~ The instruction we find in books is like fire. We fetch it from our neighbours, kindle it at home, communicate it to others, and it becomes the property of all ~ Voltaire
Social learning in agriculture refers to learning that takes place through social interactions between a farmer, and individuals in his social and economic network (aka “information networks”). Such learning can influence a farmer’s decisions at various stages such as choice of crops and inputs; method of input application; or any other form of technology adoption.
The purpose of this blog entry is to situate fieldwork observations on social learning in the context of existing theories and evidence on the importance of social networks in agricultural learning. According to researchers at Stanford University, contrary to popular belief, information about new farm technology and best practices do not flow directly from the lab to the farm. Farmers observe the decisions and experiences of their peers before adopting a new technology—a reason why agricultural technology adoption often occurs sequentially, and with a time lag (Munshi 2008). This blog entry discusses the process of social learning at length, and suggests the engagement of local information networks for successful introduction of new agricultural technologies.
My initial motivation for this blog post was a field trip undertaken last year to profile contract farming in Edar village of Gujarat, India. Several farmers in Edar have shifted cotton to potato cultivation under contracts with McCain, the suppliers of potato wedges and fries we eat at McDonalds. The farmers we interviewed use scientifically-treated potato cuttings (supplied by McCain at a cost) which meet the fast food chain’s length requirements. McCain’s agronomists ensure adoption of improved irrigation practices, weather monitoring, fertilizer application, and output grading norms. Cultivation under their assistance, each interviewed farmer reiterated, had augmented household incomes.
Yet, there is a puzzle here—the farmers we met had adopted the new technology sequentially despite being exposed to McCain’s uptake campaigns simultaneously. If the new potato cultivation arrangement under contract is indeed so profitable, why did Farmer 2 wait till Farmer 1 had reaped a profit on his potatoes to enrol under a contract, followed by Farmer 3 in the next cycle and so on…? Was this a singular case of sequential social learning I was witnessing or does existing literature hold theoretical and empirical precedent to such patterns of peer effects in agriculture? As it turns out, there is substantial, if not ample, literature that formalises reasons for lagged technology diffusion in agriculture. There are two broad explanations for the observed lag.
One, farmers differ in initial wealth, risk behaviour, education, creditworthiness, etc. (these differences are collectively termed “individual heterogeneity” in academic literature) which, in turn, determines their receptiveness to new technology and farming practices (Griliches 1957; Rogers 1962). Thus, the speed of technology diffusion varies across farms and results in an overall lag. It is possible that I witnessed farmer-level heterogeneity in Edar that November afternoon—perhaps, Farmer 1 had more personal and wealth (relative to Farmers 2, 3, 4, 5…) which aided his early foray into potato cultivation under contract with McCain. However, I was constrained by time and resources to profile each interviewed farmer’s individual characteristics, and have no means to verify if such heterogeneity was indeed the reason for the long-drawn adoption of a seemingly profitable contract farming arrangement in Edar.
Alternatively, to understand what may have motivated lagged technology diffusion in our case, the literature on social learning in developing countries (especially agriculture) is more compelling. Before I discuss theory and evidence on this subject, it is important to note that interviewed farmers 2,3,4,5…were on the same page when asked why they had delayed entry into contract potato cultivation—they were initially wary of returns under the new technology and cultivation arrangement, and waited to take cues regarding costs and revenues under the new system from Farmer 1.
Banerjee (1992) constructs a theoretical framework around such sequential patterns of social learning which fits perfectly with our observed behaviour in Edar. A detailed discussion of the said theory is beyond the scope of this blog post, but the basic intent of the model is quoted below:
We analyze a sequential decision model in which each decision maker looks at the decisions made by previous decision makers in taking her own decision. This is rational for her because these other decision makers may have some information that is important for her. We then show that the decision rules that are chosen by optimizing individuals will be characterized by herd behaviour; i.e., people will be doing what others are doing rather than using their own information.
Thus, given that information flows sequentially from one farmer to another, and there is a typical time lag in agriculture between technology adoption and harvest, it is now possible to contextualise the episode of social learning I witnessed in Edar in light of existing theories and empirics on social learning in developing countries. Munshi (2008) says that while theoretical research on social learning has made significant strides in recent years, empirical evidence on the same is limited due to econometric difficulties in studying social networks (“identification of social learning is a challenging problem”, pg 3111).
Some interesting evidence is offered by Conley & Udry (2005) on technology diffusion among pineapple growers in Ghana; and Bandiera & Rasul (2006) on social learning among farmers in northern Mozambique. Conley & Udry (2005) observe that farmers increase (decrease) fertilizer use when their “neighbours” using more (less) fertilizer than them make unexpectedly high profits. Bandiera & Rasul (2006) establish interesting relationships between the number of peer-farmers who adopt a new technology and the individual farmer’s own decision to adopt. Their main inferences are quoted below:
On the one hand, an increase in the number of adopters in the individual’s information network increases the amount of (social) information that is made available, increasing his propensity to adopt. On the other hand, having many adopters in the network increases the individual’s incentive to delay adoption and free-ride on the information that is made available by his neighbours. It can be shown that the second effect dominates once the adoption level within the network grows beyond a cut off point.
These papers inform us of the existence and extent of social learning in agriculture but not necessarily of its benefit(s). One of the few papers to simulate the impact of social learning on agriculture is Foster & Rosenzweig (1995)—the authors infer that imperfect knowledge about management of new HYV seeds among a sample of rural Indian households is detrimental to its adoption. However, with cues from neighbours’ (and own) experience, profitability from HYV seeds increases significantly. In light of this evidence, the year-on-year income augmentation of McCain’s potato farmers is perhaps a part reflection of the benefits of social learning in agriculture.
In sum, given the importance of learning prompted by social networks in agrarian settings, one would be stating the obvious in recommending that local information systems should be used purposefully and strategically to enhance technology adoption in agriculture.
The Author and Research Assistant on the project is Sourovi De, University of Oxford. The findings, interpretations and conclusions expressed herein are those of the authors and do not necessarily reflect the view of Global Development Network or its Board of Directors.
You can follow Sourovi De on Twitter @Sourovi
Bandiera, O., & Rasul, I. (2006). Social networks and technology adoption in northern Mozambique. Economic Journal, 116.
Banerjee, A. (1992). A simple model of herd behavior. Quarterly Journal of Economics, 117 (3).
Conley, T., & Udry, C. (2005). Learning about a new technology: Pineapple in Ghana. Mimeo.
Foster, A., & Rosenzweig, M. (1995). Learning by doing and learning from others: Human capital and
technical change in agriculture. Journal of Political Economy, 103(6).
Griliches, Z. (1957 ). Hybrid corn: An exploration in the economics of technological change. Econometrica,
Munshi, K. (2008). Information Networks in Dynamic Agrarian Economies. In P. Schultz, & J. Strauss, Handbook of Development Economics, Vol. 4 (pp. 3086-3113). Oxford: Elsevier Science.
Rogers, E. (1962). Diffusion of Innovations. The Free Press, New York. New York: The Free Press