Factors underlying consumers’ choices of dry common beans in Dar es Salaam, Tanzania

Abstract

Beans are widely grown, consumed and are major source of protein especially among low income households in Africa. The overall consumption is low but underlying reasons are not well established, leading to limited ability of chain actors to respond to consumers' requirements. This study was conducted in Dar es Salaam City, Tanzania to assess factors influencing consumers' choices of beans using data collected under Legume Innovation Lab Project SO2.2 in 2015 from a random sample of 600 households in a choice experiment setting. The experiment involved four beans attributes (grain colour, grain size, cooking time and gravy quality). Colour had four levels that were identified as soya supa, soya kawaida, mottled red and yellow while size had three levels (small, medium and large). Cooking time had two levels, slow cooking beans that require more than 120 minutes to cook and fast cooking beans, which require less than 90 minutes to cook. Gravy quality had two levels, poor (watery) and good (thick). Factorial design was used in the experiment where attributes were treated as factors to generate 48 choice tasks. The tasks were then blocked into six choice sets, each comprising eight hypothetical choice alternatives. Respondents were presented with each of the choice alternatives and asked to state whether they would choose it or not independent of the others, resulting into a binary dependent choice variable. Finally, the respondents were presented with all the choices made and given a fixed amount of money to spend over those choices. Cragg's double hurdle model was used to measure consumers' choices of beans and expenditure on the selected bean varieties. Results revealed that choices of beans were significantly influenced by gravy quality, cooking time, grain size as well as household size, income status of consumers, age and education of the main decision maker. Expenditure on the selected bean varieties were significantly influenced by gravy quality, cooking time, grain size, income status of consumers, education and sex of the main decision maker. Efforts to promote production and consumption of beans should focus on improving gravy quality, reducing cooking time and increasing grain size. Future research endeavours should attempt to explore further the overall effect of the three attributes and other identified factors on selection of wide range of beans varieties and expenditure on the same. This could be achieved by conducting more choice experiments using other popular beans.