Why do some farmers earn more than others? Scholars have worked to explain the large observed differences in economic outcomes across firms, households, and industries. In developing countries, where small farmers often face limited access to agricultural markets, their agricultural revenues can vary dramatically within the same region, even after accounting for differences in capital investments, labor supply, technology adoption, and agro-climatic factors.
One possible explanation for this puzzling range in agricultural revenues is that farmers do not always have equal access to information about farming practices or agricultural marketing. Particularly in contexts where formal learning channels are sparse or agricultural knowledge is complicated, farmers may find out about improved crop varieties or new marketing channels from their peers. Considerable evidence shows that peers are useful at communicating several types of information including new technologies, household decision-making, labor recruitment, and risk-coping strategies.
In our recent work, we test the importance of peers’ behavior on agricultural revenues from selling cash crops in the rural Himalaya of India. We find that farmers’ revenues from cash-crop sales are associated with as much as 60 percent of the average revenue of their close friends. These effects do not merely pick up the effect of caste or geography: farm revenues are correlated with the revenue of other members of the same caste within their village but to a far lesser extent. Further, farmers’ revenues are not influenced by the revenues of their geographic neighbors or members of the same caste group in nearby villages.
We investigate the channels through which friends’ behavior may help explain farmers’ variation in cash-crop revenues. How much land a farmer allocates to crops is affected by the land allocation of his or her peers, but only for crops that are new to the area and not long-established crops. We find that farmers’ level of input use, particularly pesticide investments but not fertilizer use, is closely related to that of their peers. However, even after combining input use and land allocation decisions, we still find that these two channels together cannot fully explain the variation in farmers’ cash-crop revenues associated with peers’ behavior. This may suggest that friends may be influential in sharing other information including farm practices, marketing strategies, and negotiation skills.
The results we show in this work are important for at least two reasons. Existing extension models, such as the farmer field school (FFS) model, increasingly rely on peer networks to spread information about agricultural practices and new technologies. The FFS focus on training only a handful of farmers, and then expect them to share new information and knowledge to their friends afterwards. Our work poses some potential concerns with this approach. While we find strong effect of peers, the effect is limited to certain types of information and relatively close networks. First, we show that in our context, information from friends is limited to tight social networks. We find that peer effects are most prevalent among closely related friends, and to a lesser degree, those who belong to the same caste within the same village. Therefore, analyses that do not account for social networks at different levels may not fully capture how information may be demarcated in a particular social setting.
Further, we show that the use of certain technologies, such as fertilizer, do not appear to be influenced by friends. This result is also found in Kenya, where the only farmers who learned how to apply the appropriate amount of fertilizer from an extension workshop, and not from their friends or neighbors. On the other hand, we show that peer effects may reach beyond standard agricultural technology adoption. Our results indicate that friends may influence farmers’ production practices and marketing decisions. Thus, our work helps motivate future research in the topic about the extent to which information peers can help farmers receive higher revenue and achieve greater productivity.
Even if there is a strong influence of friends, such influence will not always be equally beneficial for all farmers. On the one hand, friends may help those highly productive friends earn even more. On the other hand, those with few friends, or less productive ones, may not be able to improve their farm outcomes due to their friends. Thus, extension models that rely on peers may want to ensure they reach groups of farmers with lower than average productivity along with the usual ‘star farmers’ to maximize the benefits from new information to improve productivity.
 In our context, self-reported friends may also include kinship ties
Photo credit: http://prakashchandthakur.weebly.com/