Behavior‐Informed Decision Support for Urban Plastic Waste Management: A Hybrid Fuzzy DEMATEL‐Bayesian Network Approach

Plastic waste management in rapidly urbanizing regions of the Global South is shaped by informal systems, uneven infrastructure, and governance variability. Addressing this requires moving beyond infrastructure‐focused policies to incorporate behavioral determinants. This study develops a behavior‐informed decision‐support framework for urban plastic waste governance by integrating the COM‐B model (Capability, Opportunity, Motivation‐Behavior) with the theory of planned behavior (TPB). This approach advances waste behavior modeling by linking structural feasibility with intention formation. Capability and opportunity are captured through COM‐B, while the TPB explains how intentions develop. Their combination helps address the intention‐behavior gap that remains unresolved in single‐focus models. Using survey data from 415 respondents across three cities in Kerala, India, construct validation supported the integrated behavioral framework. Fuzzy DEMATEL analysis identified capability and opportunity as primary causal drivers of motivation and intention, with intention as the strongest predictor of pro‐environmental behavior. A Bayesian Network (BN), structured on DEMATEL outputs, was employed to model probabilistic dependencies and simulate intervention scenarios. These indicate that jointly enhancing capability and opportunity increases adoption of core waste‐reduction behaviors by approximately 20%, while sustainable purchasing shows only a modest improvement. The findings highlight that effective plastic waste governance requires synergy between motivational interventions and enabling infrastructure to translate intentions into practice. Policy recommendations emphasize strengthening community‐based collection systems, expanding decentralized material recovery facilities, and aligning market reforms with local governance structures. This hybrid framework demonstrates how combining causal mapping and probabilistic modeling supports evidence‐based and transferable strategies for urban waste management in the Global South.