Assessing public financing for nutrition in Bhutan, Nepal and Sri Lanka

Abstract

The objective of this study was to assess public financing for nutrition in Bhutan, Nepal and Sri Lanka to identify limitations of available data and to discuss policy implications. A variant of the Scaling Up Nutrition Movement methodology was used. Budget allocations and expenditures for relevant government ministries during 2012–2018 were identified. Nutrition‐related line items were tagged using definitions of nutrition‐specific and nutrition‐sensitive interventions. Data were aggregated by year and calculated in constant United States dollars (USD). Expenditures by year were presented as a proportion of gross domestic product and general government expenditures. The percent utilization of budget allocations and proportion of funding from central government sources were determined. Per capita expenditures on nutrition‐specific interventions varied from USD 1.08–8.76 and for nutrition‐sensitive interventions varied from USD 20.22–51.20. Nutrition‐specific expenditures as a percent of gross domestic product ranged from 0.08% in Sri Lanka in 2017% to 0.34% in Nepal in 2016. The median utilization rate was 64% for nutrition‐specific and 84% for nutrition‐sensitive interventions. Nutrition‐specific funding financed by the central government was 90.7% in Bhutan and 99.4% in Sri Lanka. This study revealed the need to prioritize and invest in evidence‐based interventions, including balancing investments in nutrition‐specific versus ‐sensitive interventions. Challenges in estimation of nutrition expenditures and cross‐country comparison were also observed, highlighting the need for appropriate nutrition line item tagging and standardized systems for data collection.

This study assesses the public financing for nutrition in Bhutan, Nepal and Sri Lanka during 2012–2018. It highlights the challenges and lessons learned from undertaking the analysis and discuss its policy implications.

Key messages

• Substantial variation in the region exists with respect to nutrition‐specific expenditures, albeit with an overall low base.

• The share of spending on nutrition‐specific versus nutrition‐sensitive interventions should be carefully considered in relation to the expected impact of the interventions being funded.

• Substantial limitations to track nutrition expenditures exist within current budget allocation records, such as a lack of cross‐ministerial tagging processes.

• Standardized global guidance is needed for defining nutrition‐sensitive and ‐specific interventions to enhance comparability between studies in different countries. This study aligned the classifications of interventions between the three countries for cross‐country comparison, which could provide indications for standardization.

1. INTRODUCTION

Despite rapid economic growth in recent decades, South Asian countries continue to face a large burden of malnutrition. In 2019, an estimated 56.1 million children in South Asia were stunted (33% in the region ranging from 15% in Maldives to 38% in India and Pakistan), which places children at higher risk for infectious disease morbidity and delays in cognitive development (Caulfield et al., 2004; Sudfeld et al., 2015; UNICEF WHO & World Bank Group, 2020). Fifteen percent (25.1 million) were wasted, with wasted children experiencing mortality rates over twice that of well‐nourished children (McDonald et al., 2013; UNICEF et al., 2020). Maternal anaemia, thinness and short stature are common in South Asia and are critical risk factors for low birth weight, neonatal mortality, stunting and wasting among children (Harding et al., 2018; Kim et al., 2017; Rahman et al., 2016). There is considerable variability between countries in the region with respect to nutrition outcomes. For example, the prevalence of stunting in Sri Lanka (17%) is half that of Nepal (32%) (Central Bureau of Statistics—Government of Nepal & UNICEF, 2020; Department of Census and Statistics & Ministry of Health, Nutrition and Indigenous Medicine, 2017). There is also a growing burden of overweight and obesity in the region, which raises the risk of cardiovascular disease, diabetes and cancer (Guh et al., 2009; World Health Organization, 2020). The prevalence of overweight and obesity among adults was estimated to vary between countries from 19% to 29% in 2016, but all countries have seen substantial growth (61%–88% increase from 2000 to 2016) (World Health Organization, 2020). Taken together, nutritional risk factors are the largest single contributor to the overall burden of morbidity and mortality in the region (Institute for Health Metrics and Evaluation, 2018). A summary of country‐specific nutrition‐related indicators in South Asia are provided in Appendix A.

Beyond the direct impacts on subsequent health outcomes, nutrition is also an important driver of economic productivity. Quasi‐experimental research studies from global literature find that stunting is associated with reduced earnings; a 1 cm decrease in height among adults is associated with a 4% decrease in wages for men and a 6% decrease for women (McGovern et al., 2017). There is limited evidence from randomized trials, but one study of a nutritional supplementation intervention during gestation and early childhood found that adult wages were 46% higher among men provided with improved nutrition (Martorell et al., 2010). While estimates of the cost–benefit ratio for different nutritional interventions vary widely, a review of studies found a median return value of 17.9 for each currency unit invested (Hoddinott et al., 2013).

The second Sustainable Development Goal aims to ‘end hunger, achieve food security and improved nutrition, and promote sustainable agriculture’, and 12 of the 17 Goals contain indicators relevant to nutrition. In alignment with these aims, many South Asian countries have developed national strategies or plans to address malnutrition (Ministry of Agriculture and Forestry, 2014; Ministry of Healthcare and Nutrition, 2010; National Planning Commission & MoH, 2018). These strategies or plans typically lay out a broad range of interventions to prevent and treat malnutrition. Strategies and plans generally include nutrition‐specific interventions, such as the nutritional care of pregnant women, the protection, promotion and support of infant and young child feeding practices and treatment of acute malnutrition. Nutrition‐sensitive interventions are also often included, such as interventions to improve water and sanitation, agricultural outputs, or family income support measures.

Once strategies or plans have been adopted, it is critical that progress be monitored to ensure that targets are reached and adjustments in approach are made as necessary. Assessments of public financing for nutrition are a critical component of any monitoring strategy, as they allow insight into whether investments are aligned with the interventions and population groups prioritized by the national nutrition strategy or plan, as well as adequate for achieving stated objectives.

However, assessing nutrition financing can be challenging because multiple ministries and sectors are often responsible for nutrition‐related activities. For example, interventions to improve dietary diversity may be linked to agricultural production. Even within ministries of health, nutrition activities (such as iron and folic acid supplementation) may be bundled as part of broader health service packages (such as pre‐natal care). To address these challenges, the Scaling Up Nutrition (SUN) Movement methodology for assessing nutrition financing has been developed and implemented in several low and middle‐income countries (Scaling Up Nutrition Movement, 2020). The objective of this study is two‐fold: to summarize the findings from three country studies that assessed the public financing for nutrition in Bhutan, Nepal and Sri Lanka; and highlight challenges and lessons learned from undertaking the analysis with recommendations of potential solutions and policy implications.

2. METHODS

This study builds on three previous country‐specific studies that analysed budget allocation and expenditure data for nutrition‐specific and nutrition‐sensitive interventions in Bhutan (Ahmed et al., 2020), Nepal (Medici et al., 2019) and Sri Lanka (Jayawardena et al., 2020). The analysis aimed to determine: the amount of budget allocated to nutrition interventions, the proportion of allocations that were expended, distribution of funding sources (i.e., central vs. lower‐level government) and the specific interventions which were funded. A summary of nutrition policies in the three countries are provided in Appendix B.

A variant of the SUN Movement methodology was used for tracking nutrition investments (Development Initiatives, 2017; Scaling Up Nutrition Movement, 2020). The methodology comprises three general steps: identification of nutrition budget line items (programs and interventions); categorization of expenditures; and weighting of expenditures (this last step is optional).

In the first step, documents of nutrition‐related strategies and plans for each country were reviewed to identify government ministries, and departments within those ministries, having a role in the delivery of nutrition interventions. Next, expenditure data sources were identified. In Bhutan, line‐item budget allocations were obtained from five relevant ministries for the years 2014–2017 (see Appendix C for list of relevant ministries). In Nepal, the Ministry of Finance Red Book—the official record of budget allocation—was obtained, along with expenditure data by donor‐funded development projects through the Aid Management Platform (AMP) website. Expenditure data from ten ministries were obtained for the years 2012–2018 (with the transition to federalism in Nepal in 2018, federal level budget allocation and expenditure data were available, but not at provincial and local levels). In Sri Lanka, a working group was formed among nine government agencies to obtain expenditure information, including donor spending when available, for the years 2014–2018. As there is no comprehensive account system in place that tracks expenditure data for nutrition, an Excel‐based template was used to manually collect data from ministries, provinces and departments. Sources of data from each country are provided in Appendix D.

In the second step, budgets and expenditure data were reviewed for key terms to identify relevant line items (see Appendix E for a sample list of terms). Data were then extracted into a standardized spreadsheet which contained information on the ministry, department, program, activity, budget allocation amount, expenditure amount, financing source and type of intervention. The type of intervention was classified as nutrition‐specific or nutrition‐sensitive following prior recommendations (Bhuttata et al., 2008; Ruel & Alderman, 2013). A complete list of nutrition‐specific and nutrition‐sensitive activities can be found in Appendix F. Capital assets that were shared between multiple ministries (such as warehouses) were excluded. For Bhutan and Sri Lanka, human resource budget allocations within the relevant ministries that supported a variety of activities in addition to nutrition assigned a proportion of the budgeted amount to nutrition activities based on consultation with governments counterparts in each country. When budget allocations were available for both central and lower‐level government bodies, records were checked to ensure no duplicate extraction of expenditures.

The final step of weighting expenditures is described as optional in the SUN Movement methodology. Given the subjective nature of weight estimation and the potential noncomparability this may create between countries, no weighting was performed in this analysis.

Nutrition‐specific and nutrition‐sensitive data were aggregated by year and calculated in constant United States dollars (USD) for the year 2018 using currency conversion rates from country central banks or the International Monetary Fund. Budget allocations by year were presented as a proportion of gross domestic product (GDP) and general government expenditure (GGE), which were extracted from government reports. The proportion of budget allocations that were expended are summarized, as are the distribution of funding sources between central and lower‐level government sources. Finally, the proportion of expenditures dedicated to different program activities were presented.

3. RESULTS

3.1. Nutrition‐related expenditures over time

Nutrition‐specific expenditures varied considerably within countries over time and between countries (Table 1). Bhutan recorded expenditures of USD 2.96 per capita in 2014, which grew by 268% to USD 7.94 in 2015 and further to USD 8.76 in 2017 due to changes in procurement for the Ministry of Education's school feeding program. Nutrition‐specific expenditures in Nepal were low relative to other countries, though a substantial increase was observed in the year following a major earthquake in 2015. Sri Lanka's nutrition‐specific expenditures have remained relatively constant, ranging from USD 3.23 to 3.90 per capita.

Table 1.

Nutrition‐specific expenditures in three South Asian countries (2018 US$ per capita)

	2012	2013	2014	2015	2016	2017	2018
Bhutan	‐	‐	2.96	7.94	7.82	8.76	‐
Nepala	1.45	1.08	1.25	1.62	2.91	1.71	1.27
Sri Lanka	‐	‐	3.45	3.90	3.48	3.23	3.79

^a 2018 data include only federal level data.

Nutrition‐sensitive expenditures were substantially higher (without weighting) than nutrition‐specific expenditures (Appendix G). In 2017 in Bhutan, when nutrition‐specific expenditures were at their peak, nutrition‐sensitive spending was 2.6 times higher than that for nutrition‐specific. Even wider discrepancies were seen for Nepal (10.7 times higher than the nutrition‐specific spending peak in 2016) and Sri Lanka (13.1 times higher in 2015). In Bhutan and Nepal, spending on nutrition‐sensitive interventions varied less from year to year as compared with nutrition‐specific interventions.

3.2. Relative nutrition‐specific expenditures

Nutrition‐specific expenditures as a percent of GDP ranged from 0.08% in Sri Lanka in 2017 to 0.34% in Nepal in 2016 (Figure 1). While substantial fluctuations of expenditures as a percent of GDP were observed for 1 or 2 years in Bhutan and Nepal, they remained generally constant in Sri Lanka. More fluctuation was seen in expenditures as a percent of GGE. Expenditures rose for Bhutan and fluctuated up and down for Nepal. Expenditures as a percent of GGE ranged from 0.42% in Sri Lanka in 2017 to 1.7% in Bhutan in 2017.

Figure 1.

Nutrition‐specific intervention expenditures as a percent of gross domestic product (a) and percent of general government expenditure (b)

3.3. Utilization of resources and level of decentralization

Governments typically did not spend the full amount that was allocated to nutrition‐specific interventions (Table 2). Bhutan spent more than two‐thirds of their allocations since 2015, while Nepal spent less than two‐thirds for 6 out of 7 years. The mean utilization rate for nutrition‐sensitive interventions was 80% in Bhutan and 85% in Nepal (Appendix H). The utilization rates for nutrition‐sensitive interventions were higher than for nutrition‐specific interventions, with Nepal showing a notably large disparity. No data was available for Sri Lanka on utilization. Central government spending was dominant for nutrition‐specific interventions in Bhutan (90.7%) and Sri Lanka (99.4%; Table 3).

Table 2.

Percent utilization of nutrition‐specific allocations in two South Asian countries ^a

	2012	2013	2014	2015	2016	2017	2018
Bhutan	‐	‐	62	83	77	76	‐
Nepalb	62	64	38	29	70	64	41

^a Sri Lanka is excluded due to incomplete data on utilization.

^b 2018 data in Nepal include only federal level data.

Table 3.

Central‐government versus lower‐level government funding distribution for nutrition‐specific interventions (% of total) ^a

	Central	Lower‐levelb
Bhutan	90.7	9.3
Sri Lanka	99.4	0.6

^a Funding distribution for the most recent year of available data: 2017 in Bhutan, 2018 in Sri Lanka. No data available for Nepal.

^b District and block in Bhutan, province in Sri Lanka.

3.4. Type of interventions

Nutrition‐specific intervention spending is heavily concentrated within a few programs (Table 4). In Bhutan, spending on the School Health and Nutrition Division by the Ministry of Education represents 56% of all nutrition‐specific expenditures. In Nepal, 80% of spending originates from the Integrated Child Health and Nutrition Program. Expenditures in Sri Lanka are more diversified, with School Meal Programs representing 46%, the Thriposha program (a supplementation program for women in pregnancy and 6 months of lactation, and undernourished children under 5 years of age) representing 32%, and Family Health Bureau expenditures (primarily on micronutrient supplements) representing 14%.

Table 4.

Intervention‐specific funding allocations for top nutrition‐specific activities (data for most recent year available)

	US$ per capita	% of total
Bhutana
School Health & Nutrition	4.94	56.4
Early childhood care and development & special education	1.27	14.5
Vaccine Preventable Disease Program	0.45	5.1
Other	2.10	24.0
Nepalb
Integrated Child Health and Nutrition Program	1.01	79.5
Other	0.26	20.5
Sri Lankac
School meal programs	1.76	46.4
Thriposha program	1.21	31.9
Family Health Bureau medicine and supplements	0.54	14.2
Glass of milk for school children	0.15	4.0
Other	0.13	3.5

^a Programs in Bhutan include: Revitalize Baby Friendly Hospital Initiative; micronutrition supplementation; strengthen infant and young child feeding; growth monitoring; acute malnutrition detection and referral; acute malnutrition treatment; immunization and integrated management of childhood illnesses, care for child development; school supplementation program (weekly iron and folic acid, deworming, vitamin A); monitoring of health and diet of school children; develop minimum nutrition package guidelines for adolescent girls; develop and implement advocacy material for social mobilization; develop pre‐conception nutrition service package; strengthening antenatal care and post‐natal care; develop nutrition counselling toolkit; early childhood care and development; healthy meals and hygiene; rice fortification; fortified food in boarding schools; review of school feeding program; mid‐day meal in all schools; nutrition and food safety training for teachers and students; improve access to minimum nutrition package for adolescent girls; build capacity of service providers.

^b Programs in Nepal include: iron folic acid supplementation; multiple micronutrient supplementation; balanced energy protein supplementation; calcium supplementation; exclusive breastfeeding; complementary feeding; vitamin A capsule and deworming tablets supplementation (6–59 months); oral rehydration solution and preventive zinc supplementation; management of severe acute malnutrition. management of moderate acute malnutrition; awareness on the use of iodized salt; scale‐up Integrated Management of Newborn and Childhood Illness program; School Health and Nutrition Program revitalization, including promotion of healthy dietary and physical activities; conduct maternal, infant and young child nutrition counselling; conduct nutrition promotion events, including iodine month, breastfeeding week, etc.; establishment of nutrition information and surveillance system.

^c Programs in Sri Lanka include: micronutrient supplementation programs; multivitamin drops, iron, folic acid for preterm and low birth weight children; therapeutic food (BP 100); zinc supplementation for children with diarrhoea; deworming treatment; Thriposha program; promote optimal maternal nutrition and infant and young child feeding (IYCF) practices; capacity building of relevant health staff in maternal nutrition and IYCF counselling; growth monitoring and promotion; information, education and communication for breastfeeding and IYCF; District Nutrition Action Plan for vulnerable population and estate sector; recommended instruments for maternal and child health clinics; morning meal/fresh milk for preschool children; Poshana Manpetha (food and nutrition awareness program); school meal program; glass of milk for school children; Food for Education Program; Tikiri Shakthi (high‐protein nutrient bar).

4. DISCUSSION

4.1. Allocation amounts for nutrition interventions

This study assessed financing for nutrition interventions in Bhutan, Nepal and Sri Lanka during the period 2012 to 2018. It found that expenditures by a government on nutrition‐specific interventions tended to remain relatively stable over time, with a few short‐term increases and a notable increase in Bhutan. Expenditures varied widely between countries on the basis of spending per capita, as a percent of GDP, and as a percent of GGE. Expenditures increased considerably only in Bhutan as a proportion of GGE, indicating a higher priority placed on nutrition in the country. Expenditures on nutrition‐sensitive interventions were 2.6–13.1 times higher than those for nutrition‐specific interventions, due to the fact that some of these interventions are quite costly and the entirety of the expenditure was counted as nutrition‐related (e.g., expenditures on agriculture) and multiple sectors are involved. Further, interventions counted as ‘nutrition‐sensitive’ are so comprehensive in nature and their impacts on nutrition outcomes are incompletely documented, that they may in fact be overcounted/overestimated. For example, much of the literature often counts all investments in food security/agriculture as ‘nutrition‐sensitive’ investments. Yet both the objectives and the impacts on nutrition outcomes of interest are often unclear.

For all countries, nutrition‐related utilization was less than the amount allocated. It may be difficult to convince government officials that increased spending on nutritional interventions is needed when the currently allocated amounts are not being spent, and thus bottlenecks for utilization must be identified and addressed.

In an analysis of 24 countries across multiple years, nutrition‐specific expenditures as a percent of GDPs varied by over 1000‐fold, but over half of data points fell between 0.01% and 1% (MQSUN, 2016). In this study, the variability in nutrition‐specific expenditures as a percent of GGE is approximately four‐fold between countries and years (from 0.42% to 1.70%). This higher degree of similarity between countries in this study compared with the other study may have to do with regional similarities in prioritizing nutrition investments, methodological similarities, or simply with the limited number of countries examined. Nonetheless, variations do exist between the three countries. Per capita nutrition‐specific expenditure in Bhutan has been US$8–9 except since 2015, US$1–2 in Nepal and around US$3.5 in Sri Lanka. However, it should be noted that there is no benchmark as to what constitutes an adequate level of expenditure, which depends on multiple factors such as overall disease burden or nutritional status of population. For instance, Sri Lanka has the lowest share of nutrition‐specific expenditure as percentage of GGE, but the rate of stunting is also lowest among the countries. Therefore, the level of expenditures should be interpreted with caution.

4.2. Limitations of current budget tagging procedures

The data collection process for this study revealed limitations in how nutrition‐related line items were tagged in budget allocations and expenditure records. In all countries, there was no pre‐existing budget tagging mechanism to identify nutrition activities. As a result, nutrition line‐items had to be identified manually and in consultation with government officials from the relevant ministries. Furthermore, nutrition activities were often bundled along with other interventions in health service packages such as antenatal care, which required additional investigation to determine the amount for a line‐item related to nutrition. This is further complicated by the fact that packages of interventions listed within a given line item might change from year to year.

In Bhutan, Nepal and Sri Lanka, the restructuring of line‐ministries (such as merging or splitting of ministries) that typically occurs when a new government takes office added further difficulty in extracting nutrition expenditure data. While nutrition‐specific interventions are primarily managed by Ministries of Health or Education that are typically less affected, nutrition‐sensitive interventions under other ministries are more subject to this concern. In Sri Lanka for instance, data on the nutrition program implemented under the Ministry of Livestock and Rural Community Development was not retrievable before 2015 as it was dissolved and integrated into the Ministry of Rural Economy in 2015.

To address these challenges, a cross‐ministry coordinated system of line‐item tagging for nutrition expenditures could be introduced, as has been implemented in Indonesia (Satriawan, 2019). This system would include sufficient disaggregation so that line‐items are uniquely related to nutrition and are disaggregated with respect to the intervention type and population. This would allow quick extraction of nutrition‐related expenditures and easy analysis, thereby facilitating more frequent review of nutrition data by governments. Such a system would need to be underpinned by sufficient technical capacity among ministry personnel, which would itself require investment and coordination. Yet, for countries with such burden of malnutrition, this would be an important step forward.

4.3. Limitations in the analytical approach

Weighting of nutrition‐sensitive interventions is controversial in two aspects. The first is a lack of documented evidence on nutrition outcomes and therefore how to assign weights to each intervention. Nepal convened a workshop of stakeholders to assign weights while Bhutan and Sri Lanka did not attempt to assign weights. Second, not assigning weights means that nutrition‐sensitive interventions dominate the nutrition expenditure as shown in our findings. In addition, the high figures on nutrition‐sensitive expenditure could potentially provide a misleading information that the government is investing sufficient funds in nutrition or that nutrition‐sensitive interventions predominantly contribute to nutritional improvement. Therefore, expenditure on nutrition‐specific and sensitive interventions should be analysed separately rather than as a whole. For this reason, standard classification between the two types of interventions becomes even more critical.

4.4. Comparability with prior studies of nutrition financing

It is important to note that the comparability of nutrition expenditure reviews across countries is limited by differences in the method used to conduct the assessment. Although the SUN methodology provides a general structure, case‐by‐case decisions are required when examining the budget allocation of a particular government. For example, the inclusion of capital assets (particularly those shared between multiple programs, such as vehicles or warehouses) or personnel costs (for staff who perform both nutrition‐ and non‐nutrition‐related activities) may be handled differently between research teams. For example, some teams may choose to exclude these expenditures, whereas others may choose to weight them. The weighting process typically relies on subjective decisions that could reasonably vary between teams. Furthermore, some studies may use different definitions of nutrition‐specific and nutrition‐sensitive activities. A typical example includes school feeding where debates had been observed even within the same country. While Nepal classified it as nutrition‐sensitive, Bhutan and Sri Lanka treated it as nutrition‐specific intervention. In other cases, some countries have included interventions that were not considered nutrition programs in this study (e.g., goat/sheep/duck population development). Finally, whether a nutrition public expenditure review includes data from off‐budget donor expenditures may have a significant impact on the study's findings, depending on the share in the total expenditure. While Bhutan and Sri Lanka did not include off‐budget expenditures in the estimates, Nepal did from data provided in the AMP (although the share of off‐budget spending was only 3% of total nutrition‐related expenditure). It is recommended for all countries to maintain a database with detailed financing information from development aid. While the government may host such database, it requires full commitment from all development partners in maintaining such database.

In summary, comparisons of spending across countries should be interpreted with a high degree of caution. One way to address the issue of different definitions of nutrition programs is for studies to report results with both country‐specific classifications and a globally agreed set of definitions.

4.5. Investing in evidence‐based interventions

The identified nutrition interventions in the country studies were generally aligned with their respective nutrition policies and strategic plans, although policies tend to be described more broadly and are implemented in a range of interventions. In terms of their design, the nutrition‐specific strategies of Bhutan, Nepal and Sri Lanka contain similar interventions. Principally, these strategies include nutrition counselling and micronutrient supplements (Vitamin A, iron and folic acid) for women and children, the care of severely wasted children and school meal programs (preschool or elementary). In Sri Lanka pregnant/lactating women and children also receive food supplements as well as vouchers for purchasing healthy food through part of the pregnancy and lactation period.

Spending on nutrition‐specific interventions comprised 4%–28% of overall nutrition spending in Bhutan, Nepal and Sri Lanka. Evidence shows that nutrition‐specific interventions have more immediate impacts than nutrition‐sensitive interventions on reductions in stunting, wasting, micronutrient deficiencies, anaemia and mortality because they directly target children and women. This suggests that an increased emphasis on nutrition‐specific expenditures may be warranted. In particular, maternal nutrition counselling, including on age‐appropriate infant and child feeding, has been shown to be both effective in reducing stunting and is highly cost‐effective (Pearson et al., 2018). It is important for governments to monitor whether spending on nutrition is being allocated to the interventions which yield the greatest impact. Nutrition‐specific interventions should be complemented with nutrition‐sensitive interventions to address underlying and basic causes of malnutrition (Bhuttata et al., 2008). However, it should be noted that there is no clear benchmark of a recommended proportion between nutrition‐specific and sensitive interventions.

School feeding programs comprised 46% of spending in Sri Lanka and 56% in Bhutan. While school feeding programs can have important benefits such as increased school attendance and reduced anaemia, they are generally not effective measures to reduce stunting or its long‐term consequences, since most stunting develops in the first 2 years of life (when the impacts on brain development are likely to be greatest). Furthermore, if a large percentage of children under 5 years old are not attending preschools (where meals are served)—which would likely be the children from more disadvantaged backgrounds—then this intervention would miss much of the vulnerable population (Best et al., 2011; Snilstveit et al., 2015). Spending is most likely to impact nutritional status when it is targeted to age groups which are most likely to benefit, particularly during the first 1000 days of life (from pregnancy until the child is 2 years of age). Populations that are sensitive to nutrition interventions and will therefore benefit from impacts must be prioritized.

Food assistance needs to be carefully considered for its effectiveness. For instance, the food allowance program for pregnant mothers in Sri Lanka was made universal in 2015 (LKR 2,000 for 10 months), which had major implications on spending. Yet, it is the kind of intervention that focuses on the right age groups aiming to diversify the diet, and that likely has high impact on nutritional status. One option could be to expand the support for a longer period, for example, for the child's first 1000 days, coupled with nutrition counselling. Conversely, nutrition supplements, such as the Thriposha program for pregnant and lactating mothers in Sri Lanka, could benefit from better targeting to undernourished women given its high energy nature that could increase overweight and obesity otherwise.

4.6. Focusing interventions on high burden and vulnerable populations

A related consideration that was discussed in country studies is whether interventions are being focused on populations that need them the most (Ahmed et al., 2020; Medici et al., 2019; Jayawardena et al., 2020). In Nepal, allocations were based on the burden of malnutrition in a community; in 2017, per capita nutrition allocations in Karnali Province (where stunting prevalence is 54%) were approximately double the per‐capita allocations in Bagmati Province (where stunting prevalence is 29%). Unfortunately, tracking nutritional financing at lower administrative levels is generally limited in the countries studied, thereby impairing insight into how resources are matched to needs. Nepal introduced a federal system in 2015, with modified fiscal arrangements implemented by fiscal year 2017–2018, which resulted in a major change in how resources are allocated. Before federalism, all nutrition budget allocations were registered at the central level. After the transition, public budget allocations were decentralized to provincial and local governments, but these were not accurately recorded. While resource allocation and expenditure in Sri Lanka is highly centralized, the distribution of program resources to provinces or districts was not always recorded. Therefore, some central expenditures were allocated to provinces proportionally assuming the number of target population (e.g., by number of school children for school health programs). Bhutan faced similar issues where budgets are not channelled from central ministries to subnational levels for implementing activities, limiting the analyses at the subnational level (indirect expenditures at subnational levels were not estimated as access to internal ministry‐level data was also limited).

4.7. Benchmarking expenditure

Beyond simply monitoring expenditure amounts, it is necessary to assess how expenditures compare to their expected cost. To benchmark the level of spending needed for a population, costing exercises are warranted to estimate the required cost for delivering essential nutrition programs stated in the national policy. Comparing actual spending against the projected cost will help identify investment gaps (in cases of inadequate spending) and potential inefficiencies (in cases of excess spending). For instance, Ministry of Health, Nutrition and Indigenous Medicine in Sri Lanka (MoH) conducted a costing study on nutrition‐specific interventions for 2017‐2021 (MoH, 2017). After limiting the expenditure to MoH programs (i.e., removing school feeding, etc.), cost and expenditure exhibited similar levels for 2017 (LKR 4,866 million and LKR 4,574 million, respectively) but not for 2018 (LKR 5,186 and LKR 6,411, respectively) indicating a higher spending than was estimated. Such comparisons would be useful in raising questions about whether the differences are due to scaling up of interventions or reflecting less efficient resource use (in this case it was due to a change in the micronutrient supplement aiming to improve quality and compliance). Equally important is to compare spending level and nutrition outcomes. This is important for a subnational analysis to assess if resources are reaching where the demand is high. The government's account system should be capable of tracking subnational allocations of resources more accurately. Along this line, analysing the optimal mix of programs for each context is warranted using optimization algorithms such as those embedded in the Optima Nutrition tool (Pearson et al., 2018), although the tool requires extensive data, particularly unit costs and interventions are limited to nutrition‐specific programs.

5. CONCLUSIONS

Nutrition expenditures within countries tend to remain constant over time, with considerable differences between countries in terms of per capita expenditures. Regular review of nutrition budget allocations and expenditures can help governments assess whether resources are allocated in alignment with national nutrition strategies and plans. Appropriate nutrition line item tagging and developing standardized systems of data collection and assigning accountability for the conduct of reviews can help ensure that these occur regularly. Furthermore, prioritization exercises led by governments, and with inputs from technical advisors, can help ensure that expenditures are focused on the right interventions, with the required balance between nutrition‐sensitive and ‐specific interventions, for the right populations. Taken together, these actions can help ensure continued progress to reduce malnutrition.

DISCLAIMER

The findings, interpretations, and conclusions in this paper are entirely those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations or to members of its Board of Executive Directors or the countries they represent.

CONFLICT OF INTERESTS

The authors declare that there are no conflict of interests.

AUTHOR CONTRIBUTIONS

HH, DNC and AT conceptualized and designed the study. CTA prepared the first draft of the manuscript. JSC and MG extracted data. All authors provided comments and approved the final manuscript.

APPENDIX A. NUTRITION‐RELATED INDICATORS IN SOUTH ASIA

Indicator	AFG	BGD	BTN	IND	MDV	NPL	PAK	LKA
Stunting prevalence among children under 5 years of age (%height‐for‐age < −2 SD)	35.1 (2020)	30.2 (2020)	22.4 (2020)	30.9 (2020)	14.2 (2020)	30.4 (2020)	36.7 (2020)	16.0 (2020)
Prevalence of wasted children under 5 years of age (% weight‐for‐height < −2 SD)	5.1 (2018)	9.8 (2019)	5.9 (2010)	17.3 (2017)	9.1 (2017)	12.0 (2019)	7.1 (2018)	15.1 (2016)
Underweight prevalence among children under 5 years of age (% weight‐for‐age < −2 SD)	19.1 (2018)	22.6 (2019)	12.7 (2010)	33.4 (2017)	14.8 (2017)	24.4 (2019)	23.1 (2018)	20.5 (2016)
Overweight prevalence among children under 5 years of age (% weight‐for‐height > +2 SD)	3.9 (2020)	2.1 (2020)	5.2 (2020)	1.9 (2020)	4.6 (2020)	1.8 (2020)	3.4 (2020)	1.3 (2020)
Prevalence of overweight among adults, BMI ≥ 25 (age‐standardized estimate)	23.0 (2016)	20.0 (2016)	27.1 (2016)	19.7 (2016)	30.6 (2016)	21.0 (2016)	28.4 (2016)	23.3 (2016)
Prevalence of obesity among adults, BMI ≥30 (age‐standardized estimate)	5.5 (2016)	3.6 (2016)	6.4 (2016)	3.9 (2016)	8.6 (2016)	4.1 (2016)	8.6 (2016)	5.2 (2016)
Prevalence of anaemia in children aged 6–59 months (%)	44.9 (2019)	43.1 (2019)	44.7 (2019)	53.4 (2019)	37.8 (2019)	44.6 (2019)	53.0 (2019)	24.1 (2019)
Prevalence of anaemia in nonpregnant women(aged 15–49) (%)	43.2 (2019)	36.5 (2019)	38.7 (2019)	53.1 (2019)	52.3 (2019)	35.4 (2019)	41.1 (2019)	34.6 (2019)
Prevalence of anaemia in pregnant women (aged 15–49) (%)	36.5 (2019)	42.2 (2019)	38.2 (2019)	50.1 (2019)	49.3 (2019)	42.5 (2019)	44.0 (2019)	34.6 (2019)
Prevalence of anaemia in women of reproductive age (aged 15–49) (%)	42.6 (2019)	36.7 (2019)	38.6 (2019)	53.0 (2019)	52.2 (2019)	35.7 (2019)	41.3 (2019)	34.6 (2019)
Infants exclusively breastfed for the first 6 months of life (%)	57.5 (2018)	62.6 (2019)	53.2 (2015)	58.0 (2017)	63.0 (2017)	62.1 (2019)	47.8 (2018)	80.9 (2016)

Source: WHO Global Health Observatory, available https://www.who.int/data/gho/data/indicators/indicators‐index, accessed 29 November 2021; WHO Nutrition Landscape Information System, available https://www.who.int/teams/nutrition-and-food-safety/databases/nutrition-landscape-information-system, accessed 29 November 2021.

APPENDIX B. SUMMARY OF NUTRITION POLICY IN BHUTAN, NEPAL AND SRI LANKA

Bhutan: Bhutan's Food and Nutrition Security policy lays out the overarching nutrition strategy. The National Nutrition Task Force—comprising focal points from different ministries and development partners—coordinates and tracks nutrition‐related interventions across different sectors in Bhutan. Although the Food and Nutrition Security policy notes the role of multiple ministries in delivering on food and nutrition, two ministries are formally mentioned as having the primary mandate to do so: (i) the Ministry of Agriculture and Forests for food security policy and strategic action plans; and (ii) the Ministry of Health for leading and coordinating the implementation of nutrition security programs. The action plan developed by the National Nutrition Task Force aims to accelerate specific interventions targeting vulnerable groups—adolescent girls, under‐five children, women of reproductive age and pregnant/lactating women—to reduce micronutrient deficiencies and improve nutrition.

Nepal: Nutrition has been a national priority for over a decade, and since 2004, the government of Nepal has been actively involved in developing nutrition policy and guidelines. The National Nutrition Policy and Strategy 2004, led solely by the Ministry of Health and Population, was the first nutrition program. The Nutrition Assessment and Gap Analysis conducted in 2009/10 provided the basis for a multisector nutrition plan, enabling the involvement of other government stakeholders. The integrated approach to nutrition was supported by the development of MSNPs, which were designed to improve nutrition outcomes under the leadership of the National Planning Commission. A National Nutrition and Food Security Secretariat has been formed in the NPC to formulate policies on nutrition and food security.

Sri Lanka: The first nutrition policy in Sri Lanka was developed in 1986 and was followed by several revisions. The value of the policy, however, has diminished due to limited intersectoral coordination, changes in government priorities, and policies not being updated in line with socioeconomic and demographic changes. A National Nutrition Policy was developed in 2010 to address this situation, and a corresponding strategic plan and action plans have also been developed. The National Nutrition Policy has five objectives: (i) to ensure optimum nutrition throughout the life cycle; (ii) to enhance the capacity to deliver effective and appropriate interventions; (iii) to ensure effective management of adequate nutrition to vulnerable populations; (iv) to ensure food and nutrition to all citizens; and (v) to strengthen research, monitoring and evaluation. The National Nutrition Policy, which is in the process of revision for 2020–2030, reflects malnutrition in all stages of life, including presently neglected nutrition among elders. It is expected to place greater emphasis on multisectoral coordination and partnership while increasing the coverage of nutrition‐specific interventions at the primary health‐care level.

APPENDIX C. MINISTRIES WITH A ROLE IN DELIVERING NUTRITION‐SPECIFIC OR NUTRITION‐SENSITIVE INTERVENTIONS

Bhutan	Ministry of Agriculture and Forests Ministry of Education Ministry of Health Ministry of Home and Cultural Affairs Ministry of Works and Human Settlement
Nepal	Ministry of Agriculture and Livestock Development Ministry of Commerce and Supply Ministry of Drinking Water and Sanitation Ministry of Education Ministry of Federal Affairs and Local Development/Ministry of Federal Affairs and General Administration Ministry of Finance Ministry of Health and Population Ministry of Physical Planning and Works Ministry of Urban Development Ministry of Women, Children and Social Welfare
Sri Lanka	Ministry of Agriculture Ministry of City Planning and Water Supply—Water Sanitation Ministry of Education Ministry of Fisheries and Aquatic Resources Development Ministry of Health, Nutrition and Indigenous Medicine Ministry of Hill Country, New Villages, Infrastructure and Community Development Ministry of Livestock and Rural Community Development Ministry of Social Empowerment and Welfare Ministry of Women and Child Affairs

APPENDIX D. DATA SOURCES

Country	Years covered	Source of data	Notes
Bhutan	2014–2017	Ministry of Finance	Nutrition‐specific and nutrition‐sensitive items were mapped to MoF's listing of budgetary activities and sub‐activities.
Nepal	2012–2018	Ministry of Finance	On‐budget data were extracted from MoF's Red Book, while off‐budget data were acquired from the Aid Management Portal (AMP) website maintained by the MoF.
Sri Lanka	2014–2018	Key ministries and government agencies (see Appendix A)	A working group formed among the nine key ministries and government agencies conducted primary data collection on expenditure information, including donor spending when available, both at the central and provincial level.

APPENDIX E. KEY SEARCH TERMS FOR IDENTIFYING NUTRITION‐RELATED BUDGET ITEM

Health: maternal, neonatal, new born, child and adolescent health, diseases, hygiene, micro nutrients, iron supplementation/zinc supplementation, vitamins, vitamin A, feeding practices, nutrition, malnutrition, undernutrition, BMI, ANC/PNC, family planning/reproductive health, child immunization, health promotion, healthcare, health services, food safety, baby‐friendly

Agriculture: fruits and vegetables, nuts, legumes, pulses, livestock, fishery sources, extension services, cooperatives, seed, fertilizers, food, food security, hunger, agriculture production, rural development, food technology, nutrition technology, food safety, food quality, food fortification, agriculture

Education: early child education, pre‐primary education, basic education, girls' education, rural education, school feeding/school meals, school health, school WASH/hygiene/hand‐washing

Social protection: women, children, safety net programs, cash and voucher transfers, orphan and vulnerable children, insurance, welfare services, emergency, humanitarian, relief, maternity leave, pro‐poor

WASH: drinking water supplies, environment, sanitation, sewages, rural/urban areas, hygiene, toilets/latrines, WASH

APPENDIX F. CLASSIFICATION OF NUTRITION‐SPECIFIC AND NUTRITION‐SENSITIVE INTERVENTIONS

Category	Intervention
Nutrition‐specific interventions
Adolescent nutrition	Adolescent nutrition
Food and micronutrient supplementation and fortification in women	Peri‐conceptional folic acid supplementation
	Antenatal care interventions
	Iron and folic acid
	Micronutrient supplementation
	Balanced energy protein supplementation
	Calcium supplementation
	Iron and folic acid supplementation for non‐pregnant women
	Iodine through salt iodization
	Staple food fortification
Maternal counselling	Nutrition counselling for improved dietary intake during pregnancy
	Nutrition counselling for improved dietary intake during lactation after delivery
	Counselling and nutrition advice to women of reproductive age and mothers
Interventions during delivery and the neonatal period	Delayed cord clamping
	Early initiation of breastfeeding (with 1 h of birth)
	Kangaroo mother care for promotion of early and exclusive breastfeeding and care of preterm and small for gestational age infants
	Neonatal Vitamin K
	Appropriate infant feeding practices and ARV for HIV‐exposed infants
Breastfeeding, complementary feeding and dietary diversification	Breastfeeding promotion including exclusive breastfeeding for the first 6 months of life
	Implementation of the marketing of breast milk substitutes
	Appropriate complementary feeding promotion in children 6–24 months of age and continued breastfeeding
	Public provision of complementary food
	Early stimulation
	Dietary diversification
Micronutrient supplementation and fortification in infants and children	Iron supplementation for children 6–59 months
	Vitamin A supplementation for children 6–59 months
	Prophylactic zinc supplementation
	Zinc supplementation with oral rehydration in the treatment of diarrhoea
Management of SAM in children	Management of severe acute malnutrition
Management and prevention of disease	Antenatal care, including HIV testing and deworming for pregnant women
	Prevention and treatment of infectious disease; deworming for children
	Prevention and treatment and promotion of insecticide‐treated bed nets for pregnant women in high‐malaria areas
	Intermittent presumptive treatment of malaria in pregnancy in malaria‐endemic regions
Nutrition interventions in emergencies	Intervention package include management of severe acute malnutrition, prevention and health promotion strategies, such as breastfeeding and complementary feeding education and support
Nutrition‐sensitive interventions
Food security, access and availability	Improved availability, access and use of locally available foods
Family planning	Access to modern family planning services
Social safety nets	Social safety nets (e.g., conditional and unconditional cash transfer, food and in‐kind transfer program)
	Provision of healthy foods in schools
	Maternity protection in the workplace
Women empowerment	Promotion of increased age at marriage and reduced gender discrimination and gender‐based violence
Early childhood development	Parenting and life skills for early childhood development
	Early childhood development
WASH	Promotion of hand washing with soap and improved water and sanitation practices
Schooling	Increased access to primary and secondary education for girls
	Early childhood education
	Nutrition and physical education in school
	School Feeding
Maternal mental health	Maternal mental health
Others	Support for birth registration and strengthening of civil registration systems

APPENDIX G. NUTRITION‐SENSITIVE EXPENDITURES IN THREE SOUTH ASIAN COUNTRIES (2018 US$ PER CAPITA)

	2012	2013	2014	2015	2016	2017	2018
Bhutan	‐	‐	23.49	20.22	23.41	22.38	‐
Nepal a	24.86	21.28	28.84	27.01	31.10	32.75	23.63
Sri Lanka	‐	‐	31.15	51.20	49.14	46.96	35.83

^a2018 data include only federal level data.

APPENDIX H. PERCENT UTILIZATION OF NUTRITION‐SENSITIVE ALLOCATIONS IN BHUTAN AND NEPAL ^a

	2012	2013	2014	2015	2016	2017	2018
Bhutan	‐	‐	85	83	90	62	‐
Nepal b	84	90	80	84	87	92	79

^aSri Lanka excluded due to incomplete data on utilization.

^b2018 data include only federal level data.

Andersen, C. T. , Cain, J. S. , Chaudhery, D. N. , Ghimire, M. , Higashi, H. , & Tandon, A. (2022). Assessing public financing for nutrition in Bhutan, Nepal and Sri Lanka. Maternal & Child Nutrition, 18, e13320. 10.1111/mcn.13320

DATA AVAILABILITY STATEMENT

Research data are not shared.