search advanced search
UNICEF Innocenti
Office of Research-Innocenti
search menu

Tunisia study explores effect of intervention packages on childhood stunting

Standard packaging of interventions has variable effects on different population groups
19 Oct 2017

(19 October 2017) A new UNICEF Innocenti Working Paper: Child Undernourishment, WASH and Policy Synergies in Tunisia, establishes an econometric strategy for implementing UNICEF’s conceptual framework on nutrition by analysing the effects of water, sanitation, and hygiene (WASH) investments on child stunting. Looking at evidence from different populations and locations in Tunisia – a country with uneven progress in child nutrition – the paper asserts that successful mitigation of child stunting cannot rely on one universal approach, but instead requires mapping and application of the most effective intervention packages by residence and socioeconomic status to meet the varied needs of children in different contexts.

A family in their kitchen in an urban low income area of Tunis, the capital city of Tunisia. Many families in the neighborhood arrived from rural areas in search of jobs, but most are now unemployed.


The paper, co-authored by UNICEF Innocenti’s Jose Cuesta and the World Bank’s Laura Maratou-Kolias, demonstrates how improvements in stunting come from successful integration of interventions from nutrition and WASH sectors. While UNICEF’s strategy has long recognized that WASH interventions can improve nutrition, this paper intends to enable policy makers and programme managers to implement more effective intervention packages to improve nutrition outcomes specific to targeted population groups.

“Typically, economists will tell you which intervention has the largest impact, but we wanted to look at which packages of interventions correlate with the best outcomes,” Cuesta, author and Social Policy Chief with UNICEF Innocenti, said. “Indicators are just one part of how to improve programing – in this paper, we’re trying to provide a bigger strategy to inform policy interventions, linking UNICEF’s conceptual framework with multidimensional indicators to follow over time.”

In the case of Tunisia, the study analysed which investments had the largest impact on improving child nutrition using data from the 2012 Multiple Indicator Cluster Survey. The estimates indicated that multi-sectoral invention packages varied in how they correlated with better nutrition outcomes depending on socioeconomic status and residence.  Among rural poor groups in Tunisia, for example, packages combining WASH, care, food security and health interventions did not show significant effects on child nutrition, whereas this combination was effective in Tunisian urban settings. Among rural poor groups, integration of care and food security interventions alone correlated with better nutritional outcomes for children. 

Integrated multi-sectoral interventions for appropriate child nutrition. Adapted from UNICEF policy review 1990


Intervention packages can be ineffective for several reasons. Lack of sufficient investment and poor efficiency in implemention can explain why some interventions are more effective than others, especially for the most vulnerable rural poor. According to Cuesta, measurement and evaluation need to be improved in order to implement better targeted solutions.

“The data from Tunisia shows that single interventions don’t work – or don’t correlate with the best outcomes for nutrition specifically,” said Cuesta. “Only packages work, but not all packages worked for every location, setting, or socioeconomic zone. Packages aren’t going to work the same for everyone.”

This research aims to lay the foundation for more effective efforts to mitigate child stunting by improving understanding of how interventions from different sectors should be packaged differentially to address undernourishment by population group and location. The paper stresses that a single intervention will not bring uniform benefits across different types of households and that since investments are limited, interventions need to selectively respond to the specific requirements of different types of households before nutrition can improve evenly for all Tunisian children.


Nutrition, stunting, WASH, programming, Tunisia

Related publications: