Conceptual framework

In Section 5.2, we present the conceptual framework of market price effects on children’s nutrition. The research data, including descriptive statistics and methods, are described in Section 5.3. The results of price decomposition and the nutrition–price analyses are presented in Section 5.4 and discussed in Section 5.5, whereafter concluding remarks are provided in Section 5.6.

direct health effects caused by heat stress, disease spreading, or clean water scarcity that influences children’s nutrient uptake and parents’ productivity and income opportunities (Brown et al., 2014; Cooper et al., 2019; Gebre and Rahut, 2021; Grace et al., 2021; McMahon and Gray, 2021; Nawrotzki et al., 2016). We account for these other channels of weather on nutrition by controlling for total rainfall and mean temperature in some of our model specifications. However, in these cases, weather shocks cannot be regarded as instrumental variable.

Extreme weather events might also encourage the implementation of protectionist or other domestic policies that affect children’s nutrition through other channels than the price. For example, trade bans can reduce employment opportunities of parents and affect children’s nutrition through policy-induced income effects. Food aid on the other hand might improve nutrition and thus counteract the expected price shock. Such policy implications would occur concurrently to the price adjustments and might be endogenous to our resulting price volatility effects. By using futures volatility to instrument local market price volatility we address this endogeneity concern.

We are aware that there are further confounding factors related to geographic characteristics of market and household locations that affect both, nutrition and market food supply. These include local institutional settings, policies, market integration and access, infrastructure, sanitation facilities, and the regional climate settings. By using market-fixed effects we account for such time-invariant local variation in our analysis. Complementarily, years with wide-spread nutrition shocks, e.g., caused by a global heat wave, are controlled for with year-fixed effects.

Changes in staple food prices are of great relevance, since staple foods comprise a large share of diets especially of food insecure children, they are required to cover the minimum caloric intake, and therefore other dietary components may be sacrificed for their consumption. Implications for different household subgroups, identified by household and parent characteristics, are likely heterogenous.

Staple food price volatility likely affects food access, which, in turn, affects children’s nutritional outcomes differently for net producing and purchasing households. High food price levels can reduce net food buyers’ access to food , unless incomes adjust accordingly, whereas net food sellers might benefit from positive income effects (Kalkuhl et al., 2016; Wodon and Zaman, 2010). Volatile prices can cause fluctuating food access with alternating periods of easy and difficult access for households with credit constraints, while wealthier households may have the option of building stocks in times of low prices. Staple food price volatility increases the risk and transaction costs for producers, discouraging investments in efficient staple production systems, while urging food-deficit agricultural households to maintain the production of staples instead of diversifying their crop choices (Poulton et al., 2006). Some part of volatility is reoccurring, such as generally higher prices during lean season (Maître d’Hôtel and Le Cotty, 2018), and is thus predictable by households and producers. If it is within their means, producers and households stock food as a coping strategy. The unexpected part of volatility, on the other hand, may drive additional risk to food security that is more difficult to foresee.

Unexpected volatility may come in the form of price spikes. Sudden and severe price increases restrict food access of net-buying poor households that lack the ability to build stocks. Price drops might, on the other hand, generate a sharp fall in producers’ income. The corresponding households may adjust their food storage and selling behavior prioritizing their own subsistence consumption, which can result in giving up the highest retrievable profits and increasing price volatility further (Maître d’Hôtel and Le Cotty, 2018). Overall, children’s nutrition in households that rely on purchased foods might benefit from price drops and suffer from price spikes, whereas the opposite might hold if the household’s income is strongly connected to food prices on the market. The latter though will not hold if price changes result from the local households’ production quantity. For example, producing households only benefit from high prices if they have the capacities to supply the market – this effect is of course limited if their own production shortfall was the underlying reason for the price increase.

Whether food access and nutrition are protected from price movements depends on the interplay of household assets, household income, parent and children characteristics. Wealthier households spend a smaller share of their income on food and are more likely able to increase their food budget to mitigate food price increases. Higher parent education can be a sign of more profound nutritional knowledge, which might result in better childhood nutrition, although impacts of parental education on stunting are found to be modest (Alderman and Headey, 2017). Having agricultural assets can be seen as an additional source of wealth. Households with agricultural assets or female occupations in agriculture may obtain income from agricultural activities and have additional means with which they can ensure subsistence.

In the former case, such households might benefit from increased agricultural prices when they sell their products. However, they might also suffer from increased input costs (e.g., for animal feed). In the case of subsistence production, farming households might be hardly affected at all.

Rural and urban households might be affected differently, as rural households might fall back on subsistence production, while urban households may be more reliant on food purchases. Rural markets are potentially less integrated and may be less strongly affected by international price transmissions. In contrast, improved access to food imports can enhance the nutrition of children in urban households as trade can be used to balance the effects of local weather shocks. We account for this with market fixed-effects.

Children’s characteristics can also determine nutrition outcomes (Buisman et al., 2019). Different shocks to food availability are likely to have differing effects depending on the child’s developmental stage. For example, older children who are no longer breastfed might be more directly affected by reduced staple food access. Mothers might buffer especially younger children’s caloric intake by reducing their own consumption (Block et al., 2004). Nutrient needs of mothers and infant children are especially high during pregnancy and the following two years, but can only be met with sufficiently diverse diets (Adu-Afarwuah et al., 2017; Dewey, 2013). If a child has a higher birth order or is born as a twin, the household food is shared with more people. A child’s sex might also influence how food

availability affects nutrition for biological and especially for cultural reasons (Keino et al., 2014; Wamani et al., 2007).