A ‘Theory of the Middle Range’ to Support Food Security and Circular Economy Value Chain Scenario Analysis

A common critique of agent-based models is that their relationship with theory is tenuous. However, arguments have been made showing both that agent-based models can draw on theories, and that agent-based models can be treated as though they are theories. At the same time, empirical agent-based models, increasingly used for policy evaluation and scenario analysis, tend to be fitted to specific cases rather than having any sense of generality to their application. This lends weight to the criticisms of the relationship between agent-based models and theory. To address this issue in a project in which we are tasked with creating an agent-based model to evaluate scenarios in ‘farm-to-fork’ value chains, with potential applications to more general circular economy case studies, we have endeavoured to design a ‘theory of the middle range’ (à la Merton) that we believe to have the necessary capability, together with a formal implementation in the form of an agent-based model. We describe the model here, giving critical attention to its proposition as a ‘theory’ (of any kind) and potential generality, with a view to furthering the debate about the relationships between empirical agent-based models and social/scientific theories.