Back to cover

World Bank Support to Reducing Child Undernutrition

Chapter 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 framework, with a mix of interventions, including nutrition-specific and nutrition-sensitive approaches. Institutional strengthening accounts 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 evidence 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 use of evidence in nutrition programming.

Global Practices often collaborate in implementing nutrition interventions, but case studies suggest that they need support in learning 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 enhancement of national leadership and subnational governments to coordinate multisectoral actions. Second, it can support the organization 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 relevant interventions to address the disaggregated needs or priorities of countries. Recent efforts to improve the alignment of interventions with community needs are promising.

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 Development financing, and $5.8 billion in RETFs, to support reducing child undernutrition. 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, supports 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 Protection 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 institutional strengthening include support to develop institutional capacities 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 determinants 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 continue to be a major support area, and agricultural approaches (for example, home gardens, livestock production, and food fortification), and to a lesser extent social safety nets, receive increased attention. Still, the portfolio continues to have important gaps.

The World Bank’s nutrition portfolio also supports nutrition-specific interventions, 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 breastfeeding 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 interventions 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 determinants 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).

Figure 2.1. Nutrition Interventions in the Portfolio

Image

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.

Figure 2.1. Nutrition Interventions in the Portfolio

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 childhood development, awareness campaigns, positive deviation modeling, breastfeeding, child feeding promotion and counseling, accompanying measures of conditional cash transfers

Health Services

  • Health and nutrition promotion and counseling, information, education, and communication 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

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 gender roles, such as those related to decision-making regarding the care of children, and social and cultural practices that may influence the nutrition status of children and pregnant and lactating women, the focus on women’s empowerment, 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

Image

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.

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 Bank Doing the Right Thing?

The evidence on what works to reduce child undernutrition and improve nutrition 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 identifying 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 nutrition-sensitive interventions is especially importance, since the list of interventions 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 interventions targeting children, most of the evidence studied the effects of providing supplementary energy-dense foods, followed by zinc supplementation, supplementary 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 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, 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 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.

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 iodine 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 for improving child use of health services, family planning and contraception services are effective to reduce birth spacing, and health system strengthening support shows positive effects in improving complementary feeding practices and household welfare. In agriculture, food fortification with vitamin 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 household 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

Image

Sources: Independent Evaluation Group; systematic review map.

Note: The legend combines the size of the evidence (number of systematic reviews) and the direction 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 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.

Figure 2.3. Effective Interventions to Improve Stunted and Linear Growth

Sources: Independent Evaluation Group; systematic review map.

Note: The legend combines the size of the evidence (number of systematic reviews) and the direction 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 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

Figure 2.4. Alignment of Nutrition Interventions with Evidence on What Works, by Intervention Area

Image

Sources: Independent Evaluation Group; systematic review map, and portfolio review and analysis.

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 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.

Figure 2.4. Alignment of Nutrition Interventions with Evidence on What Works, by Intervention Area

Sources: Independent Evaluation Group; systematic review map, and portfolio review and analysis.

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 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.

In health, the World Bank concentrates on supporting SBCC on nutrition and health practices known to work across different nutrition-relevant outcome areas, 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 prominent in the portfolio within the group of interventions with consistent positive 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 effective to improve complementary feeding and child health outcomes.

The portfolio includes effective WASH interventions, such as SBCC and community water supply. According to the literature on what works, effective interventions 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 effectiveness 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 evidence 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 areas (table 2.1). Although the World Bank emphasizes many of these interventions, some of them may not be receiving sufficient attention given their potential 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 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. 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 suggests that 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.

Impact evaluations of World Bank projects contribute to increasing knowledge of what works by supporting evidence-based learning to design and improve 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 facilitate the mainstreaming of interventions or experiences leveraged from the project support. In this way impact evaluations can support the institutionalization 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 rollout of nutrition interventions of the community-based nutrition (CBN) program. 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 evaluations 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 programs, 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.

Table 2.1. Systematic Review Map Interventions with a Broad Positive Impact

Intervention Types

Interventions (%)

Projects (no.)

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–folic acid)

0.9

15

SBCC of nutrition and health promotion (via community and groups, education, growth monitoring and promotion, home visits, mass communication, and IPC at health facility)

21.5

107

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: 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 training

2.3

34

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

Sources: Independent Evaluation Group; systematic review map and portfolio review and analysis.

Note: CCT = conditional cash transfer; CHW = community health worker; IPC = interpersonal communication; 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 behavior 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 different GPs to implement interventions toward nutrition determinants. The portfolio is also multidimensional in that it includes a range of different interventions across the nutrition-specific and nutrition-sensitive dimensions of the conceptual framework. HNP leads most projects in the nutrition portfolio (42 percent), Agriculture leads about 21 percent, and SPJ leads 20 percent. 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, Education, 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 integrated 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 nutrition 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

Image

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.

Figure 2.5. Projects by Approval Period and Global Practice

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.

Figure 2.6. Interventions in Projects by Global Practice

Image

Sources: Independent Evaluation Group; portfolio review and analysis.

Note: WASH = water, sanitation, and hygiene.

Figure 2.6. Interventions in Projects by Global Practice

Image

Sources: Independent Evaluation Group; portfolio review and analysis.

Note: WASH = water, sanitation, and hygiene.

Figure 2.7. Projects by Global Practice and Degree of Multidimensionality

Image

Source: Independent Evaluation Group.

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.

Figure 2.7. Projects by Global Practice and Degree of Multidimensionality

Source: Independent Evaluation Group.

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 interventions (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 determinants, 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.

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 interventions 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 interventions, 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.) Countries 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 countries 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 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 nutrition led by different GPs and instrument types (investment project financing, development policy loan, Program-for-Results, RETF) over the 10-year evaluation period. However, in most countries these interventions are fragmented 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 supported 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 welfare 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 improve 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.

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 approach is well established, but support to converge services across sectors is newer. In Ethiopia, support to ECD, maternal diet intake, women’s empowerment, 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 interventions 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 develop 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, maternal health, home gardens, safety nets, and other support, is ongoing.

Figure 2.8. Ethiopia Project Time for World Bank Nutrition Support

Image

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 = Agriculture; 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.

Figure 2.8. Ethiopia Project Time for World Bank Nutrition Support

Image

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 = Agriculture; 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.

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 strengthening 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, districts, and communities, and this has helped improve the accountability of nutrition 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 services), with emerging examples of how projects can strengthen multisectoral arrangements for nutrition. The stocktaking analysis of countries identified 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 delivery by initiating village scorecards and child scorecards and developing the interoperability of sectoral M&E systems. In some countries (Ethiopia, Indonesia, 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, agriculture). In Indonesia, Malawi, Rwanda, and Senegal, the World Bank has 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 within 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 coordination 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 integrated 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, actions, 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 protection, 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 other 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 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 having 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.

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 estimates (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 better nutrition outcomes (no stunted growth, no wasting, no underweight, no anemia, and no LBW) at the beginning and at the end of the evaluation period (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 interventions in health synergized with multidimensional interventions across determinants to improve outcomes.7 This synergized support has not consistently 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 related to disadvantaged determinants (low levels of food and care, WASH, and health services) can help countries to catch up to improve outcomes. Correlation 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 determinants. Conversely, improvements in nutrition outcomes to benefit vulnerable 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 determinants, particularly with regard to social norms.8 A mapping exercise using 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 intervention. The World Bank support has been especially relevant in addressing 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 association with country nutrition outcomes according to the heat map analysis. 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.

Figure 2.9. Undernutrition Determinants and Nutrition Outcomes

Image

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 2.9. Undernutrition Determinants and Nutrition Outcomes

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

Image

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.

Figure 2.10. Alignment of Portfolio Interventions with Country Needs

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. Projects 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 fragmented (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 protection, 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 interventions to promote health, WASH, caregiving, and nutritious foods have been supported by different projects and implemented in different communities.

  1. 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 portfolio 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 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.