Does information and communication technology (ICT) based provision of agricultural extension services help improve agricultural productivity in poor or developing countries? We answer this question in the case of rice production in rural Bangladesh. We exploit the spatiotemporal variation in the availability of
village-level phone services and the temporal variation in the timing of an ICT-based intervention to identify the differential impact by input use, network centrality, and geographic proximity. We observe that, in the villages with access to phone service, there is a 50 percent reduction in plot-level inefficiency after the intervention, driven by plots that used rainfed water for cultivation. We provide evidence suggesting that these effects are due to increased input use by the farmers using rainfed farming. Our results also document that the intervention benefits geographically remote farmers differentially more, whose information needs are otherwise unfulfilled by traditional extension services. However, the diffusion of information via networks remains relevant as we document significant cross-community spill overs through
geographic ties.
