Impact of agricultural commercialization on child nutrition in Ethiopia

Smallholder commercialization as nutrition-sensitive agricultural policy has been advocated by many researchers and policy makers though its effect on reducing child undernutrition remains inconclusive. Generalized Propensity Score model which allows for continuous treatments, capture selection bias and observed confounders was applied to estimate the causal relationship between smallholder crop commercialization levels and child undernutrition. Fixed effects model was used to control for the unobserved time-invariant confounders. Impact transmission channels were analyzed by looking at the role of income, gender and the possible substitution effects between the consumption of own-produced and purchased foods. Results provide strong evidence of heterogeneous child undernutritional impacts of smallholder commercialization. Estimates of the dose–response functions reveal that the average levels of underweight and wasting increase with lower and higher intensities of commercialization, but decreases with medium commercialization intensities, with the overall significant beneficial effects. On the other hand, the average level of stunting increases with lower and higher levels of commercialization, but declines with the medium commercialization doses with the overall effect remaining insignificant. Results also provide significant differentials in the estimated effects across time and space, commercialization doses, age, asset ownership and gender. Hence, regional, locational, infrastructural, temporal, gender and age sensitive policies need to be designed to increase smallholder farmers’ commercialization and reduce child undernutrition which is prevalent in the country.