 Appendix G Case Studies Promotion of latrines and water treatment. Nicaragua Level 1— • Parents connected to • Teens engaged in peer • Community promoters Engage nutritionists or learning on sexual and and social facilitators promoters in project reproductive health. implemented awareness communities to learn • Home visits to mothers campaigns to reinforce food preparation. from midwives and good sanitation and health promoters hygiene practices in promoted prevention communities. and primary care services. Level 2— • Parents learned to • Mothers learn how and • Volunteer health Learn prepare new foods in when to access promoters educated on balanced meals. prevention and primary proper latrine use • Mothers learn how to be care services. (including hand better caregivers in washing). terms of their child’s health, after receiving support from community volunteers. Level 3— • Families with children • Health unit staff Apply under five produced and monitored growth and consumed a greater development of children variety of foods in under six in the lowest- remote communities. income municipalities, • Women and children primarily in the Dry increased the number of Corridor. food groups they • Health care workers consume, from < = 4 routinely evaluated groups to > = 5 groups women’s prenatal (36 percent baseline, nutritional status and 80 percent final). provided pre-and postnatal nutritional supplements. Level 4— • Women increased their sustained use of a package of behavior maternal and change reproductive health services (antenatal care, prenatal care, contraceptives), showing ongoing behavior change. 278 Appendix G Case Studies Related Training of nutritionists and Support to local networks of Support to network of interventions promoters to support midwives and health promoters and social families. promoters in vulnerable facilitators in communities. Counseling and workshops communities. Support to local NGOs and on parenting and family Support to maternal homes in municipal sanitation relationships in communities. municipalities. campaigns. Breastfeeding support and Sexual and reproductive complementary feeding health training of youth. counseling for parents. Rwanda Level 1— • Community actors (lead • CHW visited households • Local leaders promoted Engage farmers, farm field for child disease WASH initiatives. schools, health workers, management, nutrition and self-help groups) counseling, family promoted food and planning, and referrals of dietary practices, kitchen pregnant women to gardens, savings plans, services. social funds, and others. • ECD workers selected from the community were trained to provide caregiving in first-phase areas. Level 2— • Households planted • CHW diagnosed child • Households in home- Learn kitchen gardens. undernutrition and based ECD programs • Communities reported undernutrition increased their WASH constructed drying and cases to the health practices. storage facilities for lean facility. season. Level 3— • Caregivers improved • Caregivers increased • Households adopted Apply parenting practices in preventive care visits for water filtration and the community-based children. chlorination practices. ECD program (child • Communities managed stimulation and feeding). piped water through • Households used PPPs. fortified seeds and food products, such as beans and sweet potato rich in vitamin A. Level 4— • Communities continue • Women increased their • Households improved sustained savings plans and social use of family planning, access to clean drinking behavior funds to pay for health antenatal care, and water sources. change insurance and increased postnatal care on a their food consumption. continued basis. • Women increased early initiation of breastfeeding, and 279 Appendix G Case Studies breastfeeding continues to be high. Related Agriculture extension worker PBF incentives for CHW and Local leaders and CHW interventions farmers were trained on providers in facilities. trained in WASH, including nutrition-related content Training provided to CHW in monitoring. (such as food processing, provision of a package of ECD workers supported diversification). preventive and curative clean toilets and organized Promotion of the services for children and WASH training for parents. organization of village mothers. CHW received incentives for kitchens. household use of water Community-based ECD in treatment. first-phase communities Community PPPs were registering lower-income formed under the rural water families. project. Introduction of voluntary saving plans for food security shocks. Support to self-help groups to set up social funds. Source: Independent Evaluation Group. Note: The levels of the behavior change map are defined as level 1—engage: the actor gained awareness and motivation for changing behavior, 2—learn: the actor developed new knowledge or skills; 3—apply: the actor draws on available resources and programs as needed to use new knowledge and skills and adopt new practices; 4—sustained behavior change or institutional change: a consistent change in actors to improve a nutrition-related determinant. Institutional changes can be achieved for caregivers and households in terms of consistency or norms in practices to care for children, for community groups in terms of promoting and perpetuating social norms, and for service providers in terms of functioning more efficiently and effectively. ECD = early childhood development; IYCF = infant and young child feeding; NGO = nongovernmental organization; PBF = performance-based financing; PPP = public-private partnership; SBCC = social and behavior change communication; WASH = water, sanitation, and hygiene. 280 Appendix G Case Studies Figure G.1. Ethiopia Project Timeline and Theory of Change for World Bank Nutrition Support a. Timeline 281 Appendix G Case Studies b. Theory of change Source: Independent Evaluation Group. Note: The box colors in panel a indicate the World Bank Global Practice responsible for the lending: brown = Social Protection; gray = Water; green = Agriculture; dark blue = Health; light blue = Education; shaded = active or new; nonshaded = closed. (+) = improvement; (−) = decline; (n/c) = no change; AfDB = African Development Bank; AFR = Africa; Ag = agriculture; APL = adaptable program loan; BMGF = Bill and Melinda Gates Foundation; CINUS = Comprehensive Integrated Nutrition Services; DFID = Department for International Development (UK); Ed = education; IFA = iron–folic acid; IFAD = International Fund for Agricultural Development; IFPRI = International Food Policy Research Institute; IPF = investment project financing; JSDF = Japan Social Development Fund; LLIN = long-lasting insecticidal net; M&E = monitoring and evaluation; P4R = Program-for-Results; SCD = Systematic Country Diagnostic; SDG = Sustainable Development Goal; SIL = sector investment loan; SP = Social Protection; SSA = Sub-Saharan Africa; UNFPA = United Nations Population Fund; UNICEF = United Nations Children’s Fund; USAID = United States Agency for International Development; WASH = water, sanitation, and hygiene; WFP = World Food Programme; WHO = World Health Organization. 282 Appendix G Case Studies Figure G.2. Indonesia Project Timeline and Theory of Change for World Bank Nutrition Support a. Timeline 283 Appendix G Case Studies b. Theory of change Source: Independent Evaluation Group. Note: The box colors in panel a indicate the World Bank Global Practice responsible for the lending: brown = Social Protection; gray = Water; green = Agriculture; dark blue = Health; light blue = Education. (+) = improvement; (−) = decline; (n/c) = no or minimum change; ADB = Asian Development Bank; AF = additional financing; ANC = antenatal care; AusAID = Australia Agency for International Development; CCT = conditional cash transfer; CDD = community-driven development; CIDA = Canadian International Development Association; CSO = civil society organization; DFAT = Australian Department of Foreign Affairs and Trade; EAP = East Asia and Pacific; ECD = early childhood development; ECED = early childhood education and development; Ed = Education; Germas = Healthy Living Community Movement; GFATM = Global Fund to Fight AIDS, Tuberculosis and Malaria; GFF = Global Financing Facility; GIZ = German Development Agency; HH = household; IFA = iron–folic acid; IFAD = International Fund for Agricultural Development; I out-of-pocket; PF = investment project financing; JICA = Japan International Cooperation Agency; JSDF = Japan Social Development Fund; MCAI = Millennium Challenge Account Indonesia; OOP = P4R = Program-for-Results; PIS-PK = Program of Healthy Indonesia with Family Approach; PKH = Family Hope Program; PNC = postnatal care; PNPM = National Program for Community Empowerment in Rural Areas; RAD PG = Regional Action Plan of Food and Nutrition; RAN = National Action Plan for the Prevention and Management of Malnutrition; RAN PG = National Action Plan for Food and Nutrition; SBA = skilled birth attendance; SIL = sector investment loan; SP = Social Protection; SUN = Scaling Up Nutrition; SUSENAS = National Socioeconomic Survey; UNICEF = United Nations Children’s Fund; USAID = United States Agency for International Development; WASH = water, sanitation, and hygiene; WFP = World Food Programme. 284 Appendix G Case Studies Figure G.3. Madagascar Project Timeline and Theory of Change for World Bank Nutrition Support a. Timeline 285 Appendix G Case Studies b. Theory of change Source: Independent Evaluation Group. Note: The box colors in panel a indicate the World Bank Global Practice responsible for the lending: brown = Social Protection; gray = Water; green = Agriculture; dark blue = Health; light blue = Education. (+) = improvement; (−) = decline; (n/c) = no change; AFR = Africa; Ag = agriculture; CCT = conditional cash transfer; DEC = Development Economics Vice Presidency; DPL = development policy loan; ECD = early childhood development; Ed = education; FAO = Food and Agriculture Organization; GOV = Governance; IFA = iron–folic acid; IPF = investment project financing; Macro = Macroeconomics; MPA = multiphase programmatic approach; NTD = neglected tropical disease; PER = Public Expenditure Review; RMNCH = reproductive, maternal, newborn, and child health; SDI = service delivery indicators; SIL = sector investment loan; Social = Social Development; SP = Social Protection; STI = sexually transmitted infection; UNICEF = United Nations Children’s Fund; Urban = Urban Development; USAID = United States Agency for International Development; WASH = water, sanitation, and hygiene; WFP = World Food Programme; WHO = World Health Organization. 286 Appendix G Case Studies Figure G.4. Malawi Project Timeline and Theory of Change for World Bank Nutrition Support a. Timeline 287 Appendix G Case Studies b. Theory of change Source: Independent Evaluation Group. Note: The box colors in panel a indicate the World Bank practice responsible for the lending: brown = Social Protection; gray = Water; red = Social, Urban, Rural, and Resilience; green = Agriculture; dark blue = Health. (+) = improvement; (−) = decline; (n/c) = no change; AFR = Africa; Ag = Agriculture; ASWAP = Agricultural Sector wide Approach Project; FAO = Food and Agriculture Organization; FISP = Farm Input Subsidy Program; GESD = Governance to Enable Service Delivery; GOV = Governance; IFA = iron–folic acid; IPF = investment project financing; JSDF = Japan Social Development Fund; M&E = monitoring and evaluation; Macro = Macroeconomics; MASAF = Malawi Social Action Fund; NHAP = Nutrition and HIV/AIDS Project; RETF = recipient-executed trust fund; SBCC = social and behavior change communication; SCT = social cash transfer; SIL = specific investment loan; SP = Social Protection; SUN = Scaling Up Nutrition; UNICEF = United Nations Children’s Fund; Urban = Urban Development; USA ID = United States Agency for International Development; WASH = water, sanitation, and hygiene; WFP = World Food Programme. 288 Appendix G Case Studies Figure G.5. Mozambique Project Timeline and Theory of Change for World Bank Nutrition Support a. Timeline 289 Appendix G Case Studies b. Theory of change Source: Independent Evaluation Group. Note: The box colors in panel a indicate the World Bank practice responsible for the lending: brown = Social Protection; green = Agriculture; dark blue = Health. (+) = improvement; (−) = decline; (n/c) = no change; (n/d) = no data; AfDB = African Development Bank; AFR = Africa; Ag = Ag riculture; ANC = antenatal care; DFID = Department for International Development (UK); DPL = development policy loan; ECD = early childhood development; Ed = Education; FAO = Food and Agriculture Organization; GOV = governance; IFA = iron – folic acid; IPF = investment project financing; M&E = monitoring and evaluation; Macro = Macroeconomics; ORS = oral rehydration salts; PforR = Program-for-Results; SIL = sector investment loan; SP = Social Protection; USAID = United States Agency for International Development; WASH = water, sanitation, and hygiene. 290 Appendix G Case Studies Figure G.6. Nicaragua Project Timeline and Theory of Change for World Bank Nutrition Support a. Timeline 291 Appendix G Case Studies b. Theory of change Source: Independent Evaluation Group. Note: The box colors in panel a indicate the World Bank practice responsible for the lending: brown = Social Protection; gray = Water; green = Agriculture; dark blue = Health. (+) = improvement; (−) = decline; (n/c) = no change; Ag = Agriculture; ANC = antenatal care; APL = adaptable p rogram loan; CCT = conditional cash transfer; DPL = development policy loan; Ed = Education; FAO = Food and Agriculture Organization; IDB = Inter-American Development Bank; IPF = investment program financing; IT = information technology; LAC = Latin America and the Caribbean; MIFAN = Ministry of Family; PNC = postnatal care; RWSS = rural water supply and sanitation; SIL = specific investment loan; SP = Social Protection; TA = technical assistance; UNFPA = United Nations Population Fund; UNICEF = United Nations Children’s Fund; WASH = water, sanitation, and hygiene; WFP = World Food Programme. 292 Appendix G Case Studies Figure G.7. Niger Project Timeline and Theory of Change for World Bank Nutrition Support a. Timeline 293 Appendix G Case Studies b. Theory of change Source: Independent Evaluation Group. Note: The box colors in panel a indicate the World Bank practice responsible for the lending: brown = Social Protection; green = Agriculture; dark blue = Health; light blue = Education. (+) = improvement; (−) = decline; (n/c) = no change; 3N = Nigerians Nourishing Nigerians; AFR = Africa; Ag = Agriculture; ANC = antenatal care; CL4D = Collaborative Leadership for Development; DPL = development policy loan; ECD = early childhood development; Ed = Education; FAO = Food and Agriculture Organization; HH = household; IPF = investment project financing; PARCA = Refugees and Host Communities Support Project; SBCC = social and behavior change communication; SIL = specific investment loan; SP = Social Protection; SUN = Scaling Up Nutrition; SWEDD = Sahel Women’s Empowerment and Demographics; UCT = unconditional cash transfer; UNFPA = United Nations Population Fund; UNICEF = United Nations Children’s Fund; WASH = water, sanitation, and hygiene; WFP = World Food Programme. 294 Appendix G Case Studies Figure G.8. Rwanda Project Timeline and Theory of Change for World Bank Nutrition Support a. Timeline 295 Appendix G Case Studies b. Theory of change Source: Independent Evaluation Group. Note: The box colors in panel a indicate the World Bank practice responsible for the lending: brown = Social Protection; gray = Water; pink = Macroeconomics; green = Agriculture; dark blue = Health; light blue = Education; black = multisector. (+) = improvement; (−) = decline; (n/c) = no or minimum change; AfDB = African Development Bank; AFR = Africa; Ag = Agriculture; ANC = antenatal care; APL = adaptable program loan; CBHI = community-based health insurance; CHW = community health worker; DFID = Department for International Development (UK); DPL = development policy loan; ECD = early childhood development; Ed = Education; eMBeD = Mind, Behavior, and Development; GAFSP = Global Agriculture and Food Security Program; IFA = iron–folic acid; IPF = investment project financing; LWH = land husbandry, water irrigation, hillside irrigation; M&E = monitoring and evaluation; Macro = Macroeconomics; PBF = performance-based financing; PforR = Program-for-Results; PNC = postnatal care; PPP = public-private partnership; RSSP = Rural Sector Support Project; SAIP = Sustainable Agricultural Intensification Project; SIL = sector investment loan; SP = Social Protection; SSP = Strengthening Social Protection; UNICEF = United Nations Children’s Fund; USAID = United States Agency for International Development; WASH = water, sanitation, and hygiene. 296 Appendix G Case Studies Reference UNICEF (United Nations Children’s Fund), WHO (World Health Organization), and World Bank. 2019. Levels and Trends in Child Malnutrition: Key Findings of the 2019 Edition of the Joint Child Malnutrition Estimates. Geneva: WHO. 297 Appendix H. Stocktaking of Multidimensional Approaches Objective and Methodology A stocktaking exercise was conducted to (i) develop a qualitative understanding of multisectoral approaches to nutrition in different country contexts, and (ii) understand how the World Bank helped enhance multisectoral coordination through institutional capacity building during the 10-year evaluation period. Box H.1. Selected Countries for Multisectoral Stocktaking • Bangladesh • Ethiopia • Indonesia • Madagascar • Malawi • Mozambique • Nepal • Nicaragua • Niger • Peru • Rwanda • Senegal Source: The stocktaking focuses on a purposeful sample of 12 countries (box H.1). The sample includes those countries selected for case studies in the evaluation, plus four others among the list of case study candidates from the 64 countries in the evaluation portfolio. These cases are of interest given that their country lending portfolios have a high degree of multidimensionality1 in terms of covering nutrition-specific and nutrition-sensitive 1The degree of country multidimensionality is defined as the sum of nutrition-specific or nutrition-sensitive intervention dimensions present in a country portfolio at any point during the 298 Appendix H Stocktaking of Multidimensional Approaches interventions during the evaluation period (71 percent coverage of intervention areas, compared with 52 percent in the overall portfolio). A qualitative stocktaking template has been developed to capture descriptive details consistently across countries at the national and subnational levels. Data were collected for each country by reviewing government documents on nutrition (such as plans), the Scaling Up Nutrition (SUN) joint assessments, and published case studies and articles identified through PubMed and EconLit searches. The stocktaking reviews countries’ institutional arrangements for the coordination of nutrition, delivery of interventions, and behavioral change communication (BCC). The portfolio review data and case study evidence are then used to understand how the World Bank has contributed to institutional strengthening of multisectoral arrangements in these countries (see appendix G on the country case studies). This appendix focuses on the World Bank’s support to institutional strengthening of multisectoral coordination. These findings provide the basis for developing typologies for characterizing multisectoral approaches to nutrition in different country contexts and for highlighting factors that help facilitate or hinder multisectoral coordination. Multisectoral Nutrition Approaches Tables H.1 and H.2 summarize the institutional arrangements for multisectoral nutrition in the 12 countries. Countries adopt different institutional arrangements for coordination of nutrition policies, strategies, and plans at the national and subnational levels; delivering nutrition-related interventions; and implementing BCC strategies. Some patterns of similarities arise in this sample. Countries that centralize nutrition planning at the presidential or prime minister level tend to also have a decentralized, multisectoral coordination at the subnational level. Having decentralized, multisector coordination at the subnational level appears to be consistent with providing more developed, coordinated support to services in communities. Furthermore, countries whose nutrition coordination remains under the health sector tend to have a BCC strategy embedded in various programs or sector strategies with limited overarching coordination of nutrition messages. In this sample, three pairs of countries share similar institutional arrangements across all dimensions: Indonesia and Senegal; Madagascar and Mozambique; and Ethiopia and Niger. However, the specific institutional evaluation period divided by the total possible number of specific and sensitive dimensions, which is equal to eight: social safety nets, WASH approaches, health and nutrition services, agriculture and food approaches, ECD, diet and breastfeeding support, child disease prevention and treatment, and adolescent health. 299 Appendix H Stocktaking of Multidimensional Approaches arrangements across countries differ significantly in the extent that there has been investment in their functional development. Table H.1 Institutional Arrangements for Multisectoral Nutrition Approaches in 12 Selected Countries Institutional Arrangement Typology of Nutrition Coordination Countries National Coordination by central government office • Mozambique • Bangladesh coordination (such as planning, prime minister’s office) • Madagascar • Nepal • Senegal • Indonesia Coordination led by health sector, with roles • Ethiopia • Malawi of other sectors and in some cases strong • Niger • Nicaragua links to the country’s community health program Coordination led by another social sector • Peru • Rwanda ministry or program (social development, early child development), with central government leadership Subnational Decentralized, multisectoral coordination of • Peru • Malawi coordination regions and district in planning, M&E and • Mozambique • Nepal learning, financing, and implementation of • Madagascar • Indonesia interventions • Rwanda • Senegal • Bangladesh Nutrition activities mainly coordinated by • Niger • Nicaragua health sector or other implementing sectors • Ethiopia* Delivery of Groups in communities and sectors (such as • Rwanda • Senegal services in extension agents) developed coordinated • Malawi • Indonesia communities support to households and community • Mozambique • Nicaragua and to households • Nepal • Peru Sector extension systems (health, agriculture, • Niger • Bangladesh social protection) deliver planned services, • Madagascar • Ethiopia with limited coordination Behavior Multisector communication strategy with • Rwanda • Senegal change common messages • Malawi • Bangladesh communication • Indonesia Targeted in programs for vulnerable groups • Peru • Nepal • Nicaragua Embedded in various programs and • Niger • Madagascar interventions or in sector strategies; no clear • Mozambique • Ethiopia coordination outside of sector 300 Appendix H Stocktaking of Multidimensional Approaches Source: Independent Evaluation Group. Note: * In Ethiopia, the main coordinating sector is health, and there is a regional and local coordinating body chaired by local government. M&E = monitoring and evaluation. Table H.2 Stocktaking of Multisectoral Approaches in 12 Selected Countries Subnational Behavior Country National Coordination Coordination Delivery of Services Change Bangladesh National secretariat, Multisector Health sector National chaired by prime minister, coordination by district community clinics, advocacy plan for being revitalized to and subdistrict nutrition CHWs, extension nutrition coordinate nutrition; coordination services to low- previously under health committees, chaired by income farmers, and district commissioner NGOs and Upajilla executive officer. All sectors and civil society are members Ethiopia National nutrition Multisectoral Community-level Health strategy coordinating body, coordination by Zonal, development groups includes chaired by health sector Woreda, and Kebele are entry points, such behavior change administrative offices; as health extension promotion anchored in health workers, farmers field bureau at regional level schools, and health development army Indonesia Coordinated by Ministry At province and district Health extension National social of National Development levels, planning offices system, human behavior change Planning and Office of the coordinate across development communication Vice President sectors to develop and workers support (SBCC) strategy monitor plan convergence of developed by implementation priority interventions Ministry of and volunteers and Health facilitators at village level Madagascar Anchored in prime Regional councils are Community nutrition Projects include minister’s office in a the subnational program provides a behavior change national nutrition council extension of the platform for activities; and office national office; linked to coordination across nutrition plan responsible for communal committee social sectors, and prioritizes the coordination and for social development, NGOs. Interventions in development of a implementation chaired by the mayor, health, agriculture, multisectoral which integrates WASH, education, behavior change nutrition coordination social protection, strategy into its agenda among other areas Malawi Coordinated by Multisector district Care groups provide National Department of Nutrition, nutrition coordination platform for education HIV/AIDS under Ministry committees, chaired by multisectoral learning communication of Health principal nutrition BCC in cooperation strategy outlines HIV/AIDS officer; with frontline workers key actions for subdistrict committees that make up village nutrition comprising supervisors 301 Appendix H Stocktaking of Multidimensional Approaches of extension works or nutrition coordinating frontline staff, village committee coordinators, and civil society Mozambique National platform chaired Establishing provincial- Agriculture extension Program includes by Prime Minister; and district-level policy workers, CHW, and behavior change coordinated by technical and technical community care activities; a secretariat consultation platforms groups, mother behavior change leaders, community strategy is leaders, NGOs planned Nepal National Planning Nutrition coordination Village development Behaviorchange Commission leads overall structures at the councils link with the activities for nutrition coordination provincial, district, local coordination marginalized and and facilitates strategy municipal, and ward cascading structures lowest-income implementation levels population segments Nicaragua Coordinated by health Coordinated by Integrated network of Integrated in sector and other ministries of health, health care providers, health, education ministries education, family affairs, NGOs, community and social agriculture, family and volunteers, midwives, protection community economics, social facilitators and community fishing and public works specialists carry out programs, community targeting interventions; and ECD vulnerable services groups Niger National platform, Nutrition is part of Extension services of Projects include coordinated regional and communal sectors deliver behavior change mainly by health sector, development plans in interventions, with activities despite multisector the same way as community groups committee agriculture, health, or water Peru Ministry of Development Decentralized approach Primary health Behavior change and Social Inclusion led by regional and local services centered on coordinates national government, with at community and vulnerable strategy; sector ministries district multisectoral household levels, populations implement committees for social programs via programs, such as social cash transfers, and protection government’s identity registry Rwanda Chaired by National Multisectoral nutrition Village committees National SBCC Childhood committees for the organize support with strategy aligned Development Agency in District Plans to household leaders across sectors Ministry of Gender and Eliminate Malnutrition, and local service Family Protection, with chaired by vice mayors, providers, CHWs, ECD support of president bring together relevant caregivers, and sectors (health, agriculture promoters agriculture, social protection, hygiene, and sanitation) and other 302 Appendix H Stocktaking of Multidimensional Approaches actors (NGOs, partners) in the district Senegal National coordination Governor is head of Local management National platform chaired by prime administration at committees at village communication minister regional level; at level, with NGOs and and advocacy commune level, mayor sectoral providers and strategy coordinates nutrition leadership of mayor addresses social services and norms at interventions community level Source: Independent Evaluation Group. Note: AIDS = acquired immunodeficiency syndrome; CCT = conditional cash transfer; CHW = community health worker; ECD = early childhood development; Med = medium; HIV = human immunodeficiency virus; NGO = nongovernmental organization; SBCC = social behavior change communication. Country multidimensionality is defined as the sum of nutrition-specific or nutrition-sensitive intervention dimensions present in a country, divided by the total possible number of nutrition-specific and nutrition-sensitive dimensions. Common Factors for Successes and Challenges in Multisectoral Approaches While multisectoral approaches to address nutrition are in an emerging stage in most of the countries, the stocktaking exercise highlights how important common factors facilitate or hinder coordination efforts and results, including clarity of mandates and leadership, engagement of local governments, organization of local delivery systems, strengthened financing, planning, and M&E (box H.2). Box H.2 Factors that Facilitated Multisectoral Coordination Efforts • Consistency of national leadership in relation to a mandated program or framework to coordinate actors and roles of relevant sectoral ministries • Developed role of subnational government to coordinate multisectoral actions • Organization of sectoral extension services and community actors to deliver an integrated package of interventions tailored to local needs, with consistent messaging • Strengthened financing and planning • M&E, and knowledge sharing approaches that support multisectoral interoperability of decisions, actions, and learning (rather than single-sector systems) on nutrition interventions at different levels. Mandates and National Leadership Among all countries, whether nutrition is coordinated by a central government office or by a specific sector (health or other social sector), key success factors include having consistent national leadership and a defined mandate or framework to integrate actions. Senegal offers one example of consistent national leadership. In 2001, the head of state created a national coordination unit to bring together all relevant sectors, including 303 Appendix H Stocktaking of Multidimensional Approaches education, family and social protection, health, livestock, agriculture, fisheries, trade, industry, higher education and research, decentralization, and environment. Actions of the coordination unit aligned with a nutrition policy and multisectoral strategic plan, under which the nutrition-related programs were harmonized. Rwanda offers another example of how the leadership of the president under the National Child Development Agency has increased attention, resources, and effectiveness of plans dedicated to nutrition across multiple sectors, with clear actions and ownership of districts. Where central leadership and or multisectoral mechanism for coordination are weaker (such as in Ethiopia, Niger, Madagascar, and Mozambique), a common challenge is ensuring the accountability of actions implemented among sectors. In countries with health sector leadership, nutrition is part of the health program, but a common challenge is to integrate agricultural and WASH approaches with health and other social services support. For example, in Nicaragua, the health sector mainly coordinates nutrition at the subnational level, which coordinates with education and social protection based on the Community and Family Health Model (MOSAFC) program. Other ministries involved in nutrition-sensitive initiatives—such as agriculture, public works (sanitation and hygiene), and fishing and aquaculture— implement separate sector activities, and there remains a need to strengthen coordination between these interventions and the MOSAFC support to nutrition. In Ethiopia and Niger, nutrition is also delivered as part of the health service package, with limited coordination with other sectors. Empowerment of Local Governments Another key factor among the countries is the empowerment of local government to facilitate and prioritize multisectoral actions on nutrition. In countries where local government roles are less developed, collaboration among sectors is often weak. In Bangladesh, Ethiopia, Indonesia, Madagascar, Malawi, Nepal, Rwanda, and Senegal, multisectoral committees at the district level have various levels of functionality. A strong coordination structure (such as in Senegal) often links to other levels of government, with clear lines of accountability of leaders to engage multiple sectors. For example, Peru offers a decentralized approach led by regional and local government, with multisectoral committees for programs, such as social protection. Without clarity of roles, even dedicated local leaders are limited to take actions to improve nutrition (such as in Ethiopia and Madagascar). Effective coordination appears to require centralizing some functions to facilitate learning and policy guidance to expand actions while decentralizing others to empower local authorities to plan, monitor, and make decisions. In some countries, there has been a continuing process to build buy-in of local leaders and converge services to develop an 304 Appendix H Stocktaking of Multidimensional Approaches integrated multidimensional package that addresses health, agriculture, WASH, and other local needs. Indonesia demonstrates how such support can be designed and phased in over time. To build commitment across the levels of government, the vice president brought together officials from provinces and districts to align policy and actions on reducing stunted growth through the national strategy. Heads of districts signed a pledge to hold stunting summits, implement convergence actions for nutrition interventions, collect and publish data on stunted growth and intervention delivery, formulate a behavior change communication policy, and support village-level nutrition intervention convergence. The ministry responsible for local government also organized subdistrict-level and village-level leaders. Rwanda is similarly strengthening district governments to converge health, social protection, agriculture, and WASH services and integrated nutrition in district performance contracts. Sectoral Extension Services Most countries have leveraged extension services (health, agriculture, social promoters, and early childhood development workers) and or community groups to deliver interventions. However, the organization of these actors to coordinate interventions in an integrated multidimensional package is often more advanced in countries that emphasize multisectoral coordination. Countries vary widely in the extent of capacity building that has been done to organize actors to deliver interventions, the multidimensionality of the package being delivered, and partnerships among local government, community groups, and sectors to coordinate and integrate interventions so they can benefit key groups in communities. In some countries, community groups (such as CHWs, mother leaders, women development groups, and farmers groups) are being strengthened to deliver interventions. In Malawi, care groups provide a platform for multisectoral behavior change communication in cooperation with frontline workers that make up village nutrition coordinating committees. Thus, members of care groups can deliver a multidimensional package to households with young children through home visits and cluster meetings. Implementing agencies and partners use the care groups as community entry points and so do other actors with similar targets. Care groups were also developed in Mozambique through the health sector. In Senegal, the NGO or community executive agency facilitates the community programming and implementation. In many of the countries, such as Ethiopia, Madagascar, and Mozambique, the coordination of community actors needs further strengthening, particularly links between actors delivering health and agriculture interventions, such as the promotion of nutritious food production and preparation among families with young children. 305 Appendix H Stocktaking of Multidimensional Approaches Learning is still emerging on how to converge interventions to meet the needs of communities. Barriers to effectively integrate the delivery of interventions at the local level can include problems related to geographic conditions and inadequate targeting of needs to differentiate services across households or communities to converge support for specific beneficiary groups. In Indonesia, decentralized service delivery faces substantial geographic challenges with about 75,000 villages with target households across 6,000 islands. This has presented logistical difficulties with scaling up a community platform. Rwanda is similarly learning to converge interventions of health, social protection, and agriculture to benefit households in the same communities. Other challenges to integrate interventions across sectors include that targeted beneficiaries often differ based on the design of interventions focused on differing sectoral objectives. Also, quality concerns happen when interventions are brought to scale quickly or community groups are overloaded with responsibilities. Another important factor is having consistent communication on nutrition in programs across different sectors or stakeholders, with a means of M&E and learning. In some countries, SBCC is a targeted part of the national action framework for nutrition (such as Bangladesh, Indonesia, Malawi, Rwanda, and Senegal). In Malawi and Senegal, SBCC was emphasized for years, whereas in other countries, such as Mozambique, having an SBCC strategy for nutrition is a recent or future plan. In Nepal and Peru, SBCC has been embedded in programs for vulnerable groups or more marginalized and low-income population segments. In Malawi, the National Education Communication Strategy outlines key actions for nutrition information and communication for effective behavior change, including key stakeholders and community delivery platforms and coordination. In 2018, Rwanda developed a national SBCC strategy to guide consistent messaging across sectors, as well as M&E and learning. Ethiopia does not have a stand-alone SBCC strategy for nutrition interventions rather it has been an integral part of the health program and outlined as a key element of the national nutrition program. Financing, Planning, and Monitoring and Evaluation A common accountability challenge among the countries is the coordination across sectoral systems (national and decentralized) to track nutrition financing and planned interventions. For example, Bangladesh, Indonesia, and Rwanda conducted nutrition performance expenditure reviews and are improving budget tracking for nutrition. One challenge is alignment between the decentralized flow of financing and the coordination roles of nutrition services at the decentralized level. In Indonesia, the district health office that receives the funding differs from the district planning agency in charge of coordinating, planning, and budgeting of the nutrition interventions program. Another 306 Appendix H Stocktaking of Multidimensional Approaches challenge is to track the money that has been allocated to sector services in terms of whether it is being used for the right activities and for the right target groups and geographical locations. In Bangladesh, sectoral line offices receive funds from their respective central ministry for implementation of their annual plan. In Ethiopia and Madagascar, the limited, decentralized decision-making and financing of nutrition constrains local capacities to plan multisectoral nutrition interventions. In Malawi, even though central coordination is strong, collaboration across sectors remains challenging due to the limited interoperability of sectoral systems for financing, planning, and supporting services. A related challenge is to facilitate M&E data collection, reporting, and performance measurements across multiple sectors and stakeholders and levels of implementation. For example, Mozambique establishes a national level M&E system to track nutrition- related indicators involving sectoral ministries and provincial and district departments, but the national coordination office lacks authority to enforce reporting across sectors and different levels of government. In Nicaragua, there is also a challenge related to limited data on the different dimensions of the nutrition situation, particularly at the local level that would allow for adequate decision-making and prioritization and targeting of interventions. In Ethiopia, the health sectors reporting system includes limited nutrition indicators, most of which monitor nutrition-specific programs. The agriculture sector is also improving its monitoring of nutrition, but there is limited M&E of integrated nutrition achievements and weak data systems at the decentralized level for data use and decision-making. Rwanda is in the process of improving its M&E system and has already successfully improved the interoperability of sectoral information systems for ECD, health, social protection, and birth registration. The use of M&E for multisectoral decision. In Senegal, the use of nutrition performance data and inputs of various frontline agencies by regional and local authorities in Senegal has been instrumental in ensuring the identification of problems, collaboration, accountabilities of various sectors, and a common vision of shared goals. In Peru, the Ministry of Social Development and Inclusion uses the information from sectoral monitoring systems for a dashboard report at regional, departmental, and district levels. The dashboard was launched in 2016 to provide districts with quarterly progress reports, including general characteristics of the district, health, and education indicators, and housing conditions. Malawi has a national M&E system to track key indicators for the achievement of goals stipulated in the national policy and strategic plan for nutrition. The system relies on data from the district level, uploaded on quarterly basis. As of 2019, 75 percent of districts in Malawi have reported at least some nutrition data from the health, agriculture, gender, and education sectors, as well as data on coordination and monitoring. 307 Appendix H Stocktaking of Multidimensional Approaches World Bank Support to Strengthen Institutions and Multisectoral Coordination This section focuses on the support provided by the World Bank to strengthen nutrition multisectoral arrangements and institutional capacities in countries. Some World Bank support focuses on the institutional strengthening of multisectoral approaches, and other support across the stocktaking countries focuses on capacities in specific sectors (health, agriculture, social protection). Across the countries, the main emphasis is on improving nutrition service delivery (figure H.1). This emphasis on service delivery is consistent with the overall nutrition portfolio (see appendix D). Box 3.1 provides examples of how the World Bank contributes to strengthening multisectoral approaches across the 12 stocktaking countries. This work takes place in diverse country contexts with other development partners and is often still at an emerging stage. Figure H.1. Focus of World Bank Interventions for Strengthening Institutions, 1998–19 Source: Independent Evaluation Group. Note: Figure shows data on 271 institutional strengthening interventions from 79 projects in the 12 countries included in the stocktaking exercise (Bangladesh, Ethiopia, Indonesia, Madagascar, Malawi, Mozambique, Nepal, Nicaragua, Niger, Peru, Rwanda, and Senegal). 308 Appendix H Stocktaking of Multidimensional Approaches The intensity of institutional strengthening support varies across countries, as does the multidimensionality of the country’s portfolio and success of project performance. Indonesia, Madagascar, Nicaragua, Rwanda, and Senegal, receive medium-to-high institutional strengthening support, which broadly covers nutrition-specific and nutrition-sensitive interventions and support relatively good achievement of project indicators (figure H.2). In Bangladesh, the World Bank has a low level of institutional strengthening support, a narrow intervention focus, relatively weaker project performance, and no evidence of institutional strengthening achievement in closed projects. In Niger, the diversity of interventions in the portfolio is high, and there is a low level of institutional strengthening support in the portfolio, which has been successful to develop capacities within sectors, such as health and social protection. Overall, the focus of institutional strengthening varies widely across countries. In most of the countries, the main attention of institutional strengthening is the development of community programs, whereas in some countries, such as Malawi and Senegal, there is a balance of support to develop policies and services in communities. Figure H.2. Multidimensionality of Country Portfolio and Project Performance 1.00 45 40 0.75 35 30 25 0.50 20 15 0.25 10 5 0.00 0 Overall Project Performance Project Performance of non-Institutional Strengthening results Project Performance of Institutional Strengthening results Multidimensionality at country level # of Institutional Strengthening Interventions Source: Independent Evaluation Group. Note: Overall project performance refers to the achievement rate of all projects in the country portfolio; project performance of non-institutional strengthening results account for achievement rates of nutrition outcomes, determinants, and social norms. Countries with highly multidimensional portfolios tend to have better project performance. This holds for all the results dimensions of the conceptual framework, including nutrition outcomes, nutrition determinants, social norms, and institutional strengthening contributions. 309 Appendix H Stocktaking of Multidimensional Approaches Global portfolio findings show that countries that are more successful at strengthening institutional capacities toward policy, service delivery, and stakeholder engagement also show better performance in nutrition results and its determinants. This suggests that the adequacy of the enabling environment underlies the countries’ potential for improving the underlying determinants of undernutrition, and in turn, outcomes (see appendix D for portfolio review). For instance, in Senegal, the World Bank contributes to improving nutrition policies, strategies, and development plans, including the adoption of an efficient, child-focused social cash transfer scheme; integrating nutrition indicators in monitoring tools for decision-making; enhancing the coverage and quality of health care services; and improving the engagement of citizens and civil society organizations for better accountability. Better access to health care services and maternal and care resources follows, and projects also show better breastfeeding and child feeding practices and malnutrition screening and treatment, as well as for other common childhood illness. In terms of nutrition results, the Nutrition Enhancement project and the Rapid Response Child-Focused Social Cash Transfer and Nutrition Security project contribute to increase the share of children ages 0–24 months showing adequate monthly weight gain. Country experiences suggest institutional strengthening requires consistent support to translate into improved performances. In Malawi, there are successful efforts to strengthen capacities for multisectoral coordination through the Nutrition and HIV/AIDS Project. The main challenge is the limited duration of this support to improve overall performance. Moreover, other institutional strengthening support in the portfolio is low. In Ethiopia, while investments in, and the performance of, institutional strengthening results are high, the portfolio is still young and does not yet see this support translate into high overall performance. One challenge is the need for better coordination of interventions at the subnational level. In the case study countries, successful examples of institutional strengthening supported by the World Bank include policy dialogue, leadership building, South-South knowledge exchange, evidence-based learning, support to M&E systems, and support to districts to oversee nutrition, use M&E, and strengthen extension services and community groups (table H.3). A key variation across countries is the extent of support to policy and coordination, relative to service delivery. At the national level, the World Bank supports high-level leadership; coordination of nutrition, policies, financing, and strategies; and M&E systems, diagnostics, and research and evaluation. At the district level, the World Bank supports learning, M&E, and supervision to oversee nutrition services. At the community level, the World Bank supports strengthening the targeting of services, community groups, and extension workers. 310 Appendix H Stocktaking of Multidimensional Approaches Table H.3 World Bank Contributions to Institutional Strengthening in Case Study Countries Country National Regions and Districts Community Ethiopia • Support to nutrition • Performance-based • Establishment of CBN coordination body financing and block approach, with and programs, grants for health and health extension including M&E nutrition services workers • Dialogue on nutrition • Support to develop • Support to farmer policies, such as on basic services, such cooperatives, salt iodization as education, health, livestock extension • Diagnosis and agriculture, WASH, workers, seed evaluation of and ECD groups, model nutrition situation • Capacity building of farmers, and and interventions to regional and woreda nutrition-agriculture inform policy and nutrition units cooperatives for cost-effective women to promote • Support to water programs nutrition supply and sanitation • Mobilization of schemes, including • Formation of young donor financing sanitation marketing women’s clubs, mentors, and mother • Support to support groups strengthen productive safety • Campaigns on nets program nutrition and early marriage • Cash transfers for farmers Indonesia • Study tour to Peru to • Support of • Support to teacher support National multisectoral training training, quality Strategy for Stunting programs for standards, M&E, and Reduction (Stranas) community supervision for ECD • Expansion of Stranas facilitators programs and strengthening of • District and village • Support to national leadership grants for water implement SBCC and coordination sanitation and supply strategy and WASH across sectors and good practices promotion levels of government • Support to • Community • Strengthening of implement standards performance-based M&E for greater local block grants to • Support to SBCC government and incentivize use of strategy facility performance health and education of nutrition, maternal services • Strengthening of and child priority • Leverage of nutrition financing programs community-driven • Diagnostics, policy • Strengthening of development dialogue, convening, convergence of platform to pilot a programmatic district activities frontline nutrition knowledge work (on • Strengthening convergence financing, management and approach multisectoral 311 Appendix H Stocktaking of Multidimensional Approaches nutrition, ECD, implementation of • Technical support for decentralization, nutrition activities at multisectoral social services) district level, district coordination through • Strengthening of performance subdistrict and social assistance monitoring, local village forums delivery systems and stunted growth • Training programs cash transfers surveillance, district for community diagnostics, facility facilitators accountability and human development worker mobilization Madagascar • Establishment of a • Establishment of • Establishment, multisectoral regional refinement, and coordination body multisectoral delivery of a package • Consistent flow of capacity for nutrition of CBN interventions evaluations and coordination, linked to frontline diagnostics to assess including technical services the impact of and M&E support • Expansion of community • Improved multisectoral content interventions, refine coordination and of minimum package program content and collaboration, to include key approaches, and especially across sectors, such as improve regional offices for health, education, implementation health, education, social protection, • Enhancement of and nutrition and food security (family policy dialogue, and social protection gardens, livestock programs through • Support to regions projects), ECD, and analytic work, and districts to women learning events, and supervise, oversee, empowerment study tours and provide technical • Capacity building of • Strengthening of backstopping to frontline services for M&E systems frontline service improved synergy delivery and coordination • Monitoring and supervision of service • NGO support to delivery and quality strengthen improvements, community-based including services and links to investments in basic services human resources and • Engagement of the provision of critical community and inputs community leaders in nutrition activities • Improved emergency support to communities Malawi • Support to • Learning forums to • Development of care coordination share lessons on the groups and village platforms, financing, nutrition response nutrition and M&E across districts and coordination build leadership committees 312 Appendix H Stocktaking of Multidimensional Approaches • Leadership building • Building of capacity (communication activities in relation to implement M&E materials, supplies, to policies and to practices to support training) coordinate sectors the national • Development of and stakeholders framework extension support • Development of through care groups, M&E and resource youth clubs, tracking framework community savings for nutrition and loans groups, • Support to many and farmer groups nutrition policies and • Development of the national support for home communication gardens, strategy micronutrient • Diagnostics and supplements, and evaluations to inform promotion of WASH the nutrition strategy facilities and community • Support to BCC program among farmers on nutrition Mozambique • Strengthening of • Support to deliver • Strengthening of coordination body and improve CHWs and care and its stakeholder nutrition services in groups to deliver engagement health and other nutrition services to • Support to nutrition sectors, such as benefit targeted policies, strategies, agriculture, food households M&E, and financing security, water • Engagement of (primarily though supply, and ECD NGOs to coordinate agriculture and and deliver nutrition health sector) interventions Nicaragua • Support to improve • Support to child • Support to efficiency and development centers, implement and accountability of including feeding improve community- social service delivery program, growth based health services • Dialogue on health monitoring, and • Development of cash and social policies vaccinations transfers to low- • Evaluations and • Support to maternal income families with diagnostics to inform services children program design and • Support to • Development of services strengthen package community water • Support to of services committees strengthen health integrating social • Strengthening of care model and protection, health, network of strategy to deliver and education, community community services focusing on lower- volunteers to integrating nutrition income and promote parenting vulnerable families practices; and in • Support to and children (cash agriculture to strengthen policy transfers, nutrition promote biofortified messages, 313 Appendix H Stocktaking of Multidimensional Approaches and model for social counseling, crops, livelihoods protection workshops) enterprises, nutrition • Support to develop education, and and implement gender behaviors adolescent health strategies to prevent early pregnancy and gender-based violence Niger • Support to health • Strengthening of • Support to mobilize plan and nutrition basic package of community actors in directorate in health health services, with relation to health • Diagnostics and a focus on maternal, centers to deliver evaluations of social reproductive and nutrition services in protection, nutrition child health health facilities and coordination, and • Support to district by CHWs community programs health plans • Strengthening of • Support to • Leadership building safety nets and strengthen social to implement accompanying protection systems reproductive health measures on and nutrition services nutrition, SBCC, and • Support to support on ECD for community health low-income • Support to M&E households • Support to • Training of midwives coordinate sexual • Support to school and reproductive management health and women’s committees to empowerment, engage girls in integrating nutrition school Rwanda • Building of • Support to improve • Support to CHWs leadership at highest the assessment of platform, including level nutrition in district package of nutrition • Strengthening of performance interventions coordination by frameworks and • Development of National Child plans, and safety nets for low- Development Agency stakeholder income households (policy, strategy, mobilization • Development of financing, mapping • Support to community-based stakeholder behavior agricultural family planning change, and productivity • Support to mapping nutrition- approaches agriculture extension sensitive and specific • Support to improve agents, including interventions across high-impact health farmers’ sectors) services organizations and • Diagnostics and IEs • Support to improve women’s groups, to for evidence-based M&E and supervision promote nutrition, learning on balanced diet interventions demonstration 314 Appendix H Stocktaking of Multidimensional Approaches • South-South • Rural water supply sessions, fortified knowledge exchange public-private foods, and rollout • M&E systems partnerships kitchen gardens interoperability • Support to districts • Learning to converge • Learning to measure to converge sectoral services based on behavior change services needs in selected approaches • Engagement of communities • Community health leaders to support • Development of strategy SBCC and WASH home‐based ECD promotion groups • Support to improve child and gender • Support to sensitivity of social implement SBCC protection systems strategy • Performance-based • financing strategy • Strengthening of agriculture strategies Source: Independent Evaluation Group. Note: CCT = conditional cash transfer; CHW = community health workers; ECD = early childhood development; M&E = monitoring and evaluation; NGO = nongovernmental organization; PER = public expenditure review; SBCC = social behavior change communication. Country multidimensionality is defined as the sum of nutrition-specific or nutrition- sensitive intervention dimensions present in a country, divided by the total possible number of nutrition-specific and nutrition-sensitive dimensions, which is equal to eight (this excludes all institutional strengthening and social norms dimensions). 315 Appendix I. Multivariate Regression Analysis Approach IEG conducted an econometric analysis anchored in the conceptual framework to uncover predictors of project performance. Project performance is measured for closed projects as the share of achieved results framework indicators.1 The analysis provides additional evidence for answering the third evaluation question on the extent to which World Bank interventions contribute to reduce child undernutrition outcomes and improve nutrition determinants. Based on the results of various exercises conducted in the evaluation, including country case studies and the portfolio review, seven hypotheses were tested: 1. Higher country multidimensionality is associated with better nutrition results. 1. Higher institutional strengthening (IS) achievements are associated with better project performance in improving nutrition outcomes and determinants. Moreover, the higher the intensity of IS achievement (more IS sublevels achieved within a project), the better the project’s performance. 2. Project design, community-based implementation, country ownership and institutional arrangements, and monitoring and evaluation (M&E) are important factors associated with project performance. 3. Investing in effective interventions as documented in the global literature are positively associated with project performance. 4. A better match between World Bank supported nutrition interventions and country needs is associated with better project performance. 5. Core nutrition projects (those with “nutrition” or “stunted growth” in the title or PDO and a nutrition content share in the top 40 percent of the distribution) tend to perform better than sectoral projects (noncore projects) in improving nutrition determinants outside their area of expertise (cross-sector support). 6. Projects with analytical support perform better than projects without it. The project-level analysis is based on the cross-section of 131 closed nutrition projects with available indicator achievement information and their characteristics identified in the portfolio review and analysis, including nutrition interventions, factors of success and failure, and indicators, among others. Appendix D describes these projects in detail. 316 Appendix I Multivariate Regression Analysis Empirical Evidence on Project Performance Although much research has focused on predictors of World Bank project performance, to IEG’s knowledge no previous work has investigated the drivers of nutrition-related project performance as measured by indicators. Nonetheless, the relevant empirical literature, in which performance is measured with project outcome ratings, reveals several project-level drivers of better performance that need to be considered to minimize omitted variable bias. These drivers include project design and the quality of M&E (Hussein, Kenyon, and Friedman 2018; Raimondo 2016)2; shorter project duration and the presence of additional financing (Bulman, Kolkma, and Kraay 2015); task team leader (TTL) record and predicted performance, TTL turnover, project preparation time and support from analytical work (Geli, Kraay, and Nobakht 2014; Hussein et al. 2018). Some country-level characteristics are also important, including the ratings of Country Policy and Institutional Assessments (CPIA) (Hussein et al. 2018). However, evidence indicates that project-level drivers are more important than country-level ones. For example, Denizer, Kaufmann, and Kraay (2013) found that for 6,000 World Bank projects evaluated in PPARs, ICRs, or ICRRs between 1983 and 2011, 80 percent of the variation in project outcomes could be explained by within-country and across-project variations rather than by country characteristics. Similarly, Hussein et al. (2018) found that for IPFs approved between 2005 and 2009 and evaluated by IEG, 25 percent of the variation is explained by project M&E flags, 12 percent by other project characteristics, and only 6 percent and 5 percent by staffing characteristics and country characteristics, respectively. The evaluation uses proxies for several of these project performance correlates, including a taxonomy of factors of success and failure from the portfolio review and analysis related to project design and M&E quality, among others (see appendix D). Other project-level controls are defined as they are in the literature, including project duration. Methodology The basic model for investigating the project-level drivers of nutrition indicator achievement is based on the conceptual framework and the evidence on World Bank project performance described above. The model is as follows: = + ′Χ + ′Ψ + () where yij is project performance, that is, the share of indicators achieved in project i of country j, and used in one of its three versions (overall, nutrition outcomes and determinants, and cross-sector); Χ ij is a vector of project-level characteristics, including 317 Appendix I Multivariate Regression Analysis project multidimensionality, a noncore project dummy, the quantity of achieved institutional strengthening indicators or its three sublevels (policy, service delivery, and stakeholder engagement), selected factor topics as shares of total factors in a project (project design, community-based implementation, country ownership and institutional arrangements, M&E, country context, and World Bank systems and performance), the percent match of interventions to country needs, the share of effective intervention outcomes according to the literature, an analytical support dummy, an emergency project dummy, the share of nutrition outcome indicators in a project, the share of factors with positive direction, project duration time, and project approval period; Ψj is a vector of country characteristics, including the country portfolio multidimensionality score broken down by quartiles, region dummies, income level, fragile and conflict- affected situations (FCS) status, and the non-World Bank Group nutrition-relevant foreign aid per 1,000 population in $, millions α is an intercept and εij is the error term. As a first step for testing the seven hypotheses, we use equation 1 to estimate bivariate regressions through ordinary least squares (OLS) between each of the relevant independent variables and the three measures of project performance. We then perform the multivariate regression analysis by estimating equation 1 through OLS with additional controls. Here, several model specifications are estimated for each of the three dependent variables, always including the variables of interest for testing the seven hypotheses. The choice of controls is based on the evidence in the literature and on an effort to ensure parsimony of the model given the small sample of projects. For example, three available proxies capture the enabling environment: the baseline composite score for five nutrition determinants, the CPIA rating, and the baseline government effectiveness ranking of the World Governance Indicators, yet only the latter is included as a control. Similarly, the log of nutrition commitments and a dummy on additional financing do not add explanatory power to the model across several specifications so they are excluded in the end. Results of preliminary estimations are available on request. In addition, we estimate an augmented version of equation (1) to test the second part of hypothesis number 2 on the positive association between the intensity of institutional strengthening achievement and project performance: yij = α + β1 ISachieved + β2 broad + β3 medium + β4 narrow + β5 broad × ISachieved + +β6 medium × ISachieved + β7 narrow × ISachieved + γ′Χ ij + δ′Ψj + εij (1)’ where ISachieved represents the quantity of IS indicators achieved in a project, and broad, medium and narrow represent dummy variables capturing the intensity of IS achievement. The broad dummy captures projects with three IS sublevels achieved (policy, financing and coordination; nutrition service delivery; and stakeholder engagement), irrespective of the number of IS indicators achieved within each sublevel. 318 Appendix I Multivariate Regression Analysis Similarly, medium and narrow dummies capture projects with two IS sublevels achieved and one IS sublevel achieved, respectively. A fourth dummy that is used as the base category captures projects with no IS sublevels achieved. In the augmented model, we interact the broad, medium, and narrow dummies with the total quantity of IS indicators achieved as an additional element of intensity. Limitations The econometric analysis attempts to reduce the risk of omitted variable bias by including relevant controls from the literature. However, there is still a risk that proxy variables for relevant controls do not fully capture such controls. For example, the World Bank systems and performance factor topic proxies for several relevant controls at once, like TTL record and predicted performance and TTL turnover, which are important in the literature, but it remains an imperfect measure. Also, omitted variable bias may remain a risk to the extent that there is reverse causality between project performance and factors of success or failure. For example, past project performance could influence current performance through improvements in project design and implementation that confound our factors of success and failure variables; and since the evaluation does not control for past performance this may bias the coefficients. Another limitation is that the regressions are based on a small sample of projects. The evaluation team addresses this issue by imputing missing values with regional averages for all variables used, so that the number of observations is always maximized to 131 projects. In addition, it is assumed that the sample of closed projects is representative of the whole nutrition portfolio. Results Table I.1 shows descriptive statistics of project-level variables that are used in the regression analysis. Table I.2 shows bivariate regression results of estimating equation 1 by OLS. The bivariate regression results provide preliminary evidence favoring most of the hypotheses. Country portfolio multidimensionality, institutional strengthening achievement (including all three of its subcategories and the broad and medium intensity measures), community-implementation and M&E factors, support from analytical work, and percent of interventions matching country nutrition needs, are all significantly associated with at least one measure of project performance without controlling for other variables and show the expected sign. For example, a higher country portfolio multidimensionality score is associated with better overall achievement; and the coefficients for community-based implementation and M&E factors are highly significant regardless of the dependent variable and have a positive and negative sign, respectively, consistent with their top ranking as success and failure factors. Project 319 Appendix I Multivariate Regression Analysis design factors and the share of intervention outcomes with positive evidence in the literature, are not significantly associated with project performance in these bivariate regressions. A closer look at these results is offered next in the multivariate regression analysis. Table I.3 shows the project-level multivariate regression results and shows suggestive evidence favoring most of the hypotheses. Hypothesis one: In terms of country portfolio multidimensionality, the multivariate regressions suggest there is some evidence, although not robust across all model specifications that projects in countries in the top quartile of country portfolio multidimensionality perform better than projects in the bottom quartile. On average and all else equal, projects in the top quartile perform about 13 percentage points better compared with projects in the first quartile, both on overall performance and nutrition determinant and outcome performance (Panel A, column 1 and Panel B, columns 5 and 6). Hypothesis two: Higher institutional strengthening achievements are associated with higher cross-sector achievement. This result is robust to changes in specification. On average one additional IS indicator achieved is associated with an increase of 15 percentage points in cross-sector achievement, other controls in the model constant. However, the magnitude of this positive association diminishes as the number of achieved IS indicators increases, as shown by the statistical significance of its squared term (Panel C, columns 25 and 26). The positive relationship is driven by the sublevel of policy, financing, and coordination, which is positively and significantly associated with cross-sector achievement and is again robust to changes in specification. A one unit increase in the number of achieved policy, financing and coordination indicators is associated with about a 10 percentage point increase in cross-sector achievement (Panel B, columns 18, 20, 22, and 24). In terms of IS intensity, estimating equation (1)’ shows suggestive evidence that projects with broad IS achievement are associated with better performance in terms of both nutrition determinants and outcomes’ support, and cross-sector support, compared with projects with no IS achievement. However, for cross-sector support the significant negative coefficient of the interaction term shows that this positive association is reduced in magnitude as the quantity of achieved IS indicators increases. All other controls in the model constant, if the quantity of achieved IS indicators is equal to the mean (approximately two indicators), projects with broad IS achievement perform approximately 33 percentage points better compared with projects with no IS achievement3 (Panel C, column 28). 320 Appendix I Multivariate Regression Analysis Hypothesis three: The results show evidence consistent with the literature. The clearest finding is that M&E matters for achievement rates and this is robust across specifications and choice of dependent variable. All else in the model being equal, a one standard deviation increase in the share of M&E factors in a project (16 percentage points) is associated with about a 6 percentage point reduction in overall achievement (Panel A, columns 1–4). This effect is a bit higher in magnitude for nutrition determinants and achievement of outcomes and in turn for cross-sector achievement (all columns in Panels B and C). Recall that M&E is one of the more frequent negative factors, so the results are also consistent with this fact. Project design matters for overall performance and nutrition determinant and outcome performance, showing robust results to changes in specification, though only with coefficients significant at the 10 percent level (Panel A, columns 1 and 2, and Panel B, columns 6–9, 11–12, and 14–15). All other controls in the model constant, a one standard deviation increase in the share of project design factors (21 percentage points) is associated with about a 4 percentage point increase in performance (coefficients are similar across performance measures). This result suggests that project design is important for better nutrition results. Similarly, there is some evidence that community-based implementation is positively associated with the three measures of performance, but it is not robust to changes in specification. The strongest and more statistically significant association is seen for cross-sector support. A one standard deviation increase in the share of community- implementation factors (18 percentage points) is associated with about an 8 percentage point increase in cross-sector indicator achievement, all else in the model constant (Panel C, column 17). There is no evidence that country ownership and institutional arrangement factors matter for project performance when controlling for other factors and irrespective of the performance measure. Interestingly, the share of World Bank systems and performance factors in a project is positively and significantly associated with cross-sector performance with results robust to changes in specification. In its largest estimated effect, a one standard deviation increase in the share of this factor (11 percentage points) is associated with an 8 percentage point increase in the cross-sector achievement rate (Panel C, column 24). Recall that this factor captures internal World Bank processes affecting project implementation, including adequacy of financing, timeliness of disbursements, procurement, quality of supervision, and quality of team composition. This finding is consistent with the literature on the importance of staffing for project performance. For example, Hussein et al. (2018) find that predicted practice manager and TTL performance during the second half of supervision has a strong influence. 321 Appendix I Multivariate Regression Analysis Hypothesis four: There is no evidence supporting the hypothesis that investing in effective nutrition interventions as per the global literature is positively associated with project performance. This suggests that having the right interventions is not enough— implementation factors also matter. Hypothesis five: Similarly, there is little evidence supporting a positive association between a higher percentage of nutrition interventions that match a country’s needs and project performance. Although the associated coefficient is positive and significant in the bivariate analysis, the results do not hold in the multivariate regressions, not even in the more parsimonious specifications (Columns 1 and 2 in Panel A, 5 and 6 in Panel B, and 17 and 18 in Panel C). A higher percentage of nutrition interventions that match a country’s needs is not significantly associated with better nutrition results once other controls are considered. Hypothesis six: There is strong evidence in favor of hypothesis six: noncore projects are, on average, worse performers than core projects (those with “nutrition” or “stunted growth” in the title or PDO and a nutrition content share in the top 40 percent of the distribution) in terms of cross-sector performance. On average they perform 13 percentage points below core projects, all other controls kept constant (Panel C, columns 19–22, 24, 26, and 28). There is no evidence of a statistically significant difference in the effects of noncore and core projects on the achievement rates either overall or for nutrition determinants. Hypothesis seven: There is strong and robust evidence that projects that are supported with analytical inputs perform better on all three measures of performance, especially cross-sector achievement, compared with those without analytical support. In its largest effect, on average and holding all other controls in the model constant, projects with analytical support perform about 17 percentage points higher in terms of cross-sector achievement than projects with no support (Panel C, column 26). Results of cross-sector achievement are robust across specifications. Finally, there are other interesting findings in addition to the original hypotheses. First, the nature of indicators included in the results frameworks of projects matter for project performance. The higher the share of nutrition outcome indicators included, the lower the project performance. This is consistent with the finding that nutrition outcome indicators (such as anthropometric measures and micronutrient status) are more challenging to achieve during the project cycle, and hence with the predominance of project objectives focusing on nutrition determinants. Second, emergency projects perform better than nonemergency projects in terms of cross-sector achievement. Third, there are also some regional differences in performance, and non-FCS countries perform better than FCS countries in terms of overall and cross-sector support. Finally, better 322 Appendix I Multivariate Regression Analysis World Bank systems and performance (disbursements, team composition, quality of supervision) matter for improved project performance in terms of cross-sector support, which is consistent with findings in the literature. References Bulman, D., W. Kolkman, A. Kraay, A. 2015. “Good Countries or Good Projects? Comparing Macro and Micro Correlates of World Bank and Asian Development Bank Project Performance.” World Bank Policy Research Working Paper 7245. Denizer, C., D. Kaufmann, and A. Kraay. 2013. “Good countries or good projects? Macro and Micro correlates of World Bank Project Performance.” Journal of Development Economics 105: 288‐302. Geli, P., A. Kraay, and H. Nobakht. 2014. “Predicting World Bank Project Outcome Ratings.” World Bank Policy Research Working Paper 7001. Hussein, M., T. Kenyon, and J. Friedman. 2018. “A New Look at Factors Driving Investment Project Performance,” Development Economics (DEC) Policy Research Talk, September. Washington, DC: World Bank. Raimondo, E. 2016, “What Difference Does Good Monitoring & Evaluation Make to World Bank Project Performance?” World Bank Policy Research Working Paper 7726. Notes 1Project performance was measured in three versions, including (i) overall nutrition-related indicator achievement (nutrition outcomes, immediate determinants, underlying determinants, social norms, and institutional strengthening); (ii) nutrition determinant and outcomes indicator achievement (nutrition outcomes; immediate determinants, underlying determinants and social norms); and (iii) cross-sector indicator achievement (immediate and underlying determinants in sectors different from the project’s leading GP, in addition to social norms. For example, in a project led by the Social Protection and Jobs GP, this measure would exclude achievement of social safety net indicators). 2Hussein et al. 2018 define good project design with higher quality of the results framework, lower number of components, and lower number of intermediate indicators in a project, while Raimondo 2016 shows several aspects of good quality of M&E, including the simplicity of the M&E framework and its degree of alignment with the client’s M&E system, a clear institutional set-up in relation to M&E, and good integration with operational tasks. ∂y 3 ( ) = β2 + β5 × 2 = 0.810 − 0.241 × 2 = 0.328 ∂broadISachieved=2 323 Annex I.1 Table AI.1.1 Descriptive Statistics—Project-Level Variables Variable Mean SD Min Max Obs. Share of all indicators achieved 0.671 0.287 0.000 1.000 131 Share of nutrition determinants and outcomes’ indicators achieved 0.648 0.301 0.000 1.000 131 Share of cross-sector indicators achieved 0.657 0.347 0.000 1.000 131 No. of institutional strengthening (IS) indicators achieved-All 1.652 1.616 0.000 8.000 131 No. of IS indicators achieved – Stakeholder engagement and ownership 0.535 0.677 0.000 4.000 131 No. of IS indicators achieved – Policy, financing and coordination 0.423 0.554 0.000 3.000 131 No. of IS indicators achieved – Improving nutrition service delivery 0.887 1.016 0.000 6.000 131 Project multidimensionality score 0.230 0.139 0.000 0.625 131 Project w/ Broad IS achievement (3 IS sublevels achieved) (dummy 0–1) 0.053 0.226 0.000 1.000 131 Project w/ Medium IS achievement (2 IS sublevels achieved) (dummy 0–1) 0.183 0.388 0.000 1.000 131 Project w/ Narrow IS achievement (1 IS sublevel achieved) (dummy 0–1) 0.260 0.440 0.000 1.000 131 Project w/ No IS achievement (0 IS sublevels achieved) (dummy 0–1) 0.504 0.502 0.000 1.000 131 Country portfolio multidimensionality score 0.621 0.209 0.000 1.000 131 Country portfolio multidimensionality – first quartile (dummy 0–1) 0.359 0.481 0.000 1.000 131 Country portfolio multidimensionality – second quartile (dummy 0–1) 0.221 0.417 0.000 1.000 131 Country portfolio multidimensionality – third quartile (dummy 0–1) 0.237 0.427 0.000 1.000 131 Country portfolio multidimensionality – fourth quartile (dummy 0–1) 0.183 0.388 0.000 1.000 131 Non-core nutrition project (dummy 0–1) 0.710 0.456 0.000 1.000 131 Matching score between nutrition interventions and country needs 0.859 0.180 0.250 1.000 131 Emergency projects (dummy 0–1) 0.282 0.452 0.000 1.000 131 Analytical support (dummy 0–1) 0.450 0.499 0.000 1.000 131 Share of nutrition outcome indicators in project 0.039 0.095 0.000 0.500 131 Project duration (years) 5.762 2.434 0.512 13.285 131 324 Appendix I Multivariate Regression Analysis Non-World Bank Group nutrition-relevant foreign aid per 1,000 population 0.047 0.074 0.001 0.567 131 (US$, millions) Share of intervention outcomes with positive evidence in the literature 0.185 0.217 0.000 1.000 131 Share of factors with positive direction 0.618 0.316 0.000 1.000 131 Share of project design factors 0.183 0.209 0.000 1.000 131 Share of community-implementation factors 0.108 0.175 0.000 1.000 131 Share of country ownership and institutional arrangements factors 0.159 0.189 0.000 1.000 131 Share of World Bank systems and performance factors 0.064 0.105 0.000 0.500 131 Share of M&E factors 0.119 0.161 0.000 1.000 131 Share of country context factors 0.087 0.212 0.000 1.000 131 AFR 0.466 0.501 0.000 1.000 131 SAR 0.130 0.337 0.000 1.000 131 ECA 0.053 0.226 0.000 1.000 131 EAP 0.092 0.290 0.000 1.000 131 MENA 0.061 0.240 0.000 1.000 131 LCR 0.198 0.400 0.000 1.000 131 Low-Income country 0.656 0.477 0.000 1.000 131 Lower-Middle-Income country 0.266 0.443 0.000 1.000 131 Upper-Middle-Income country 0.063 0.243 0.000 1.000 131 Non-FCS country 0.740 0.440 0.000 1.000 131 Approval period 1998–08 0.366 0.484 0.000 1.000 131 Approval period 2009–13 0.504 0.502 0.000 1.000 131 Approval period 2014–19 0.130 0.337 0.000 1.000 131 Source: Independent Evaluation Group portfolio review and analysis. Note: Three regional projects are excluded from income-level statistics. 325 Appendix I Multivariate Regression Analysis Table AI.1.2 Nutrition Performance and its Predictors—Bivariate OLS Regressions Dependent Variables Share Of Nutrition Determinants And Share Of All Outcomes’ Indicators Indicators Share Of Cross-sector Independent Variables Achieved Achieved Indicators Achieved Matching score between nutrition interventions 0.190 0.142 0.284** and country needs (0.151) (0.159) (0.140) Non-core project (dummy 0–1) -0.0504 -0.0418 -0.0898 (0.0537) (0.0566) (0.0662) Project multidimensionality score 0.189 0.145 0.141 (0.165) (0.189) (0.221) Country portfolio multidimensionality score 0.231* 0.203 0.0848 (0.117) (0.125) (0.148) Analytical support (dummy 0–1) 0.0615 0.0443 0.139** (0.0499) (0.0527) (0.0600) Emergency projects (dummy 0–1) 0.0347 0.0360 0.123* (0.0544) (0.0578) (0.0640) No. of IS indicators achieved-All - 0.0353** 0.0337 - (0.0142) (0.0227) No. of IS indicators achieved – Policy, financing - 0.0999** 0.170*** and coordination - (0.0474) (0.0482) No. of IS indicators achieved – Improving nutrition - 0.167* 0.227 service delivery - (0.0953) (0.148) No. of IS indicators achieved – Stakeholder - 0.0738** 0.0165 engagement and ownership - (0.0367) (0.0454) - 0.204*** 0.162 326 Appendix I Multivariate Regression Analysis Project w/ Broad IS achievement (3 IS sublevels - (0.0735) (0.1369) achieved) (dummy 0–1) Project w/ Medium IS achievement (2 IS sublevels - 0.113* 0.136* achieved) (dummy 0–1) - (0.0632) (0.0745) Project w/ Narrow IS achievement (1 IS sublevel - 0.003 0.008 achieved) (dummy 0–1) - (0.0584) (0.0682) Share of project design factors 0.111 0.136 0.0269 (0.123) (0.131) (0.125) Share of community-implementation factors 0.332*** 0.327** 0.346** (0.126) (0.145) (0.146) Share of M&E factors -0.280** -0.394*** -0.415** (0.115) (0.126) (0.180) Share of country ownership and institutional 0.0333 -0.00354 0.150 arrangements factors (0.136) (0.146) (0.144) Share of intervention outcomes with positive -0.0899 -0.0587 -0.0387 evidence in the literature (0.134) (0.140) (0.0998) Source: Independent Evaluation Group portfolio review and analysis. Note: Table shows coefficients resulting from bivariate regressions between each of the dependent variables and each of the independent variables; no controls are included. Observations are always 131 projects. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Table AI.1.3 Nutrition Project Performance and its Predictors—Multivariate OLS Regressions a. Dependent variable is the share of all indicators achieved Variables (1) (2) (3) (4) Country portfolio multidimensionality – second quartile 0.0634 0.0295 -0.0305 -0.0535 (dummy 0–1) (0.0742) (0.0875) (0.0912) (0.0985) Country portfolio multidimensionality – third quartile 0.106 0.102 0.0299 0.0154 (dummy 0–1) (0.0717) (0.0795) (0.0826) (0.0873) Country portfolio multidimensionality – fourth quartile 0.136* 0.0826 0.0490 0.0352 (dummy 0–1) (0.0734) (0.0754) (0.0767) (0.0783) 327 Appendix I Multivariate Regression Analysis Project multidimensionality score -0.153 -0.187 -0.0791 -0.0488 (0.188) (0.192) (0.185) (0.187) Non-core project (dummy 0–1) -0.0359 -0.0479 -0.0311 -0.0369 (0.0565) (0.0579) (0.0569) (0.0586) Matching score between nutrition interventions and country -0.109 0.0564 0.0417 0.0583 needs (0.162) (0.206) (0.195) (0.207) Share of intervention outcomes with positive evidence in the -0.102 -0.129 -0.130 -0.127 literature (0.148) (0.160) (0.164) (0.168) Share of project design factors 0.222* 0.205* 0.189 0.213 (0.118) (0.116) (0.117) (0.132) Share of community-implementation factors 0.281* 0.234 0.198 0.188 (0.150) (0.157) (0.150) (0.159) Share of country ownership and institutional arrangements 0.0496 0.0129 -0.00720 -0.00547 factors (0.147) (0.153) (0.153) (0.158) Share of M&E factors -0.415*** -0.462*** -0.434*** -0.444*** (0.154) (0.158) (0.151) (0.156) Share of country context factors -0.0688 -0.0630 0.0231 0.0263 (0.166) (0.160) (0.167) (0.169) Share of World Bank systems and performance factors 0.216 0.218 0.280 0.355 (0.227) (0.223) (0.223) (0.253) Analytical support (dummy 0–1) 0.116** 0.114** 0.103* 0.0912 (0.0544) (0.0558) (0.0546) (0.0566) Emergency projects (dummy 0–1) 0.0293 0.00528 -0.00385 0.00962 (0.0566) (0.0591) (0.0567) (0.0625) Share of nutrition outcome indicators in project -0.627*** -0.568** -0.638*** -0.635*** (0.212) (0.233) (0.225) (0.231) Project duration (years) -0.0118 -0.0104 -0.00936 -0.0123 (0.0117) (0.0128) (0.0123) (0.0128) 328 Appendix I Multivariate Regression Analysis Non-World Bank Group nutrition-relevant foreign aid per -0.542* -0.642* -0.403 -0.517 1,000 population (US$, millions) (0.306) (0.328) (0.344) (0.368) Share of factors with positive direction 0.133 0.136 0.134 0.149 (0.0914) (0.0935) (0.0894) (0.0903) Approval period 2009–13 0.0108 0.0396 0.0582 0.0400 (0.0608) (0.0623) (0.0646) (0.0683) Approval period 2014–19 0.160** 0.162** 0.185** 0.204** (0.0801) (0.0799) (0.0793) (0.0961) EAP -0.0180 -0.0354 -0.0259 (0.0871) (0.0894) (0.0979) ECA 0.0647 0.0952 0.0987 (0.155) (0.143) (0.151) LCR -0.140* -0.144* -0.142 (0.0806) (0.0797) (0.128) MENA -0.110 -0.0594 -0.00920 (0.130) (0.137) (0.154) SAR -0.126* -0.103 -0.0996 (0.0710) (0.0680) (0.0725) Non-FCS country 0.162** 0.168** (0.0715) (0.0832) Lower-Middle-Income country -0.0235 (0.0953) Upper-Middle-Income country 0.0340 (0.159) Constant 0.691*** 0.642** 0.506** 0.507* (0.200) (0.253) (0.253) (0.263) Observations 131 131 131 128 329 Appendix I Multivariate Regression Analysis R-squared 0.284 0.320 0.351 0.359 b. Dependent variable is the share of nutrition determinants and outcomes’ indicators achieved Variables (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) Country portfolio 0.0780 0.0754 0.0595 0.0581 0.0247 0.0213 -0.0179 -0.0232 0.0870 0.0642 0.0830 0.0638 multidimensionality – second quartile (0.0707) (0.0753) (0.0841) (0.0879) (0.0862) (0.0903) (0.0900) (0.0931) (0.0709) (0.0836) (0.0734) (0.0868) (dummy 0–1) Country portfolio 0.0881 0.0856 0.101 0.0985 0.0583 0.0531 0.0280 0.0212 0.0762 0.0848 0.0859 0.100 multidimensionality – (0.0753) (0.0767) (0.0814) (0.0832) (0.0826) (0.0849) (0.0849) (0.0864) (0.0759) (0.0813) (0.0809) (0.0838) third quartile (dummy 0–1) Country portfolio 0.147** 0.140* 0.0868 0.0852 0.0669 0.0644 0.0437 0.0399 0.133* 0.0738 0.144* 0.0916 multidimensionality – (0.0740) (0.0750) (0.0795) (0.0826) (0.0805) (0.0834) (0.0826) (0.0867) (0.0745) (0.0807) (0.0765) (0.0829) fourth quartile (dummy 0–1) Project -0.192 -0.205 -0.253 -0.269 -0.187 -0.202 -0.179 -0.189 -0.194 -0.253 -0.246 -0.313 multidimensionality (0.214) (0.221) (0.219) (0.223) (0.217) (0.221) (0.209) (0.214) (0.213) (0.219) (0.212) (0.225) score No. of IS indicators 0.0105 -0.00548 -0.00356 -0.00356 0.0764 0.0521 0.0160 -0.0343 achieved-All (0.0158) (0.0175) (0.0178) (0.0179) (0.0513) (0.0487) (0.110) (0.108) No. of IS indicators -0.00987 -0.00864 achieved squared-All (0.00659) (0.00626) No. of IS indicators 0.0506 0.0329 0.0370 0.0404 achieved – Policy, (0.0389) (0.0385) (0.0381) (0.0402) financing and coordination No. of IS indicators 0.00171 -0.0177 -0.0185 -0.0159 achieved – Improving (0.0270) (0.0279) (0.0283) (0.0283) nutrition service delivery No. of IS indicators 0.0165 0.0101 0.0160 0.0118 achieved – (0.0444) (0.0464) (0.0454) (0.0452) Stakeholder 330 Appendix I Multivariate Regression Analysis engagement and ownership Non-core project -0.0200 -0.0212 -0.0494 -0.0521 -0.0375 -0.0407 -0.0474 -0.0484 -0.00765 -0.0384 -0.0139 -0.0370 (dummy 0–1) (0.0631) (0.0673) (0.0659) (0.0699) (0.0657) (0.0697) (0.0680) (0.0725) (0.0649) (0.0678) (0.0663) (0.0699) Matching score -0.126 -0.141 0.0429 0.0180 0.0296 0.000594 0.0725 0.0464 -0.109 0.0806 -0.145 0.0102 between nutrition (0.160) (0.167) (0.215) (0.221) (0.214) (0.219) (0.220) (0.226) (0.163) (0.213) (0.163) (0.229) interventions and country needs Share of intervention -0.0753 -0.0740 -0.127 -0.125 -0.124 -0.124 -0.117 -0.112 -0.0834 -0.133 -0.0961 -0.151 outcomes with (0.150) (0.151) (0.167) (0.167) (0.171) (0.171) (0.176) (0.176) (0.147) (0.165) (0.154) (0.169) positive evidence in the literature Share of project 0.214* 0.203* 0.229* 0.215* 0.218* 0.202 0.264* 0.245* 0.204 0.217* 0.185 0.195 design factors (0.122) (0.121) (0.122) (0.121) (0.123) (0.123) (0.135) (0.137) (0.124) (0.122) (0.124) (0.122) Share of community- 0.292* 0.260 0.267 0.229 0.242 0.196 0.257 0.215 0.268* 0.252 0.225 0.182 implementation (0.161) (0.180) (0.171) (0.190) (0.169) (0.188) (0.173) (0.192) (0.161) (0.170) (0.181) (0.193) factors Share of country 0.0391 0.0173 0.00768 -0.00825 -0.00366 -0.0225 0.0188 -0.00109 0.0243 -0.000426 -0.0208 -0.0729 ownership and (0.154) (0.154) (0.165) (0.168) (0.165) (0.168) (0.170) (0.172) (0.150) (0.163) (0.154) (0.173) institutional arrangements factors Share of M&E factors -0.474*** -0.484*** -0.544*** -0.547*** -0.523*** -0.527*** -0.530*** -0.533*** -0.496*** -0.564*** -0.551*** -0.643*** (0.171) (0.175) (0.184) (0.190) (0.184) (0.191) (0.185) (0.192) (0.169) (0.183) (0.185) (0.200) Share of country 0.0194 0.00741 0.0305 0.0135 0.0831 0.0663 0.105 0.0901 0.0191 0.0235 -0.0271 -0.0186 context factors (0.184) (0.184) (0.182) (0.182) (0.196) (0.196) (0.201) (0.201) (0.181) (0.181) (0.186) (0.183) Share of World Bank 0.149 0.172 0.159 0.181 0.194 0.222 0.298 0.328 0.150 0.162 0.161 0.167 systems and (0.242) (0.251) (0.246) (0.255) (0.251) (0.259) (0.273) (0.280) (0.245) (0.250) (0.263) (0.264) performance factors Analytical support 0.113** 0.111** 0.122** 0.120** 0.115** 0.113** 0.101* 0.0972 0.113** 0.123** 0.116** 0.121** (dummy 0–1) (0.0542) (0.0555) (0.0569) (0.0575) (0.0563) (0.0567) (0.0581) (0.0586) (0.0541) (0.0571) (0.0566) (0.0587) 0.0390 0.0373 0.0208 0.0180 0.0157 0.0123 0.0305 0.0271 0.0323 0.0149 0.0348 0.0172 331 Appendix I Multivariate Regression Analysis Emergency projects (0.0594) (0.0597) (0.0619) (0.0627) (0.0606) (0.0614) (0.0656) (0.0666) (0.0603) (0.0625) (0.0601) (0.0622) (dummy 0–1) Share of nutrition -0.693*** -0.668*** -0.655*** -0.641** -0.692*** -0.682** -0.706*** -0.690** -0.692*** -0.659*** -0.657*** -0.606** outcome indicators in (0.220) (0.224) (0.249) (0.256) (0.260) (0.269) (0.259) (0.270) (0.220) (0.248) (0.233) (0.275) project Project duration -0.0205* -0.0212* -0.0169 -0.0178 -0.0165 -0.0173 -0.0194 -0.0204 -0.0217** -0.0182 -0.0259** -0.0223* (years) (0.0109) (0.0111) (0.0126) (0.0128) (0.0126) (0.0128) (0.0132) (0.0134) (0.0109) (0.0126) (0.0111) (0.0130) Non-World Bank -0.336 -0.323 -0.514 -0.494 -0.367 -0.338 -0.422 -0.388 -0.253 -0.417 -0.251 -0.469 Group nutrition- (0.320) (0.325) (0.340) (0.341) (0.358) (0.358) (0.380) (0.384) (0.335) (0.355) (0.377) (0.396) relevant foreign aid per 1,000 population (US$, millions) Share of factors with 0.160 0.160* 0.159 0.155 0.159 0.152 0.184* 0.180* 0.162 0.165 0.135 0.132 positive direction (0.0971) (0.0961) (0.100) (0.0987) (0.0987) (0.0967) (0.101) (0.0996) (0.0980) (0.101) (0.0996) (0.102) Approval period 2009– -0.0145 -0.0144 0.0190 0.0163 0.0297 0.0271 0.00599 0.00357 -0.0190 0.0115 -0.0167 0.0221 13 (0.0624) (0.0617) (0.0652) (0.0655) (0.0682) (0.0681) (0.0721) (0.0719) (0.0625) (0.0658) (0.0647) (0.0679) Approval period 2014– 0.0897 0.0843 0.114 0.101 0.125 0.110 0.137 0.126 0.0692 0.0960 0.0831 0.118 19 (0.0919) (0.0924) (0.0920) (0.0947) (0.0924) (0.0952) (0.109) (0.110) (0.0956) (0.0956) (0.105) (0.103) EAP 0.0417 0.0305 0.0297 0.0146 0.0363 0.0242 0.0495 0.0195 (0.0976) (0.103) (0.103) (0.108) (0.110) (0.114) (0.0987) (0.0998) ECA 0.0850 0.0861 0.0974 0.0993 0.112 0.115 0.109 0.0896 (0.169) (0.176) (0.164) (0.171) (0.170) (0.177) (0.176) (0.171) LCR -0.158* -0.154* -0.159* -0.155* -0.206 -0.199 -0.151* -0.168* (0.0899) (0.0882) (0.0902) (0.0884) (0.147) (0.151) (0.0887) (0.0922) MENA -0.151 -0.142 -0.120 -0.109 -0.0677 -0.0523 -0.126 -0.187 (0.128) (0.127) (0.129) (0.128) (0.157) (0.159) (0.124) (0.128) SAR -0.104 -0.107 -0.0891 -0.0930 -0.0928 -0.0926 -0.103 -0.120 (0.0785) (0.0811) (0.0768) (0.0792) (0.0830) (0.0846) (0.0794) (0.0800) Non-FCS country 0.0965 0.103 0.106 0.113 (0.0734) (0.0743) (0.0862) (0.0873) 332 Appendix I Multivariate Regression Analysis Lower-Middle-Income -0.00944 -0.0112 country (0.107) (0.112) Upper-Middle-Income 0.174 0.174 country (0.166) (0.173) Project w/ Broad IS 0.411** 0.464** achievement (3 IS (0.198) (0.221) sublevels achieved) (dummy 0–1) Project w/ Medium IS 0.217 0.170 achievement (2 IS (0.194) (0.189) sublevels achieved) (dummy 0–1) Project w/ Narrow IS 0.0425 0.0875 achievement (1 IS (0.144) (0.150) sublevel achieved) (dummy 0–1) Project w/ Broad IS -0.0608 -0.0380 achievement × No. of (0.112) (0.112) IS indicators achieved- All Project w/ Medium IS -0.0561 -0.0146 achievement × No. of (0.120) (0.117) IS indicators achieved- All Project w/ Narrow IS 0.0344 0.0326 achievement × No. of (0.130) (0.130) IS indicators achieved- All Constant 0.710*** 0.729*** 0.673** 0.709*** 0.591** 0.630** 0.564** 0.593** 0.654*** 0.600** 0.776*** 0.782** (0.218) (0.235) (0.258) (0.268) (0.268) (0.274) (0.277) (0.282) (0.229) (0.268) (0.259) (0.311) Observations 131 131 131 131 131 131 128 128 131 131 131 131 333 Appendix I Multivariate Regression Analysis R-squared 0.284 0.291 0.322 0.326 0.332 0.337 0.351 0.356 0.297 0.332 0.314 0.357 c. Dependent variable is the share of cross-sector indicators achieved Variables (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) Country portfolio 0.0737 0.0481 0.0111 -0.00866 -0.0453 -0.0671 -0.0490 -0.0692 0.0919 0.0208 0.0727 -0.0106 multidimensionality – second quartile (0.0855) (0.0876) (0.0905) (0.0960) (0.0966) (0.103) (0.104) (0.107) (0.0838) (0.0865) (0.0889) (dummy 0–1) (0.0907) Country portfolio 0.0546 0.0448 0.0950 0.0811 0.0266 0.00898 0.00173 -0.0148 0.0305 0.0619 0.0290 0.0496 multidimensionality – third quartile (dummy (0.0831) (0.0803) (0.0880) (0.0870) (0.0922) (0.0929) (0.101) (0.0996) (0.0837) (0.0825) (0.0926) 0–1) (0.0895) Country portfolio 0.0678 0.0488 -0.0625 -0.0752 -0.0946 -0.108 -0.0958 -0.111 0.0383 -0.0895 0.0466 -0.0897 multidimensionality – fourth quartile (0.0893) (0.0852) (0.0914) (0.0901) (0.0928) (0.0919) (0.0916) (0.0898) (0.0946) (0.0919) (0.102) (dummy 0–1) (0.0992) Project -0.351 -0.377 -0.441* -0.473* -0.333 -0.367 -0.308 -0.341 -0.356 -0.441* -0.347 -0.455 multidimensionality score (0.256) (0.261) (0.263) (0.265) (0.261) (0.262) (0.267) (0.268) (0.252) (0.264) (0.261) (0.279) No. of IS indicators 0.0155 0.00181 0.00492 0.00681 0.149** 0.122** 0.0936 0.117 achieved-All (0.0252) (0.0230) (0.0222) (0.0228) (0.0651) (0.0606) (0.103) (0.103) No. of IS indicators -0.0200** -0.0181** achieved squared-All (0.00886) (0.00769) No. of IS indicators 0.125*** 0.0964** 0.103** 0.104** achieved – Policy, (0.0466) (0.0431) (0.0440) (0.0451) financing and coordination No. of IS indicators -0.00939 -0.0284 -0.0298 -0.0280 achieved – Improving (0.0333) (0.0300) (0.0288) (0.0298) 334 Appendix I Multivariate Regression Analysis nutrition service delivery No. of IS indicators -0.0157 -0.00738 0.00198 0.00517 achieved – (0.0568) (0.0502) (0.0463) (0.0478) Stakeholder engagement and ownership Non-core project -0.0968 -0.106 -0.150** -0.162** -0.131* -0.144* -0.125 -0.137* -0.0717 -0.127* -0.0768 -0.138* (dummy 0–1) (0.0739) (0.0771) (0.0724) (0.0739) (0.0708) (0.0726) (0.0755) (0.0780) (0.0728) (0.0729) (0.0755) (0.0743) Matching score 0.0952 0.0486 0.147 0.141 0.126 0.113 0.124 0.110 0.131 0.226 0.127 0.234 between nutrition interventions and (0.139) (0.130) (0.213) (0.216) (0.216) (0.219) (0.229) (0.227) (0.141) (0.206) (0.171) country needs (0.235) Share of intervention 0.0791 0.0766 0.0869 0.0784 0.0910 0.0799 0.105 0.0947 0.0627 0.0732 0.0560 0.0451 outcomes with positive evidence in (0.134) (0.133) (0.131) (0.129) (0.139) (0.137) (0.146) (0.144) (0.132) (0.128) (0.144) the literature (0.136) Share of project 0.139 0.119 0.181 0.156 0.163 0.136 0.185 0.153 0.117 0.155 0.113 0.142 design factors (0.128) (0.132) (0.120) (0.123) (0.121) (0.124) (0.140) (0.143) (0.129) (0.121) (0.133) (0.123) Share of community- 0.471*** 0.433** 0.313* 0.275 0.272 0.223 0.292 0.239 0.422** 0.282 0.411** 0.253 implementation factors (0.168) (0.171) (0.182) (0.193) (0.178) (0.190) (0.191) (0.201) (0.164) (0.183) (0.175) (0.196) Share of country 0.169 0.0981 0.0803 0.0326 0.0619 0.00984 0.0589 0.00441 0.139 0.0633 0.144 0.0576 ownership and institutional (0.166) (0.174) (0.162) (0.169) (0.162) (0.170) (0.172) (0.181) (0.164) (0.165) (0.172) arrangements factors (0.175) 335 Appendix I Multivariate Regression Analysis Share of M&E factors -0.388* -0.441** -0.525** -0.565** -0.492** -0.533** -0.476** -0.517** -0.432* -0.567** -0.434* -0.607** (0.222) (0.219) (0.225) (0.226) (0.215) (0.218) (0.222) (0.226) (0.219) (0.227) (0.230) (0.236) Share of country 0.160 0.124 -0.0283 -0.0675 0.0567 0.0164 0.0833 0.0400 0.159 -0.0430 0.149 -0.0622 context factors (0.202) (0.203) (0.199) (0.195) (0.202) (0.197) (0.205) (0.201) (0.195) (0.196) (0.201) (0.199) Share of World Bank 0.505* 0.568* 0.552* 0.613** 0.610* 0.679** 0.693* 0.765** 0.506 0.560* 0.489 0.521 systems and performance factors (0.303) (0.307) (0.297) (0.300) (0.308) (0.309) (0.353) (0.354) (0.311) (0.306) (0.327) (0.319) Analytical support 0.165*** 0.153** 0.170*** 0.163*** 0.159*** 0.152** 0.158** 0.150** 0.165*** 0.171*** 0.158** 0.164*** (dummy 0–1) (0.0586) (0.0608) (0.0584) (0.0586) (0.0603) (0.0600) (0.0636) (0.0631) (0.0590) (0.0590) (0.0611) (0.0604) Emergency projects 0.163** 0.154** 0.154** 0.145** 0.146** 0.136** 0.156** 0.147** 0.150** 0.142** 0.150** 0.136* (dummy 0–1) (0.0658) (0.0677) (0.0669) (0.0684) (0.0664) (0.0677) (0.0719) (0.0728) (0.0673) (0.0669) (0.0710) (0.0704) Share of nutrition -0.824*** -0.772*** -0.632* -0.607* -0.692** -0.671* -0.726** -0.705* -0.822*** -0.639* -0.810*** -0.637* outcome indicators in project (0.248) (0.262) (0.330) (0.338) (0.334) (0.343) (0.354) (0.364) (0.249) (0.323) (0.263) (0.346) Project duration -0.00787 -0.00898 -0.0103 -0.0112 -0.00967 -0.0105 -0.0113 -0.0123 -0.0104 -0.0129 -0.0105 -0.0135 (years) (0.0124) (0.0124) (0.0130) (0.0130) (0.0127) (0.0127) (0.0132) (0.0132) (0.0119) (0.0127) (0.0129) (0.0132) Non-World Bank -0.0709 -0.0965 -0.726** -0.694** -0.488 -0.445 -0.269 -0.244 0.0976 -0.522 0.0542 -0.577 Group nutrition- relevant foreign aid (0.351) (0.344) (0.329) (0.333) (0.358) (0.359) (0.514) (0.509) (0.375) (0.345) (0.407) (0.375) 336 Appendix I Multivariate Regression Analysis per 1,000 population (US$, millions) Share of factors with 0.172 0.175 0.138 0.141 0.137 0.136 0.145 0.145 0.176 0.149 0.167 0.131 positive direction (0.112) (0.117) (0.111) (0.113) (0.105) (0.108) (0.106) (0.110) (0.111) (0.109) (0.115) (0.114) Approval period 2009– -0.0785 -0.0691 -0.0300 -0.0300 -0.0129 -0.0129 -0.0213 -0.0230 -0.0876 -0.0457 -0.0708 -0.0169 13 (0.0713) (0.0692) (0.0666) (0.0655) (0.0682) (0.0669) (0.0731) (0.0720) (0.0709) (0.0661) (0.0750) (0.0699) Approval period 2014– -0.0937 -0.0712 -0.0557 -0.0488 -0.0381 -0.0337 -0.0161 -0.0156 -0.135 -0.0930 -0.110 -0.0598 19 (0.112) (0.119) (0.104) (0.113) (0.102) (0.111) (0.117) (0.123) (0.112) (0.107) (0.121) (0.112) EAP 0.0485 0.0552 0.0292 0.0300 0.0488 0.0448 0.0648 0.0561 (0.110) (0.112) (0.102) (0.102) (0.109) (0.107) (0.107) (0.110) ECA -0.250 -0.193 -0.230 -0.172 -0.214 -0.159 -0.200 -0.243 (0.184) (0.189) (0.175) (0.181) (0.184) (0.187) (0.151) (0.179) LCR -0.323*** -0.313*** -0.324*** -0.315*** -0.281** -0.283** -0.307*** -0.322*** (0.0894) (0.0883) (0.0900) (0.0890) (0.136) (0.138) (0.0880) (0.0949) MENA -0.0558 -0.0271 -0.00567 0.0248 0.0481 0.0717 -0.00267 -0.0237 (0.121) (0.123) (0.127) (0.129) (0.159) (0.160) (0.114) (0.122) SAR -0.221* -0.220* -0.196* -0.198* -0.178 -0.183* -0.217* -0.245** 337 Appendix I Multivariate Regression Analysis (0.115) (0.111) (0.112) (0.109) (0.112) (0.110) (0.118) (0.121) Non-FCS country 0.156* 0.164* 0.194* 0.199* (0.0935) (0.0919) (0.108) (0.106) Lower-Middle-Income -0.0522 -0.0422 country (0.106) (0.113) Upper-Middle-Income 0.000773 0.0116 country (0.196) (0.202) Project w/ Broad IS 0.782** 0.810*** achievement (3 IS sublevels achieved) (0.350) (dummy 0–1) (0.248) Project w/ Medium IS 0.207 0.201 achievement (2 IS sublevels achieved) (0.205) (dummy 0–1) (0.182) Project w/ Narrow IS 0.0723 0.132 achievement (1 IS sublevel achieved) (0.170) (dummy 0–1) (0.183) Project w/ Broad IS -0.205 -0.241** achievement × No. of IS indicators achieved- (0.127) (0.106) All Project w/ Medium IS -0.0993 -0.136 achievement × No. of IS indicators achieved- (0.111) (0.108) All -0.0443 -0.0935 338 Appendix I Multivariate Regression Analysis Project w/ Narrow IS (0.137) (0.141) achievement × No. of IS indicators achieved- All Constant 0.445* 0.527** 0.695** 0.738** 0.563** 0.612** 0.508* 0.563* 0.330 0.542* 0.369 0.580* (0.246) (0.249) (0.284) (0.286) (0.280) (0.277) (0.291) (0.286) (0.262) (0.287) (0.312) (0.321) Observations 131 131 131 131 131 131 128 128 131 131 131 131 R-squared 0.292 0.319 0.389 0.407 0.408 0.429 0.409 0.430 0.331 0.419 0.323 0.423 Source: Independent Evaluation Group portfolio review and analysis. Note. In the table, the base region is AFR; the base approval period is 1998–2008; the base for the country portfolio multidimensionality quartiles is the first quartile; the base for the institutional strengthening (IS) achievement intensity levels (narrow, medium, broad) is projects with zero IS intensity (projects with no IS sublevels achieved); non-core projects are compared against core projects, defined as those with the words “nutri” or “stunt” in their title or PDO and having a share of nutrition content in the top 40 percent of the distribution. Table excludes 14 countries that had no closed nutrition projects (Nigeria, Haiti, Zimbabwe, Mali, Congo Rep., Côte d’Ivoire, Comoros, Philippines, Liberia, Bhutan, Armenia, Lesotho, Guinea-Bissau, and Marshall Islands). Three regional projects are excluded from columns that include income-level regressors. Robust standard errors in parentheses. *** p<0.01 ** p<0.05 * p<0.1. 339 The World Bank 1818 H Street NW Washington, DC 20433