Assessing the sustainability of agri-food systems through Bayesian Networks applications: an exploratory study

Sustainability in the agro-food sector appears as a particularly complex issue due to the articulation of the sector structure, while several policies affect sustainability through interacting with this complex set of variables. Most studies analyse sustainability under the perspective of one or few individual indicators or determinants. However, there is a clear need for more holistic tools able to take into account the interconnections among the different components of the system. The objective of this paper is to discuss the potential use of Bayesian Networks tools to represent the connections between the different sustainability issues in the agri-food systems, by providing an empirical application based on a case study in Emilia Romagna. A Bayesian network has been built on the data obtained by a survey of 300 farms, focusing on the connections between networking choices and sustainability indicators. This experience shows that this tool can be helpful in representing complex relationship in the agro-food chain, taking into account a mix of information derived from survey data. However, ways of fully exploiting inputs from economic theory and from known technical relationships in association with this tool should be further investigated in order to provide more reliable and policy relevant simulations.