A latent class analysis of improved agro-technology use behavior in Uganda: Implications for optimal targeting

Abstract

This study uses a large dataset that covers a wide geographical and agricultural scope to describe the use patterns of improved agro-technology in Uganda. Using latent class analysis with data on more than 12,500 households across the four regions of Uganda, we classify farmers based on the package of improved agro-technologies they use. We find that the majority of farmers (61 percent) do not use any improved agricultural practices (the “nonusers”), whereas only 5 percent of farmers belong to the class of “intensified diversifiers,” those using most of the commonly available agro-technologies across crop and livestock enterprises. Using multinomial regression analysis, we show that education of the household head, access to extension messages, and affiliation with social groups are the key factors that drive switching from the nonuser (reference) class to the other three (preferred) classes that use improved agrotechnologies to varying degrees. Results reveal the existence of heterogeneous farmer categories, certainly with different agrotechnology needs, that may have implications for optimal targeting.