How place, scale and governance capacity shape the evolution of circular economy initiatives

Regional circular economy (CE) transitions remain under-examined compared with urban and industrial contexts, despite growing recognition that place, scale and governance configurations strongly shape circular pathways. Drawing on a mixed-methods investigation of two Australian States (Victoria and Tasmania), including a comparative analysis of CE actors and activities and exploratory case studies in three regional local governments, this paper examines how regional capacities, governance networks and institutional conditions influence the evolution of CE initiatives. The findings reveal two systemic patterns: (1) CE activity in both States was concentrated on long-loop recycling and waste-management functions, with minimal medium-loop (repair, remanufacture) capability and modest short-loop social and behavioural initiatives; and (2) actor ecosystems varied markedly, with Victoria exhibiting broad actor diversity while Tasmania showed near-complete dependence on local governments, indicating a form of regional governance centralisation under constraint. Regional case studies showed, local governments act as promoters, enablers and partners, but their agency is mediated by crisis signals (e.g., impending landfill closures), and regional development imperatives. Social enterprises and community organisations played significant roles in short-loop activities and people-centred services, yet their capacity to scale was limited by inadequate infrastructure and weak industrial bases. These results demonstrate how structural economic geography shapes the feasible boundaries of CE transitions. CE policy in regional contexts like Australia – export oriented, thin markets and institutions, with limited manufacturing bases and at the extraction end of global supply chains - requires place-sensitive strategies that build medium-loop capacity, diversify actor ecosystems and institutions, and recalibrate CE models beyond urban-industrial assumptions.