South Africa is food secure at the national level; however widespread food insecurity persists
at the household level. To understand the dynamics of micro-level food insecurity this paper
investigates how two different aspects of ‘food access’ – diet quality and diet quantity – affect
two outcomes of ‘food utilisation’ – hunger and nutrition. Diet quantity is captured by food
expenditure in Wave 1 of the National Income Dynamics Study (NIDS). To capture diet quality
I use dietary diversity, which is not directly available in NIDS. I build and test a food group
dietary diversity score and a food variety dietary diversity score using NIDS Wave 1. Both
dietary diversity indicators are found to usefully summarise information about food security
in South Africa by using methods found in the dietary diversity literature. The paper then
turns to testing whether the theoretical differences between diet quality and quantity play
out empirically in the case of nutrition (adult BMI) and hunger (self-reported household
hunger). The results reveal that food variety and food quantity are complementary in
explaining the chance of household hunger, with food quantity having a slightly more
important effect. The pathways to BMI differ by gender. Dietary diversity and food
expenditure are substitutes in the case of male BMI; however, food variety and food
expenditure are complementary to explaining female BMI when food expenditure enters into
the model as a quadratic. Overall, food variety proved to be a stronger and more significant
correlate of both outcomes than the food group dietary diversity score.