All hands on deck: new evidence on the need for a multi-sectoral approach to reducing childhood stunting


This page in:

In Indonesia, chronic malnutrition is widespread with more than one-third of young children being stunted. Despite the reduction of the poverty rate from 16.6% to 11.4% from 2007 to 2013, the rate of stunting amongst children under the age of five has remained alarmingly high, exceeding 37% in 2013, although that figure has declined in the last two years. Stunting has important lifelong consequences for health, as well as for cognitive development, education, human capital accumulation, and ultimately for economic productivity.

However, to reduce stunting it’s not only important to focus on the health sector. It also requires improvements in other sectors such as agriculture, education, social protection, and water, sanitation, and hygiene (WASH). As originally emphasized by the UNICEF conceptual framework, to ensure that a child receives adequate nutrition depends on four critical factors: care, health, environment, and food security, areas that straddle multiple sectors.

Naturally, interventions to reduce childhood stunting that target one particular sector may not achieve the desired impacts if they aren’t complemented with appropriate measures across other sectors. For example, integrating nutritional considerations in the agricultural sector is unlikely to take into account the fact that WASH services may be poor and inadequate in some communities. As a consequence, the impacts of such interventions on key nutrition outcomes could be impeded considerably by the absence of adequate WASH facilities. The impacts of the same nutrition-sensitive agricultural interventions could be enhanced considerably if they were to be accompanied by simultaneous improvements in the water and sanitation facilities in the same communities.

Given such interlinkages, we launched a new report with the National Institute of Health Research and Development (NIHRD) of the Ministry of Health of the Republic of Indonesia that lays the groundwork for a more effective multi-sectoral action on reducing stunting in Indonesia.  The report provides a lens for a more systematic monitoring of the progress in access to the four main drivers of child malnutrition.  It also identifies potential “binding constraints” to reduce child stunting.

The analysis reveals that in 2013, 23% of the children between 0 and 3 years did not have adequate access to any of the four determinants of nutrition, down from 39% in 2007. On the other end, less than 1% of children had simultaneous access to all four of the underlying determinants of nutrition (see Figure 1). The differences in simultaneous access to two or more drivers become larger between rural and urban areas and by quantile of wealth.
The low proportions of children with simultaneous access to more than one of the drivers of nutrition suggests that the success of sector-specific nutritional initiatives could be constrained by inadequate access to the other underlying determinants of malnutrition.

Further analysis also reveals that simultaneous access to two or more of the four key determinants of nutrition is associated with a decrease in the likelihood of a child being stunted (see Figure2).  The likelihood that children between 0 and 36 months of age are stunted is lower when they have simultaneous access to adequate levels to two of the four drivers of nutrition and is even lower among children with simultaneous access to adequate levels to three of the four drivers of nutrition (Figure 2). 

In combination, these results suggest that progress toward reducing stunting in Indonesia can be enhanced by coordinated multi-sectoral interventions that address effectively the four key underlying determinants of nutrition. For example, nutrition-sensitive intervention in the agricultural sector, such as giving incentives to poor and undernourished households to grow fruits and vegetables in their own home gardens, could be accompanied by coordinated interventions to improve the WASH services prevailing in these communities. Such coordinated efforts would not only decrease the incidence of diarrhea and illnesses associated with an inadequate environment but also strengthen the positive nutritional impacts of consuming micronutrient-rich fruits and vegetables.

The report also highlights the numerous data gaps that need to be addressed for a more informed and better coordinated approach across line ministries in Indonesia.  In essence, it concludes that for Indonesia to fight against childhood malnutrition requires clarity and prior agreement on the common yardstick used among the different sectors involved so that the country can achieve efficient improvements in its policies.


Emmanuel Skoufias

Lead Economist at the Poverty & Equity Global Practice of the World Bank Group

Jack Clift (R4D)
May 30, 2017

This is very interesting work with substantial policy relevance. One question: while it makes sense that having access to increasing numbers of the drivers has an increasingly positive association with lower stunting rates, were you able to determine whether the whole is greater than the sum of the parts? - i.e. is their a positive interaction effect, whereby (say) having access to improved food security and improved WASH has a bigger combined impact than just the impact you get from improving each individually?
Given limited resources and the need to scale up many different interventions in different sectors in different geographies, it would be helpful to understand if there are positive (or conceivably negative) interactions associated with scaling up some combinations in the same geography. e.g. we may not get the full benefits from improved food if we don't reduce diarrheal disease and loss of nutrients; conversely, increasing exclusive breastfeeding rates may have less marginal benefit in areas with improved water quality compared with areas with poor water quality, if part of the benefit of exclusive breastfeeding is avoiding contaminated baby formula and water-borne disease.
Clearly the eventual goal should be strong programs across all sectors in all places, but until we get there there may be better and worse ways of deploying scarce resources for the scale up. Would love to hear any insights on this from your work.

Emmanuel Skoufias.
June 01, 2017

Dear Jack,
Thank you for your comment. We share your desire for ranking different combinations of interventions and, for any given context, choosing the best set. From our regression analysis however, we have been unable (in spite of really wanting to do that) to make the claim that “whole is greater than the sum of the parts,” and much less which specific intervention combinations work the best. The lack of positive interaction effects in the analyses may be due to the small share of children with access to the specific sets of adequacies. For more definitive answers much more detailed data would be needed.
We do find that there are differences in the probability of stunting depending on which of the four determinants a child has access to. Specifically in Indonesia, among children with access to just one driver, access to adequate food security alone is associated with the largest increases in the average height-for-age, than access to adequate care alone which is not associated with significant gains. However, it is important to keep in mind that while the analyses describe the current state of nutritional drivers and their correlation with nutrition outcomes the models are not causal.
Emmanuel Skoufias

Jack Clift (R4D)
June 04, 2017

Dear Emmanuel - thanks for the clear response. All makes sense, and I appreciate you taking the time to reply. Best, Jack