McGovern-Dole Food for Education and Child Nutrition in Kyrgyzstan – Baseline Study on Nutritional Variables

In this “Baseline Study on Nutritional Variables”, we address three broad research questions around nutrition in the context of the national Food for Education and Child Nutrition (FFE) program in Kyrgyzstan. The three questions are:

RQ1: What is the status of nutrition in Kyrgyz households with primary grade children?

Here, we place particular and novel emphasis on measuring and studying a) nutrition knowledge, preferences and practices among children and their caregivers and b) nutrition outcomes and home, such as dietary diversity.

RQ2: What is the impact of nutrition on child health and education among Kyrgyz children?

Here, we place particular and novel emphasis on studying a) nutrition knowledge, preferences and practices among children and parents, and b) “intrapersonal foundations” of learning in the sense of children’s school participation, health and cognitive and non-cognitive skills.

RQ3: What is the impact of the FFE program on nutrition as well as on health and education among Kyrgyz children?

Here, we place particular and novel emphasis on a) identifying net links, b) uncovering the causal pathways at work, distinguishing nutritional and alternative channels (such as school feeding incentivize school attendance based on purely economic terms), and c) estimating the separate and combined impacts of hot meal provision and community- and family-based behaviour change program components.

To maximize learning from this Baseline Study, we developed an innovative mixed-method approach that combines survey-based quantitative research with in-depth ethnographic insights. In the quantitative survey conducted in the period November 2019 till February 2020, we collected detailed information from 3035 grade 1 and grade 2 students from 154 FFE program schools, as well as their primary caregivers. Schools and children were selected in a two-stage random sampling process. In addition, we conducted two waves of qualitative research: one before the household survey to inform the quantitative design and one after the household survey to complement quantitative analysis. The qualitative research included 40 in depth conversations in 21 FFE communities, with school representatives and teachers, caregivers, children and experts of nutrition and education familiar with the FFE program. The results derived from this methodology are preliminary to the extent that they rely on baseline data only; a validation of all the preliminary causal findings requires a full set of endline data.

Our baseline analysis suggests that a lot of children have good nutrition knowledge and healthy food preferences in general, but also that they eat many snacks, which are often unhealthy. Their caregivers have often limited nutrition knowledge, but typically have healthy food preferences. Nutrition at home is now well diversified for a majority of households.

In terms of health and learning, attendance rates among Kyrgyz children are very high, they are generally in fine health, and they exhibit very good other “intrapersonal characteristics” that determine learning outcomes, such as low short-term hunger in class, high cognitive function and strong noncognitive skills. Against this encouraging background, we find that healthy nutrition plays a key role and suggests strong benefits: avoiding unhealthy snacks is strongly linked with better executive function and better literacy and numeracy scores (primarily among grade 2 children). In addition, household diets that are rich in vitamin A are strongly associated with better child health, less short-term hunger, better executive function, and higher literacy and numeracy.

Our preliminary impact analysis drawing on a novel design using baseline data of children who just started grade 2 suggests that the FFE program has a critical role for improving nutrition and fostering learning. For learning outcomes, our results indicate that purely economic and purely social impact channels are less relevant than nutritional channels. Specifically, we document with baseline data that one year of the FFE program led to improved nutritional practices at home, which may be adapted even without strongly shifting caregivers’ knowledge or preferences. In turn, we observe large gains in terms of learning due to strong FFE impacts on executive function and numeracy in particular.

Quite strikingly, we observe similar differences between children from 2018 and 2019 FFE communities who just started grade 1. A potential explanation points to positive spill over effects in FFE communities, which would be an encouraging result from a programming perspective, as it would suggest that the FFE program can also benefit children who are not direct beneficiaries, i.e. who did not receive hot meals in the past year. Yet, this is an indicative finding only relying on baseline data only; we cannot definitively rule out using baseline data only the possibility that children, households, and/or communities differ structurally across 2018 and 2019 FFE schools.