Leadership structure and performance in dynamic contexts: A multidimensional network approach ¹

Abstract While leadership structure is recognized as a key mechanism for enabling teams to adapt to dynamic environments, there remains limited understanding of which leadership structural configurations yield optimal performance under such contexts. To address this gap, we draw on adaptive leadership theory and adopt a multidimensional network approach to examine how leadership structure influences team performance in dynamic contexts. Specifically, this study considers not only the dispersion (i.e., centralization) and magnitude (i.e., density) of leadership but also the transition of leadership (i.e., centrality variance). A field study with a sample of 115 entrepreneurial project teams was conducted to test the hypotheses. The findings reveal that, in comparison to the other two basic configurations, the leadership structure with high centralization, low density, and high centrality variance is associated with the lowest level of status conflict, while the structure with high centralization, low density, and low centrality variance is associated with the highest level of status conflict. This three‐way interaction further exerts an indirect effect on team performance through status conflict. To enhance the validity of the findings, supplementary analysis was performed using qualitative data collected from five new venture teams, which provided additional support for the hypotheses. This study advances theoretical understanding of leadership structures in dynamic environments and offers actionable insights for fostering effective team adaptation.