Can Peers Improve Agricultural Revenue? TISORN SONGSERMSAWAS a , KATHY BAYLIS b , ASHWINI CHHATRE c and HOPE MICHELSON b,* a International Fund for Agricultural Development (IFAD), Rome, Italy b University of Illinois, Urbana, USA c Indian School of Business, Hyderabad, India Summary. — Crop revenues vary greatly among farmers and the source of that variation is not fully understood, even after controlling for factors including input use, technology adoption, and other agro-climatic factors. One hypothesis that may explain the variation in outcomes among farmers is differential access to information through peers. Using a household survey from India containing detailed information about personal relationships, we estimate peer effects on cash crop revenue using a novel spatial econometric technique to control for reflection. Our results show that 60% of farmers’ revenue is explained by peers. Peer effects are particularly large in pesticide use and in the cultivation of a new crop. However, peer effects in input expenditures and land allocation cannot fully explain the variation in revenue, implying peers may also associate with management, negotiation, and marketing. We find that peer effects are significant among farmers’ self-reported peers, especially among those peers who are farmers’ main advisors for agricultural matters. Although caste-based networks (both within the same and in adjacent villages) are important, their effect is smaller than that of self-reported peer networks. We empirically rule out that our effects are driven by other factors such as geographically correlated unobservables, farmers following a lead farmer or economies of scale. Our findings speak to both the potential and the limitations of peers as sources of agricultural information, and highlight the need for future research about how to best integrate peers into agricultural extension. Ó 2016 Elsevier Ltd. All rights reserved. Key words — peer effects, India, South Asia, crop revenue, spatial econometrics, social networks 1. INTRODUCTION AND MOTIVATION Economists have long tried to explain the considerable observed variation in economic outcomes across firms and households. In developing countries where market frictions are frequently high, economic outcomes can vary dramatically (Bloom, Eifert, Mahajan, McKenzie, & Roberts, 2013). Agri- culture in the developing world provides a special example of this puzzle; farm revenues and profits tend to exhibit wide dif- ferences within the same region, even after accounting for inputs, technology use, and agro-climatic factors (Fan, Hazell, & Thorat, 2000; Murgai, Ali, & Byerlee, 2001). Why do some farmers earn more than others? One possible explanation is that differences in revenues reflect differential farmer access to information about production practices and marketing (Aker, 2010; Birthal, Kumar, Negi, & Roy, 2015; Conley & Udry, 2010; Jensen, 2007). While agricultural extension services offer one mechanism to disseminate new techniques and market opportunities (Roeling, 1984), infor- mation flow through such official channels may still be limited, particularly for more complex practices which benefit from demonstration (Anderson & Feder, 2004; Duflo, Kremer, & Robinson, 2011; Waddington, White, & Anderson, 2014). Par- ticularly in contexts characterized by limited formal informa- tion, peers have been found to be an important mechanism for disseminating information about new technologies (Foster & Rosenzweig, 1995; Liverpool-Tasie & Winter- Nelson, 2012; Magnan, Spielman, Lybbert, & Gulati, 2015), credit (Okten & Osili, 2004; Wydick, Karp Hayes, & Hilliker Kempf, 2011), labor recruitment (Mano, Yamano, Suzuki, & Matsumoto, 2011), household decision-making (Kandpal & Baylis, 2015), and risk mitigation (Di Falco & Bulte, 2013). These sorts of endogenous social effects are often called peer effects in the literature and refer to the relationship between the behavior or outcomes of an individual and the behavior or outcomes of a social group of which the individual is a member (Manski, 1993). This paper tests how a farmer’s agricultural revenues are influenced by the revenues and characteristics of his or her peers. In our setting, one’s peers may include family members, relatives, friends, and social and religious ties. Based on this definition of peers, the intensity of interaction in the networks of each individual may vary, which has an important implica- tion for peer effects (Munasib, Roy, & Birol, 2015). We study the revenues of small farmers in Thaltukhod Valley, Himachal Pradesh, India. These small farmers are highly dependent on crop revenues, and even a small increase in crop prices or pro- duction can greatly increase their income. Understanding how social connections might influence farmer outcomes can pro- vide critical insights into why some farmers escape poverty and others remain stuck at a low level of productivity. Our study site is the Thaltukod Valley in the foothills of the Himalayas in Himachal Pradesh, India. Our study analyzes data from a survey of all 522 households in the valley’s 17 * This research is funded by the ADM Institute for the Prevention of Postharvest Loss, University of Illinois. The authors thank the editor and two anonymous referees for insightful comments and suggestions. We are also grateful to Mary Arends-Kuenning, Benjamin Crost, Alessandra Garbero, Phil Garcia, Barrett Kirwan, Xin Li, Mindy Mallory, Carl Nelson, Atul Nepal, Alex Winter-Nelson, and David Zilberman for generous comments on earlier drafts. Satya Prasanna provided excellent help with data. We thank the seminar participants at the 2014 AAEA/EAAE/CAES Joint Symposium in Montreal, the 2014 AAEA Meeting in Minneapolis and the 2014 NEUDC Conference in Boston for useful comments and suggestions. All remaining errors are our own. Final revision accepted: January 21, 2016. World Development Vol. 83, pp. 163–178, 2016 0305-750X/Ó 2016 Elsevier Ltd. All rights reserved. www.elsevier.com/locate/worlddev http://dx.doi.org/10.1016/j.worlddev.2016.01.023 163