Introduction: Malawi provides subsidies on fertilizer and improved seeds under the Farm Input Subsidy Programme (FISP) to resource-constrained smallholder farmers. This is considered to have contributed to achieving national maize self-sufficiency. However, critics point to nutritional consequences of diets that continue to be dominated by maize, which lacks essential nutrients. Relatedly, nutrition indicators for Malawi are below global standards: 37% of under-five children are stunted, presenting developmental consequences. The FISP package has included provisions for legumes, as well as maize, in recent years. Accordingly, legume production has risen, but it is unknown whether such increases have resulted in increased dietary diversity (DD).
Methods: This study uses Integrated Household Panel Surveys data, collected by Malawi’s National Statistical Office with support from the World Bank as part of the Living Standards Measurement Surveys. This is nationally-representative data collected from a sample of randomly selected households across the country. Data used here were from 2013 and 2016, to estimate Panel Poisson Regressions undertaken to investigate whether redemption of legume coupons is associated with dietary diversity. We removed outlier observations from the data and obtained a usable sample of 1,980 observations. To improve accuracy of estimates, bootstrap estimation with 50 replications was conducted. Robust standard errors accounting for any heterogeneity in the data were obtained. The model presented below was estimated:
γ_it = α_0+δ_t+β_i (〖FISP〗_i )+X_it+ε_it
where, i individual Household, t season of survey (2013 and 2016); γ_it outcome of interest (Dietary Diversity defined as count of food groups consumed); δ_t is season when consumption module was administered; 〖FISP〗_i defined as a dummy variable whether household redeemed legume coupon; alternative definition as in receipt of any subsidized coupon is also used; X_it other control variables (socioeconomic factors, market participation, wealth, demographic factors). We also used descriptive analysis to determine emerging patterns of dietary diversity across seasons.
Findings: Redemption of legume coupons influences dietary diversity: While receipt of any subsidy coupon was not associated with dietary diversity, redemption of the legume coupon was positively and significantly associated with DD. Households that redeemed legume coupons are expected to have one additional food group in their DD. Household wealth drives dietary diversity: As in previous studies, we found a significant positive association between household wealth and DD, with durable assets being more prominent (17.9 percent increase in number of food groups consumed) than ownership of livestock (2.5 percent) in increasing DD. Poverty was associated with lower DD at household level. Market participation affords dietary diversity: Some types of market participation were associated with dietary diversity. Results show sale of maize was positively associated with DD, whilst the sale of legumes was not. Overall, however, crop sales value was associated with higher DD. Seasonal variation affects dietary diversity: Accounting for seasonality, we found households consumed fewer food groups during lean seasons, with precarious falls in consumption of food in the ‘meat,’ ‘eggs,’ ‘milk,’ and ‘legumes, nuts, & pulses’ food groups. Households interviewed in post-harvest and planting seasons experienced higher DD relative to those interviewed in lean period.
Conclusions: Farming systems in Malawi dominated by maize production are supported with government input subsidies, which translates into consumption of calorie-dense foods associated with nutrient deficiencies. Since 2009, legumes are subsidized to diversify production and consumption of foods available. Panel Poisson Regression results from nationally-representative data point to relevance of input subsidies, especially redemption of legume coupons, in facilitating access to diversified diets. This suggests the importance of addressing varied availability of legume seeds in FISP markets and scaling up the legume component of FISP, which has been declining from 2.8 MT in 2014/15 farming year to 1000 MT in 2017/18.