When does power disparity help or hurt group performance?

Power differences are ubiquitous in every social setting. However, the question of whether groups with higher or lower power disparity achieve better performance has thus far received conflicting answers. To address this issue, we identify three underlying assumptions in the literature that may have led to these divergent findings, including a myopic focus on static hierarchies, an assumption that those at the top of hierarchies are competent leaders, and an assumption that equality is not possible. We employ a multi-method set of studies to examine these assumptions. First, our agent-based simulation analyses show that the common implicit assumptions in power research can explain earlier disparate findings – power disparity only benefits group performance when it is dynamically aligned with leader task competence. Second, our empirical findings in both a field study fraud investigation groups and a multi-wave laboratory study corroborate the simulation results by showing that high power disparity helps group performance if the leader is competent and harms group performance if the leader is not competent. We thereby contribute to research on power by highlighting a dynamic understanding of power in groups and explaining how current implicit assumptions may lead to opposing findings.