Evolving Structures of Interagency Collaboration in Disaster Response: A Social Network Analysis of Cyclone Idai

The complexity of large-scale disasters requires coordinated action among diverse humanitarian organizations. This study examines the interagency collaboration during the response to Cyclone Idai using Social Network Analysis to trace how organizational relationships evolved. Based on ReliefWeb reports, we build temporal networks for early, mid, and late phases of the crisis, linking agencies and reports in a bipartite structure projected into agency-only networks. Metrics such as degree centrality, betweenness, community structure, and role assortativity are used to identify key actors and structural patterns. The results show a shift from centralized coordination led by major UN agencies to more decentralized and specialized collaborations during recovery. Using Complex Adaptive Systems theory, we interpret these changes as signs of emergent, adaptive behavior within the coordination system. The study highlights the importance of inclusive and flexible network structures in disaster governance and emphasizes the joint role of global and local actors in achieving resilient outcomes.