Explaining Variation in the Agency of Transnational City Networks: From Institutional Architecture to Embeddedness in EU and UN Climate Governance

Contemporary climate governance no longer operates through a centralized, vertical state hierarchy. Instead, it has evolved into a horizontal, indirect, and transnationally interconnected ecosystem in which authority and coordination are dispersed across levels and actors. Transnational city networks (TCNs) have emerged as key connective infrastructures within this ecosystem. Operating from the side of formal authority, they aggregate data from cities worldwide, observe policy implementation across territories, and channel information upward and across governance levels. In doing so, TCNs hold significant potential to stabilize fragmented governance arrangements, identify systemic gaps, and support scaling of climate innovations, particularly in governance systems increasingly constrained by geopolitical conflict and domestic backlash. Yet this potential is neither automatically nor fully utilized. While TCNs’ roles have been actively welcomed and institutionalized by governance authorities, we still lack a systematic understanding of how this institutional embedding is organized and how it shapes TCNs’ agency across different multilevel governance (MLG) contexts. This paper addresses this gap by advancing a governance-oriented institutional perspective that moves MLG research toward a causal theory of action capacity. It examines how institutional factors in MLG systems are associated with differentiated modes of TCN agency through the lens of institutional embeddedness. Institutional embeddedness is understood as the structurally fixed positioning of TCNs within governance architectures that conditions action. In this sense, the institutionalization of embeddedness is approached as a factor that may stabilize participation while simultaneously constraining the scope and form of agency available to TCNs. Building on this analysis, the paper further examines how variation in MLG regimes generates distinct embeddedness–agency configurations. From this perspective, rigid and flexible MLG regimes are understood as complementary opportunity structures. Rigid regimes tend to stabilize participation and channel agency into procedural and rule-based forms, while more flexible regimes create space for experimentation, innovation, and boundary-pushing strategies. Crucially, it is TCNs’ agency that enables complementarity connecting these otherwise fragmented political spaces. By operating simultaneously across governance contexts, TCNs create and sustain linkages between the European Union and the United Nations. Through their involvement in both arenas, TCNs carry practices, ideas, and governance innovations across institutional boundaries and test them at the local level. Implementation experiences are subsequently circulated back to higher levels of governance, enabling cross-regime learning, recombination, and diffusion rather than parallel and disconnected governance trajectories. Empirically, the paper draws on a qualitative comparative research design based on document analysis and interviews. It focuses on the European Union as a key engine of decarbonization and benchmark for climate governance, and on the United Nations as the global arena where climate policies are forged. As for TCNs, the analysis examines the same two networks operating simultaneously within the EU and UN governance contexts, enabling a structured comparison of how identical actors navigate different MLG configurations.