Do Egocentric information networks influence technical efficiency of farmers? Empirical evidence from Ghana

Abstract

We investigate the impact of farmers’ egocentric information network on technical efficiency and its distribution in the network, using observational data of 600 farmers from northern Ghana. We exploit community detection algorithms to endogenously identify homogeneous network communities with known structures to account for spatial heterogeneity, in a spatial stochastic frontier model that controls for social selection bias. The empirical results reveal that at the global network level, farmers’ technical efficiency strongly correlate with that of farmers in their egocentric networks. Our findings also show that farmers who are technically less efficient tend to depend on the more efficient farmers in their networks to improve efficiency. We further find that estimating spatial dependence of technical efficiency without accounting for spatial heterogeneity can lead to underestimation of technical efficiency of high (efficiency score >0.6) performing farmers, while overestimating that of medium (efficiency scores between 0.36–0.5) and low (efficiency scores between 0.1–0.35) performing farmers. The findings suggest that identifying central farmers in egocentric networks and improving their technical knowledge in a farmer-to-farmer extension organization, can contribute to improving the productivity of many farmers.