Choice experiment and experimental auction data to assess consumers' willingness to pay for legume based processed products and their attributes in Malawi

The data sets are part of the outputs of a survey conducted in Central and Northern provinces of Malawi between April and May 2023 by the socio-economics department of Lilongwe University (LUANAR) in collaboration with the Alliance of Bioversity and CIAT and Virginia Tech. The study had two components. The objective of collecting the data under component one was to evaluate consumer willingness to pay for attributes of a hypothetical fast-cooking bean product. This product was locally processed from smallholder-grown bean varieties and was specifically designed to enhance consumer nutrition.. The second component aimed to 1. estimate consumer willingness to pay for whole dehydrated beans, 2. measure the effect of informational interventions on consumers' willingness-to-pay (WTP). These interventions educated participants about the products' economic and nutritional advantages3. identify potential consumer segments for the products. The bean based processed products are new to many consumers as they were recently introduced on the market by local processors. Methodology: A team of 12 enumerators and five supervisors were first trained in survey procedures and the experimental auction protocol by the Alliance of Bioversity & CIAT and LUANAR. They then collected data from 792 households in two cities and three rural districts of Malawi through a two-round process. In the first round, all 792 respondents participated in a choice experiment, answering four choice sets where they selected between two dehydrated bean products or a \"no-buy\" option, alongside a structured questionnaire. The survey collected data from a choice experiment to estimate consumers’ willingness to pay for different attributes of hypothetical legume-blended flour products. The attributes considered were flour composition, packaging material, grind texture, and price. For the second component, half of the initial sample (n=396) was randomly selected to participate in an experimental auction. These participants attended group sessions which were conducted in a central village location and were facilitated by a supervisor following a standard script. Using the Becker-DeGroot-Marschke (BDM) mechanism, they placed bids over three rounds to establish a baseline and then measure the impact of information (convenience and nutritional)interventions. The session concluded with individual interviews on product perceptions and attitudes. Our survey materials included the main household questionnaire, a follow-up questionnaire administered to the experimental auction participants, printed CE visuals, about 300 packets of the pre-cooked products (dehydrated beans), and bidding sheets. The main questionnaire included the following modules: 1) statement of informed consent 2) identification variables and socio-demographic variables, including gender, age, education level, occupation, and household composition, 3) bean consumption (frequency, quantity, methods of preparation, cooking time, expenditure, source of unprocessed bean currently consumed by the households), 4) knowledge and use of processed bean-based products, 5) choice experiment including four choice sets and the choice experiment script, , 6) attitudes and perceptions about processed bean products, and 7) household income status and asset endowments. (2025-11)