Improving nutritional status through behavioural change: lessons from Madagascar

The authors provide evidence for intermediate and long-term effects of a large scale intervention that focuses on quality of nutritional and child-care inputs during the early stages of life. Their empirical strategy uses a combination of difference-in-difference and weighting estimators in a longitudinal survey spanning ten years to estimate the effect of the availability of the programme at the community level on nutritional outcomes. They also provide indirect evidence to support their main identification assumption using falsification tests. They find that the programme helped 0-5 year old children in the participating communities to bridge their gap in weight-for-age z-score and the incidence of underweight. The programme also had significant effects in protecting long-term nutritional outcomes (height-for-age z-scores and incidence of stunting). Importantly, the effect of the programme exhibits substantial heterogeneity: gains in nutritional outcomes are larger for more educated mothers and for villages with better infrastructure. The results are suggestive of important complementarities between child-care, maternal education and community infrastructure.