The role of food preferences in determining diet quality for Tanzanian consumers

Abstract

Consumer preferences can be leveraged to magnify the impacts of agricultural investments and interventions on diets for all consumers in an economy, not just farmers. Using nationally representative panel data from Tanzania, we estimate demand for 19 food groups using an Exact Affine Stone Index censored demand system, which is flexible, utility-theoretic, controls for unobserved heterogeneity, and accounts for bias arising from endogenous prices. We find strong links between growth in household expenditures and improved diet quality. Also, staple grain prices are important determinants of nutrient intake. For poor consumers, e.g., protein and iron intake are more sensitive to maize price changes than to changing prices of other foods that contain more protein and iron. We use simulations to show that cash transfers and price vouchers targeting staple grains, pulses & nuts, and starchy staples could be effective in shrinking gaps between recommended and actual dietary intake for poor consumers.