Farmers' Information Management in Developing Countries-A Highly Asymmetric Information Structure

In developing countries, governments, non- governmental organizations, and social entrepreneurs are disseminating agricultural information to farmers to improve their welfare. However, instead of having direct access to the information, farmers usually acquire information from local social networks, and, thus, they may have very different information channels. We establish a general framework that accommodates highly asymmetric information structures to study farmers' information management and utilization problems. In our model, a bipartite graph describes which subset of signals is accessible to a farmer. We characterize a unique Bayesian Nash equilibrium and express farmers' strategies and expected profits in closed form. We discuss properties of this equilibrium and show that asymmetric information structures can lead to various novel results. For example, a farmer may produce more (less) when observing a pessimistic (optimistic) signal, may benefit from the improvement of a signal she cannot observe, may want to share her signal with others, and may become worse off when another farmer releases a signal to her. We conduct comprehensive studies on the equilibrium in the weak signal limit, where signals are subject to substantial noise. We examine the government's optimal information allocation in this limit when its goal is to maximize farmers' total profits or the social welfare. To improve farmers' total profits, the government should provide all its information to (and only to) one farmer. We establish an index to determine which farmer should get the information. In contrast, to maximize the social welfare, the government should provide all its information to all farmers.