Causal network maps of urban circular economies.

Urban systems have a central role in the transition toward circular economy. Systematic analysis of drivers is needed because of the complex interplay of social, economic, and political factors. Such analysis requires a good understanding of direct and indirect influences on urban circularity. Because of the presence of indirect influences, visualizing the causal networks is necessary for systematic analysis. Existing methods for formulating causal network maps (CNMs) rely on subjective approaches which inhibit robust assessment. The generation of robust CNMs can provide more accurate representation of direct and indirect influences. Therefore, this study generates a robust CNM for drivers of urban circular economy through a hybrid decision-making and trial laboratory-fuzzy cognitive map (DEMATEL-FCM) framework. DEMATEL is used for building the initial structure of the network map. The network is then trained using FCM with data obtained from the Sustainable Cities Index. A 70:30 training–testing ratio is used to partition the training and testing datasets. The trained CNM has 92.75% accuracy during training and 96.77% accuracy during testing. The trained CNM provides an empirical depiction of driver interrelationships in urban circular economies. It indicates the importance of 'affordability' and 'economic development' in the network structure. The network yields significant insights for the development of city-level plans and policies to stimulate a transition to a more circular economy. Data-driven visualization of interactions among drivers give stakeholders insights on the most effective measures to implement. [ABSTRACT FROM AUTHOR]