Data-driven approaches to agricultural production are being touted for their potential to revolutionize smallholder agriculture. This "datafication" (Cukier and Mayer-Schoenberger 2013) process is already evident in the Global North, where the agricultural sector is being reorganized by “smart” and “precision” data-intensive technologies (Wolf and Wood 1997). However, a parallel movement can be seen at work in the Global South in the suite of approaches collectively termed "ICT for Ag." Private companies, international development organizations, philanthropic foundations, and non-governmental organizations are embracing the potential of digital data to generate profit and foster agrarian development in Africa, Latin America, and other parts of the developing world. Their initiatives promote flows of digital data both to and from small farmers; many involve collecting data from farmers—such as farm location, crop and variety, inputs applied, yields, farm expenses—and most offer better data to farmers—including market prices, weather forecasts, soil data, input recommendations. Frequently, data collection and dissemination go hand-in-hand, as in cases when location and operations information collected from farmers is combined with remotely sensed weather and soil data to provide them with recommendations for seed varieties and fertilizer application. These ag-data initiatives are promoted for their potential to transform the lives of poor farmers by helping them increase their productivity, tap into new markets, adapt to climate change and more.
However, celebratory narratives about the revolutionary potential of digital agricultural data may obscure the technology’s potential to shape food system vulnerabilities. Combining insights from critical agrarian studies and critical data studies, this paper examines the political implications of the shift toward more data-intensive practices. Drawing from over 40 interviews with ag-data practitioners, as well as participant observation at development sector conferences focused on agricultural data, we offer a series of critical interventions for further research. First, we urge reflection on how data-intensive farming changes the nature of agricultural knowledge, legitimizing some ways of knowing while devaluing others. The long-standing tendency of agricultural experts to dismiss peasant and indigenous knowledge is exacerbated by their exclusion from the development of ag-data platforms and classificatory systems. Second, we problematize how data-based solutions may deepen market penetration of agrarian economies by promoting chemical- and capital-intensification and by increasing peasant visibility to corporate actors. Third, we suggest that privacy, consent, and data ownership are receiving insufficient attention, a finding consistent with emerging research on “digital humanitarianism” (Crawford and Finn 2015) in other realms. These preliminary research findings suggest a need for more research on how "agro-digital assemblages" (Carolan 2016) are being mobilized in the Global South, leading to the creation of new data-producing farmer subjects and data-enabling agrarian environments.
