Evaluating farmers’ preferences for sweet potato variety traits in Western Kenya

Abstract

Despite the high investment put in place so far in developing high-yielding sweet potato varieties, their uptake by farmers in the country has been low. While both farm and farmer characteristics have influenced farmers' choice decisions for different crop varieties, the role of varietal traits has only received limited attention in the empirical literature. This study investigated farmers' preferences towards sweet potato variety traits in western Kenya (i.e., Busia, Bungoma Kakamega, and Vihiga Counties). More specifically, the study sought to assess farmers' willingness to pay (WTP) for better quality sweet potato varietal traits; evaluate farmers' preference certainty and consistency in their choice decisions for sweet potato varietal traits; explore the role of preference variation in the choices made by farmers towards sweet potato variety traits and; finally, ascertain attribute non-attendance of the farmers in their choice decisions for sweet potato variety traits in the study area. The discrete choice experiment (DCE) approach was used to investigate these objectives. The empirical analysis was based on primary data collection from 400 randomly selected farmers in western Kenya (Busia, Bungoma, Kakamega, and Vihiga Counties). Functional estimations followed several models, namely: the multinomial logit (MNL), mixed multinomial logit (MXL), scaled multinomial logit (S-MNL), generalized multinomial logit (G-MNL), generalized mixed logit (G-MXL) and the equality constrained latent class (ECLC) models, which were estimated using Nlogit V software. Farmers, in the first objective, were willing to pay the highest amount for sweetness of the flesh attribute KES 334.70, P<0.01) followed by tolerance attribute (KES 187.09, P<0.05), yield level attribute (KES 137.58, P<0.05), color of the flesh attribute (KES 100.85, P<0.1) and maturity period attribute (KES 100.28, P<0.05) respectively. As for the second objective, results showed that farmers were considerably specific about their preferences, with an average certainty score of 7.0 on a scale of 0-10. They were also consistent in their choice behavior since they chose the same option on the second occasion as on the first occasion in 65.9% of the cases. Results of the third objective show significant preference variation in farmers' affinity towards Sweet Potato traits such as yield level, sweetness of the flesh, color of the flesh, maturity period, and price change. Similarly, the final fourth objective show survey results that ranked flesh color as the most ignored attribute from both stated (61.8%) and inferred (59.2%) attribute non-attendance. In conclusion, due to solid statistical evidence, all four null hypotheses of the study were rejected. The Kenya Agriculture and Livestock Research Organization (KALRO) should consider the production of improved seed varieties (sweet potato breeding programs in Western Kenya) since sweet potato farmers were willing to pay positive amount of money to get sweet potatoes with high-quality attributes. At the National and County level, the Ministry of Agriculture (MoA) should provide regular agricultural extension services to sweet potato farmers in Western Kenya. Respondents in the survey did not pay attention to all attributes when selecting sweet potato varietal traits, implying a need for researchers to take note of this phenomenon in future undertakings involving farmers. Sweet potato breeding research institutions, such as KALRO should use the findings of this study to support development of high-quality sweet potato varieties in Kenya by employing discrete choice approach; ostensibly choice certainty, choice consistency, preference variation, and attribute non- attendance of individual farmers.