Relationship between production diversity and dietary diversity depends on how number of foods is counted

A diverse diet is more likely to meet nutrient requirements (1), and so diverse farm production has been encouraged as a means to increase dietary diversity (2). However, Sibhatu et al. (3) argue that the impact of market access and the buying and selling of food swamps and eliminates the production diversity–consumption diversity relationship in small farms. They argue that excessive production diversity sacrifices the benefits that would come from specialization. They use large datasets and strong analytical methods to support their arguments. However, they make an important oversight.

Production diversity (PD) is counted as the number of crop and livestock species produced on a farm. Dietary diversity (DD) is counted as the number of food groups consumed. The food groups used by Sibhatu et al. (3) are cereals; white tubers and roots; legumes, nuts, and seeds; vegetables; fruits; meat; eggs; fish and fish products; and milk and milk products.

For illustration, consider the following farms. One grows only rice, and so has a PD of 1. Another farm grows rice and corn (PD = 2). A third farm grows rice, corn, and barley (PD = 3). If the families on these farms ate only the foods they produced (rice; rice and corn; or rice, corn, and barley) then the DD would be 1 in all cases, because all of the food is from the cereal food group. In this simplified scenario there would be no relationship between PD and DD. PD varies from 1 to 3, but DD could only be 1.

Consider a different scenario. One farm produces only rice (PD = 1). Another farm grows rice and beans (PD = 2), and a third farm grows rice and beans and raises chickens (PD = 3). If the families in these farms ate their own foods, their DD could be as high as 4 (cereals, legumes, meat, and eggs), and there could be a relationship between PD and DD.

The reality for farming households is surely more complex, with some (most?) farms being a combination of the two types of scenarios (growing/raising multiple types of food within and between food groups), as well as most farms being involved as sellers to and purchasers from the markets. In these scenarios a relationship between PD and DD is possible, though not assured, and it would be helpful to know whether such a relationship exists. It is likely that at least some of the failure to observe a relationship between DD and PD in Sibhatu et al. (3) is that they measured DD and PD using different scales. If PD is measured using the same food groups as DD then there would more likely be a relationship between the two. When nutritionists promote increased production diversity, they implicitly intend for it to be foods from various food groups. Testing for a PD–DD relationship in this sense would provide support for continuing or abandoning such promotions.