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 interventions 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 countries. 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 interventions in a country portfolio can be accomplished through multidimensional projects that deliver a broad set of nutrition interventions or through a multisectoral approach by coordinating and integrating nutrition interventions 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 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 improving 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 opportunities 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 multisectoral 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 nutrition outcomes through supporting nutrition determinants, social norms, 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 design 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 arrangements 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 government. At the local level, institutional strengthening has been important (for example, in Rwanda and Senegal) to operationalize the delivery of interventions (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 important to achieve expected results from multisectoral nutrition approaches.

• The quality and extent of subnational targeting of multisectoral interventions 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 communities. Recent approaches to improve the convergence of interventions and needs in countries, such as Indonesia and Rwanda, are promising. Multidimensional 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 other 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 promoting 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 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 achievements toward nutrition determinants. The evaluation consistently identifies M&E of nutrition indicators as a pathway to improve 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 community level and those that focus on the institutional strengthening of actors involved in delivering services, since the goal of behavior change interventions 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 interventions 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 evaluation’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 nutrition-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 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 protection 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 nutrition-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 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 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, Madagascar, 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 implementation of multisectoral nutrition interventions and (ii) increase focus on subnational targeting of interventions to reflect areas of greatest disadvantage 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.