Towards a behavioural system dynamics: Exploring its scope and delineating its promise

• Published in: European journal of operational research Vol. 306; no. 2; pp. 777 - 794
• Main Authors: Lane, David C., Rouwette, Etiënne A.J.A.
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 16.04.2023
• Subjects: Behavioural or; Intervention; System dynamics; Intervention; Behavioural or; System dynamics
• DOI: 10.1016/j.ejor.2022.08.017
• ISSN: 0377-2217

•Critically evaluates the ‘behavioural turn’ in OR and re-visits the core ideas.•Applies to system dynamics the behavioural perspective of biases and deviations from rationality.•Presents a new framework for system dynamics studies which reveals a range of behavioural effects.•Proposes five ‘Behavioural SD’ axioms involving interdisciplinarity and experiments.•Makes the case that ‘Behavioural SD’ can explain behavioural phenomena and improve model-based interventions. Attempts to examine so-called ‘behavioural effects’ have reached into many different fields and are in full sway across OR. This paper considers whether this can be applied to System Dynamics modelling in a useful way. The idea is that humans frequently do not employ strict rationality in their daily lives but make errors and are subject to fallacies. This can shed light on the cognitive and human interaction aspects of the process and outcomes of modelling. The paper first raises concerns about the current state of ‘Behavioural OR’ – BOR. To refocus the underlying ideas it then returns to the Decision Theory roots, takes a broader view using an illustration from the history of science and then builds on work which first proposed a link to System Dynamics (SD). The core of the paper then explores in depth how behavioural ideas apply to SD. This is done using examples dealing with complex systems in a ‘naïve’ versus a ‘sophisticated’ way, and then using a mind map. It then offers a new and detailed framework of the stages of an SD-based intervention, indicating the presence of behavioural effects and providing a fine-grained discussion of those effects as they apply to SD. It builds on this by proposing a definition of ‘Behavioural System Dynamics’ (BehSD) in terms of its perspective on phenomena, five new, constitutive axioms, and its potential for improving practice. The paper closes by reprising the nature and potential of ‘BehSD’ and by sketching how research that adopts this perspective might go forward.