Using Social Economics to Design Peer-to-Peer Farmer Extension Programmes

RESEARCH QUESTION

The nature of agricultural production is changing across the world, nowhere more so than in rapidly growing economies such as India and China. With more prosperous and demanding urban customers, consumption patterns shift towards high value crops such as oilseeds, vegetables, animal feed crops, and fruits. This transformation opens new opportunities to improve the livelihood of small farmers, but also raises a new challenge, namely, how to teach millions of farmers how to grow new crops. Government extension services are unable to cope alone with the massive simultaneous demand for location-specific, detailed advice on numerous crops, many of which are delicate and sensitive to pests, while private extension along the standard government model would not be cost-effective. Policy makers and researchers have hence been keen to experiment with alternative models of extension.

One idea that has been attracting growing interest is that if targeted farmers are asked to diffuse new knowledge to other individuals in their social networks the impact of extension services can be greatly enhanced at a low costs (Banerjee et al 2012, Ben Yishay and Mobarak 2012). Across the developing world, much of farmers’ learning about new technologies still occurs as a result of observation of and discussion with fellow farmers (Foster and Rosenzweig, 1995; Munshi, 2004; Bandiera and Rasul, 2006; Conley and Udry, 2010; Krishnan and Patnam, 2012). In one study from rural India, for example, more than 50 percent of farmers report discussing with peers about crop prices, weather predictions, crops to plant, cultivation practices and inputs (Fafchamps and Minten 2011).

To best design farmer-to-farmer interventions, however, policy makers and researchers require an understanding of the formation of farmers’ learning networks and of how farmers’ investment decisions (for example, for the acquisition of new information) are affected by embedded-ness in social networks. In this project we will use the lab-in-the-field method to provide empirical evidence on these two questions.
 

PROJECT

We are currently running two field experiments in the agricultural districts around Pune, Maharashtra. The first experiment tests whether farmers can form learning networks that maximise access to information and whether social homophily decreases the efficiency of the resulting networks. We observe  farmers’ link formation decisions in a sequential link formation game and can test for the rationality of individual play, the efficiency of the final networks, and for whether the introduction of minimal social identities in some sessions prompts individuals to deviate from economic efficiency to avoid linking to members outside the group.

The second experiment tests the prediction of an established model of strategic investment decisions in an exogenous network.  The experiment is enriched by a task that measures social preferences and with the selected introduction of status differentiation among participants.

Results from the two experiments will be combined with rich observational data to identify the most problematic aspects of diffusion processes and will propose designs of farmer-to-farmer interventions that overcome these, for example, interventions that foster the circulation of information across social groups.