Nutrient adequacy for poor households in Africa would improve with higher income but not necessarily with lower food prices

Sub-Saharan Africa faces a significant challenge of undernutrition, stemming from a combination of high food prices and low incomes. Historically, advancements in global food security have been driven by both reducing food prices and increasing incomes, as exemplified by the Green Revolution. However, Africa's progress has lagged behind other regions, resulting in a disproportionately high burden of nutrition deficiencies and associated health issues, especially among children under five. While the need to address high costs of nutritious diets relative to poor households' purchasing power is widely acknowledged, there's a scarcity of recent evidence from Africa comprehensively quantifying how consumers adapt their diets in response to price and income changes. This study aims to fill this gap by employing a structural approach to model consumer demand, leveraging nationally representative panel data from five sub-Saharan African countries (Malawi, Niger, Uganda, Tanzania, and Nigeria). The study utilizes improved demand system modelling techniques to account for factors like consumer substitution between food and non-food expenditures, price endogeneity, and unobserved heterogeneity in consumer preferences. This rigorous approach enables the researchers to quantify the sensitivity of diet quality to changes in income and prices of specific foods, providing insights into the differences in dietary patterns between poor and wealthy consumers.

The existing literature on food demand in Africa reveals varying income elasticities for different food groups and countries. A meta-analysis of previous studies suggests generally income-inelastic demand for basic staples and more elastic demand for aspirational foods, with significant cross-country variations. Most earlier studies relied on cross-sectional data, limiting their ability to capture dynamic consumer behavior. Experimental methods, such as randomized controlled trials and choice experiments, have also been employed to study consumer preferences, often showing differences from predictions based on demand models. The literature reveals conflicting findings on the relationship between food prices and nutrient intake, highlighting the importance of substitution patterns across food groups and the varying price responses of poor and wealthy consumers. However, evidence on price elasticities of demand, especially from multiple African countries and using panel data, remains scarce. This study addresses these limitations by using a panel approach, advanced modeling techniques, and data from multiple countries to provide a more comprehensive understanding of consumer behavior.

The study utilizes nationally representative panel data from the Living Standards Measurement Study-Integrated Surveys on Agriculture (LSMS-ISA) for five sub-Saharan African countries. The data includes household-level information on food consumption, expenditures, and prices. The researchers employ the Exact Affine Stone Index (EASI) demand model, a flexible incomplete demand system that avoids the assumption of separability between food and non-food consumption. The model incorporates log price indices, log of real total expenditures, a polynomial of real total expenditures, price-expenditure interactions, and demand shifters to control for observed household characteristics. Community-level correlated random effects address time-invariant unobserved heterogeneity in preferences. A Tobit model accounts for censoring (non-consumption of a food group). The study addresses endogeneity by instrumenting the Stone Price Index and using a Fisher Ideal Price Index and household-level price instruments to account for quality heterogeneity and price search behaviour. The model estimates parameters for food demand and uses these parameters to calculate expenditure and price elasticities. The study assesses diet quality using three measures: intake sufficiency (comparing household consumption to estimated average requirements), dietary balance (shares of carbohydrates, fat, and protein in total dietary energy intake), and the Nutrient-Rich Food Index (NRFI). Finally, the study simulates the impact of a cash transfer (CT) and price discounts (PDs) on diet quality, considering their costs of implementation.

The study reveals several key findings. First, it demonstrates that food demand is significantly more expenditure-elastic than found in previous studies, particularly for animal-source foods and beverages. This suggests a greater potential for income growth to improve food intake, but also a higher vulnerability to income shocks. Second, the researchers find a strong positive relationship between expenditure growth and nutrient intake adequacy, especially among poor consumers. The elasticity of nutrient intake with respect to expenditure is greater for poorer households than for wealthier households. Simulating a cash transfer (CT) shows substantial increases in the probability of sufficient dietary energy intake among the poorest households across all five countries. The CT also leads to improvements in the intake of key macro- and micronutrients, such as protein, iron, zinc, and folate. Furthermore, the CT generally improves dietary nutrient density, as measured by the NRFI, but the effect varies across countries and expenditure groups. The CT reduces dietary imbalance, mostly by lowering the proportion of households with excessive carbohydrate intake. Third, the study highlights the significant impact of food prices on dietary intake, particularly in countries with a single dominant staple crop. In these countries, the price of the dominant staple strongly affects the intake of multiple nutrients, even those not concentrated in that staple, highlighting the importance of cross-price effects. This effect is less pronounced in countries with multiple staple foods. In countries with a single dominant staple crop, price discounts targeting staple grains were almost as effective as cash transfers at increasing dietary energy adequacy amongst poorer households. In multi-staple countries, this was not the case. Finally, the cost-effectiveness of different price discount programs are also examined. Price discounts targeting pulses and nuts appear the most promising in terms of improving nutrient adequacy amongst poorer households with relatively lower implementation costs when compared to other price discount programs.

The study's findings strongly support the notion that poverty reduction is crucial for improving diet quality in sub-Saharan Africa. The findings highlight the significant impact of increased household expenditure on multiple aspects of diet quality, particularly for the poorest households. The effectiveness of cash transfers in improving nutrient intake is demonstrably clear, though it does not eliminate all intake gaps nor perfectly address dietary imbalance. The influence of food prices is also significant, especially in single-staple economies, showcasing the role of cross-price effects and the complexity of predicting dietary responses to price changes. The study offers valuable insights for policymakers, suggesting a multi-pronged approach combining income-focused interventions (such as cash transfers) with carefully designed price interventions targeting specific food groups, while recognizing the potential for tradeoffs and differing effects across countries and income groups. The significant expenditure elasticity of demand for certain food items suggests that poverty reduction is the most impactful route to nutritional improvement.

This study provides robust evidence of the strong link between household income and improved diet quality in sub-Saharan Africa. While price discounts can play a role, particularly targeting pulses and nuts in certain contexts, the most impactful strategy for improving nutritional intake remains poverty reduction. Future research could investigate the intra-household distribution of food, refine the modelling of food consumption outside the home, and explore interventions that address the complexities of cross-price effects and dietary imbalance.

The study acknowledges several limitations. The analysis is conducted at the household level, not at the individual level, potentially obscuring inequitable intra-household distribution of food. Seven-day food consumption estimates are subject to measurement errors. The study does not capture food consumed away from home, and the nutritional content of consumed food items may be misrepresented. The model's simplification of the cash transfer and price discount simulations may not fully reflect real-world scenarios. Despite these limitations, this is a robust study with important policy implications.