Do Behavioral Interventions Increase the Intake of Biofortified Foods in School Lunch Meals? Evidence from a Field Experiment with Elementary School Children in Ethiopia

Background: Many African countries are seeking to improve nutrition by introducing biofortified foods in school feeding programs. These programs are generally designed to create demand for biofortified foods both in and outside of school. Finding ways to encourage child acceptance of novel biofortified foods is key to the success of this strategy. Objectives: The aim was to assess effects of 2 behavioral interventions in promoting the consumption of biofortified foods as part of school lunch meals. Methods: The study is based on a field experiment involving 360 school-going children of in the third, fourth, and sixth grades. We tested if structured provision of information about the nutritional benefits of a biofortified food and its association with an aspirational figure influence its consumption when served alongside a favorite local food as part of school lunch meal. Six schools in Tigray, Ethiopia, were randomly selected to participate, with 4 participating in the Orange-Fleshed Sweetpotato (OFSP) Program. Of the 4 participating in the program, 2 were assigned to provide educational information about sweetpotato, whereas 2 presented the same information plus depictions of an aspirational figure (a famous local athlete) associated with the sweetpotato. Results: Provision of information on the nutritional benefits of biofortified food combined with an aspirational figure resulted in increased consumption of biofortified food by children. However, provision of the information alone did not detectably increase consumption. An analysis of trends over the course of the study revealed no discernable decay effect. Conclusions: Our results highlight the potential for relatively inexpensive behavioral interventions to increase acceptance of novel biofortified foods among children in a developing-country context. Larger studies with more varied interventions and larger numbers of participating schools could address several of the weaknesses in this study and establish more robust findings