Endogenous Drivers of Cascading Rebound: Mapping Feedback Loops in Digitally-enabled Sharing Platforms

Monday, July 20, 2026, 11:00 AM

Session: SOC Poster Session (In-person)

Digitally-enabled sharing platforms (DSPs) are increasingly promoted as potential enablers of circular economy transitions. Yet, growing evidence shows that DSPs can trigger rebound effects—systemic responses that offset or reverse anticipated sustainability gains. Existing research typically conceptualizes rebound mechanisms in isolated, static, and linear terms, overlooking how they emerge from endogenous, feedback-driven interactions across micro, meso, and macro levels. In response, this study adopts a system dynamics perspective to investigate how rebound mechanisms arise from the underlying, inherently complex system structure of DSPs. Building on a systematic literature review followed by a Gioia-informed coding process, we identified key variables, causal relationships, and partially described mechanisms. These were translated into three integrated causal loop diagrams (CLDs), which articulate the dominant reinforcing loops driving consumption dynamics, ownership-related dynamics, and socially mediated dynamics. This study shows that rebound effects in DSPs result from the system's structure—including their multi-level nature. More specifically, our CLDs reveal how feedback loops interact across micro, meso, and macro levels, showing how these rebound effects cascade across the complex system. It uncovers interdependencies, nonlinearities, and feedback structures, explaining why rebound effects persist. The resulting CLDs provide a foundation for future quantitative system dynamics modeling, enabling exploration of leverage points, counteracting balancing loops, and policy interventions that can alter system behavior. This work contributes to research on rebound effects by clarifying how DSP-induced rebound effects unfold endogenously through complex, multi-level feedback processes and how systems dynamics can inform this understanding.