Data-driven pathways to circular E-waste management

This study integrates environmental policy from the perspective of the circular economy (CE) concerning e-waste management in Canada, technological change, and regional governance through the application of the theory of Change (ToC). It applies ToC (Figure 1) as a spatially grounded analytical tool, highlighting the need for regionally differentiated pathways toward CE. By leveraging K-means machine learning (Figure 2) and NLP to mine scholarly knowledge, we identify the interdependencies, feedback loops, and multi-actor governance dynamics that shape e-waste transitions within and across Canadian provinces. Furthermore, it offers policymakers, managers, environmental agencies, and other stakeholders an efficient, transparent, and replicable tool for identifying leverage points across the e-waste value chain. By employing an innovative ToC framework informed by machine learning and natural language processing, we have mapped the multi-actor, multi-scalar, and multi-dimensional dynamics that shape regional pathways toward circularity. Our findings highlight the urgency of transitioning from fragmented, linear approaches to e-waste governance toward more integrated, inclusive, and regionally responsive systems of resource recovery.

While this study focuses on the Canadian landscape, our ML-informed ToC model has been created for cross-country adaptation. Our ToC model will allow other nations, predominantly those characterized by patchwork policies and jurisdictional fragmentation, to substitute the Canadian information with their own localized policy data. By using our ToC as a roadmap informed by NLP-derived thematic clusters, stakeholders in diverse contexts can identify their specific 'policy-to-action' gaps.

The Canadian case reveals both opportunities and barriers typical of advanced economies with strong extractive and industrial legacies. Despite its material wealth and policy expertise, Canada has yet to implement a cohesive national framework for circular e-waste governance—resulting in uneven capacities and inconsistent implementation at the sub-national level. In this context, regions are not simply administrative units but key arenas for experimentation, policy learning, and innovation. Effective circular transitions will require localized action plans, cross-sector collaboration, and the engagement of emerging actors such as youth entrepreneurs and community-based initiatives.

Conceptually, this research extends the application of Theory of Change to the domain by reframing it as a tool for exploring place-based transformation rather than merely evaluating linear interventions. The AI-assisted approach not only provides a scalable model for synthesizing complex knowledge systems but also supports evidence-based policymaking at the regional level.Ali, A., ShaabanNejad, S., Shirazi, F., Hajli, N. (2026).  Data-Driven Pathways to Circular E-Waste Management, Business Strategy and the Environment.