Networks in development economics
Innovations in agriculture production like compositing (using crop residue from previous harvests for compost) and pit planting (planting seeds shallowly to allow for greater exposure to moisture) can have massive consequences for developing countries. For Malawi, where the overwhelming majority of farmers (97 percent) grow maize, and where over 60 percent of all calorie consumption comes from that crop, the benefits of better agriculture practices are especially salient. Despite that maize enjoys a rather important place in the Malawi diet, maize production remains low for how popular a crop it is. Pit planting and compositing have enjoyed some success in other regions in Africa, and the researchers in this study were looking for the most effective way to disseminate information about these practices in Malawi.
Two factors seem to have provided a lot of the motivation for the structure of this experiment: trained representatives from Malawi’s Ministry of Agriculture and Food Security are low in number compared the number of farmers they need to reach, and farmers who interact regularly with their peers may be more credible or trustworthy and thus better at disseminating information to this demographic anyway.
This study was done as a randomized controlled trial, a research design inspired by the clinical studies done in the field of medicine. 168 villages were divided into three treatment groups and a control group. Treatment groups constituted those that received information about composting or pit planting, whether by lead farmers, peer farmers, or government officers, while the control group received no information. There was also a reward scheme in place, where half of all information disseminators were offered a reward if, at the end of one year, they increased their constituents’ knowledge of the agricultural techniques by at least 20 percent as measured by a short questionnaire. I’m also leaving a lot of stuff out, so the paper should be read to get a more accurate picture of the research.
In a broader context, this gets at one of the questions of especial interest to advertisers and or anybody who wants to market something or make something “viral”: are stronger ties or weaker ties more useful for disseminating information about that something (in this case, agriculture practices, but in another, perhaps a sports car or the Premier Rose Diamond or less trivially, a protest)? In the context of networks, there would be one node to denote the communicator (a government officer or a lead farmer or a peer farmer), with edges branching off of this one node to other nodes denoting farmers, with edges from those nodes (hopefully) branching off to even more farmers, ad infinitum. The key difference here would be tie strength. Government officers would have ostensibly weaker ties, while peer and lead farmers would have ostensibly stronger ones. Intuition suggests that stronger ties might be better at information dissemination, since peer and lead farmers are more likely to be trusted by the targeted demographic. As an additional note: peer and lead farmers are trained by government officers; in this case, the edge between the node representing a given officer and the node representing a given peer or lead farmer can be seen as a bridge across which new information is transferred into the farmers’ social network.
The results of the study corroborate this, though they vary based on participation in the rewards program. Generally, lead farmers successfully increased constituents’ knowledge of these farming practices, and achieved gains in usage when offered a reward. Peer farmers who participated in the rewards program organized more informational activities and were more likely to have their constituents attend, when compared to lead farmers and their constituents/activities. And perhaps most notably, “peer farmers who were seen as most similar to other farm families in their input use and land holding size were most effective.”
Paper: http://faculty.som.yale.edu/mushfiqmobarak/papers/MalawiAg.pdf
Policy brief: http://faculty.som.yale.edu/mushfiqmobarak/featuredresearch/malawiag.pdf