IMPROVING COMMUNICATION IN DROUGHT PREPAREDNESS CAMPAIGNS IN IRRIGATION SYSTEMS: A Network Analysis

One common finding in recent diffusion research is that ideas, beliefs, and behaviors tend to diffuse more quickly in disassortatively than in assortatively mixed social networks. This outcome may have important implications for the design of communication strategies to increase the participation of farmers in drought preparedness programs. For instance, if the underlying social network is disassortatively mixed by degree, i.e., if highly connected farmers tend to preferentially associate to less connected ones, community involvement programs should focus on convincing opinion leaders on the need to adopt the recommended measures. If, otherwise, the network is assortatively mixed by degree, communication should be more focused on horizontal campaigns targeted to the average farmer. We tested a typical irrigation system in the semi-arid region of Brazil for degree and geographic assortativity. The main finding was that the underlying informational network is disassortative by degree and assortative by geographic location of the farmers. This result suggests that community involvement campaigns in similar systems should be designed based on some variation of the basic “two-step flow ” model of communication, in which most of communication and marketing researches are based on. Specifically, it appears that policies seeking to improve the response of irrigation systems to severe droughts should focus on influencing local leaders able to act as mutually reliable information bridges between irrigators and government agencies