In this paper, we investigate the association between smallholder livestock commercialization and household dietary diversity. Our measurement of dietary diversity relies on the six food groups specified in the Zambian Food Based Dietary Guidelines (FBDGs) as well as the standard household dietary diversity scores (FAO, 2021). We examine livestock commercialization at both the extensive and intensive margins, where the extensive margin refers to whether households sell livestock, and the intensive margin refers to the actual number of livestock sold. To have a better sense of the degree of commercialization, we also use the livestock commercialization index which defines livestock commercialization as a ratio of total livestock sold to total livestock owned. To enable comparison between livestock and hinge this on efficient measurements, we convert livestock resources to Tropical Livestock Units (TLUs) since different livestock offer unique contributions.
In the interest of understanding which livestock contributes more to dietary diversity and adherence to Zambian FBDGs, we perform some heterogeneity analysis across five livestock portfolios: cattle, pigs, small ruminants (sheep and goats), and poultry. Related to this, we also look at how the six different food groups respond to livestock commercialization. To further enhance our analysis, we incorporate a gender dimension to understand how gendered decision-making matters. Different studies have shown differential access and control over decisions in households – including livestock production and sales. Our interest is in examining whether these gender dynamics matter for dietary diversity. We differentiate between households where men are the primary decision-makers and those where women play this role besides the standard differentiation by male and female headed households. Beyond this, we also estimate the association between gendered decision making and dietary diversity. Additionally, we consider the implications of livestock commercialization on income, off-farm income and farm production diversity, with the aim of demonstrating associations between livestock commercialization and these pathways as well as with dietary diversity.
We rely on a rich two-wave micro-data from Zambia and estimate different panel data estimation to control for both observed and unobserved heterogeneities. Particularly, we employ the Mundlak Chamberlain device (correlated random effects) to control for possible time invariant unobserved heterogeneity in the relationship between commercialization and dietary diversity. (Mundlak, 1978; Chamberlain, 1984). To further address possible issues of endogeneity, we employ the two-stage least squares estimation approach which relies on the specification of instrumental variables. To identify the association between livestock commercialization and dietary diversity, we use the village-level livestock commercialization level as an instrument following a rich literature on smallholder commercialization that has relied on this instrumental variable (Ogutu et al., 2020; Geffersa and Tabe-Ojong, 2024). Given some of the potential limitations with this instrument (which we discuss below), we perform some robustness checks by estimating the Kinky least squares regression as an alternative identification strategy (Kripfganz, and Kiviet, 2021).
Our results show that livestock commercialization is associated with dietary diversity. The consumption of fruits, vegetables, legumes, pulses and nuts as well as cereals, starchy roots and tubers respond more to livestock sales. In terms of which livestock contributes more to dietary diversity, we show that dietary diversity responds more to the sale of cattle and pigs at the extensive level of commercialization while poultry sales matter more at the intensive level. Exploring what may explain the relationship between livestock commercialization and dietary diversity, we identify three possible pathways - livestock income, off-farm income, and crop production diversity. From a gender perspective, we find similar gains for households where women and men share decision making and control, but the effects are more pronounced when females are overseeing decision making of the household. Additionally, we show that income, off-farm income, farm production diversity and gendered decision making are also associated with dietary diversity, further strengthening the identified pathways. These insights point to the importance of livestock commercialization as an important entry and leveraging point for improving dietary diversity.
Our study offers two main contributions to different strands in the empirical literature on smallholder commercialization and dietary diversity. First, we offer insights from livestock commercialization beyond crop commercialization. While crops are certainly important, livestock also play a crucial and increasingly significant role amongst smallholder households (Brockington, 2021). Livestock offers immense opportunities for households, including consumption smoothening, precautionary savings and serving as buffers against shocks such as droughts and conflicts making them relevant for addressing food consumption concerns. The second contribution directly stems from the first where we explore how livestock commercialization may lead to dietary diversity. Here, we contribute to an important part of the literature on the role of livestock for consumption smoothening. We offer additional insights into this literature by specifically looking at how livestock commercialization contributes to dietary diversity.
