A Latent Class Analysis of agricultural technology adoption behavior in Uganda: Implications for Optimal Targeting

Agricultural productivity is still lower in Africa. This is largely attributed to the lower than expected adoption of modern agricultural technologies. Existing on studies are marred by univariate analyses on single technologies over a limited scope while assuming that the uniform effects of the explanatory variables across farm households. In this study, we use a large dataset that typically covers a wider geographical and agricultural scope to describe modern technology use in Uganda. Using statistical data reduction approaches, we show distinct classes of farmers based on the package of modern technologies mix used. Overall, we find that improved seeds, pesticides and fertilizer are the most commonly used crop technologies while veterinary drugs are the most commonly used technology for livestock farmers. We also find that the majority of farmers, 61% do not use any modern agricultural technology and thus consider them as non-adopters. On the other hand, we find only 5% of farmers belonging to the intensified diversifiers, adopting most of the commonly available agro technologies across crop and livestock enterprises. Using multinomial regression analysis, show that education of the household head, access to extension messages and affiliation to social groups, but with varying intensities, are the key factors that drive switching from the non-adopter reference class to the other three preferred classes that use modern agricultural technologies to varying levels.