Ex-Ante Adoption of New Cooking Banana (Matooke) Hybrids in Uganda Based on Farmers' Perceptions

ABSTRACT Despite the research efforts to introduce the newly developed, improved banana ’Matooke’ hybrids to the farming communities in Uganda, to date no attempt has been made to document the likelihood of farmer adoption of these hybrid bananas in Uganda. The paper has analyzed farmers’ perceptions regarding the newly developed improved Matooke hybrid banana attributes in Uganda to ex ante understand farmers’ likelihood of adoption of these varieties. Descriptive statistics and data reduction techniques (like factor analysis) were used to define the potential explanatory variables affecting adoption. Following this, a Zero Inflated Poisson (ZIP) regression model was applied to estimate the effect of farmers’ perceptions about the hybrid banana attributes and other factors on the likelihood of adoption of the Matooke hybrid banana variety. The results show that, compared to Mbwazirume – a local variety used as a reference, four of the hybrid banana varieties considered are perceived to be better in terms of production characteristics (resistance to sigatoka, weevils, nematodes, tolerance to poor soils, good bunch size, sucker production) but are regarded as inferior in terms of consumption characteristics (taste, colour when cooked, flavour). The hybrid M9 is regarded as having a relatively good performance with respect to most of the production and consumption characteristics. The results of the ZIP regression analysis, including farmer characteristics like gender, family size, age and farmer perceptions of varietal attributes, disease and pests, yield and agronomic attributes, were positively associated with the likely adoption of most of the hybrid bananas. In collaboration with extension agents, variety M9 could be disseminated to a wider farming community targeting, larger households, younger farmers and relatively farmers with large size of land