Abstract

In an attempt to enhance agricultural development in sub-Saharan Africa, the region has witnessed an influx of approaches to agricultural technology development and dissemination in the past decades. Yet there is paucity of empirical evidence that links these past approaches to productivity indicators or justify the phasing out of the existing approaches with the new ones. We use cross-sectional baseline data, which were collected in 2008 over 2130 smallholder farmers and 242 villages in East and Central Africa before the implementation of the most recent approach to agricultural research and development known as integrated agricultural research for development (IAR4D), to examine whether the adoption of soil and water conservation technologies (SWCT) generated and disseminated through the past approaches have had an impact on technical efficiency. Taking into account the endogeneity of technology adoption and assuming that impact is heterogeneous across the population, we use an instrumental variable approach to estimate local average treatment effect (LATE). The data suggest that adoption of SWCT has had no significant impact on technical efficiency of smallholder farmers in Rwanda and the Democratic Republic of Congo. However, the impact is negative for smallholder farmers in Uganda as well as for the pooled sample, although the magnitude is small. Thus, the findings justify the need for the introduction of another approach such as IAR4D, which aims at internalising external factors that constrain adoption of improved technologies and technical efficiency.  

Key words: Impact, adoption, soil and water conservation, technical efficiency, integrated agricultural research for development (IAR4D), local average treatment effect (LATE), Sub-Saharan Africa.