The discrete choice experiment (DCE) was undertaken in rural areas of two districts of Malawi, Lilongwe District in Central Malawi and Phalombe District in Southern Malawi - specifically, from four enumeration areas in one 'traditional authority' area in each of Lilongwe District and Phalombe District. Data were collected from 400 households (200 in each district) in February/March 2018 representing the lean season (usually October to March) with maize prices expected to be high. The DCE was undertaken to understand how food choices would respond to a change in the price of maize. It is designed so that we can look at how much people will change their diets based on different maize price scenarios (and to what food types – e.g., to another starch product such as rice, to a vegetable such as cabbage, to more protein-rich foods such as the dried fish, or to an unhealthy soft drink) – and how sensitive people are to price. It consists of ten randomly selected scenarios. Each scenario contained THREE baskets. The preference attributes indicated by food types in each basket were maize, and a maximum of rice, a cabbage, dried fish, and a bottle of the soda ‘Frozy’ (an unhealthy product). The assumption we made based on the data collection undertaken prior to the design of the DCE is that people commonly buy maize (the staple food of most Malawians) – and less commonly purchase the other food types. Rice was chosen as a substitute for maize. Small dried fish were chosen to represent a high-protein food more common in this setting where an alternate food like meat is rarely consumed. Cabbage is chosen to represent a less-commonly consumed vegetable. The soft drink was chosen to represent an unhealthy item that is part of the nutrition transition associated with diets composed of higher proportion of sugars, fats, and saturated fats. We used a d-error-minimising efficient design to generate two sets of five DCE tasks, using a modified Federov algorithm in NGENE software. One set of five tasks had maize at a high price (400 MK/kg), and the other had maize at a low price (; 100 MK/kg), associated with prices in the lean (February/March) and harvest seasons (May). The study estimated that on average, a household consumed food products valued at MK1000.00 (US$1.40) each 2-3 days, which informed the set value of each food basket. Thus, as each basket costs between MK900 and MK1100, the quantity of non-maize products in the baskets varies. We displayed 3 hypothetical baskets in each task using an unlabelled design where each alternative represents a comparable basket. In addition, participants could opt-out of choosing a basket; if they did so, a forced choice task from the three baskets was asked immediately afterwards to assess differences between conditional and unconditional demand.

Social Sciences