Social networks, heterogeneity, and adoption of technologies: Evidence from India

This study examines the role of caste-based affiliations in the smallholders’ social network interactions for adoption choices. In particular, whether lower-caste, namely Scheduled Castes/Scheduled Tribes, farmers rely more on social networks for information than their counterparts. We further explore whether social network effects are more pronounced when farmers interact within their caste than otherwise. Finally, the study tests whether the effects (intra-caste and inter-caste) vary by caste—SC/ST versus non-SC/ST farmers. The study uses a survey of 478 mustard farmers in Rajasthan, India. Econometric concerns related to unobserved heterogeneity are addressed by employing specifications with village fixed effects and a series of robustness tests. Simultaneity concerns are addressed by analyzing the social network effects in a dynamic adoption framework. Results show that the adoption choices regarding hybrid mustard seeds are more pronounced for the lower-caste farmers than for their counterparts. Findings reveal that social network effects are significant in intra-caste but not in the case of inter-caste. Finally, the result shows that the likelihood of accepting advice in technology adoption is higher when SC/ST farmers interact with non-SC/ST network members than when non-SC/ST farmers interact with SC/ST network members.