When a team is told it will run without a single designated leader, the reaction is often a quiet eye-roll. It sounds like the kind of arrangement that produces warm meetings and slow decisions. In an era of flatter org charts and agile squads, that skepticism deserves a straight answer rather than a slogan, and the research provides one. Shared leadership is not the absence of leadership. It is leadership performed by more than one person, and it shows up in performance data.
From heroic to shared
The conventional model locates leadership in a person: the manager, the lead, the one whose name is on the org chart. Self-management redistributes that function. In self-managing organizations, leadership is contextual and distributed across roles rather than concentrated in individuals, which is part of what it means to eliminate the manager-to-subordinate relationship formally and systematically (Lee and Edmondson, 2017). The obvious objection is that leadership done by everyone is leadership done by no one. If that were true, teams with shared leadership would underperform teams with a clear single leader. They do not.
A meta-analysis of shared leadership in teams found a positive relationship between shared leadership and team performance (Nicolaides, LaPort, Chen, Tomassetti, Weis, Zaccaro and Cortina, 2014). This is not a single suggestive study. It is a synthesis across many, which is why it carries weight against the intuition that distributed authority must be soft. Teams in which leadership functions are genuinely shared tend to perform better, and the effect is robust enough to survive pooling across studies. That is a meaningful bar to clear. A pattern that holds only in a favorable single case can be dismissed as an artifact, but one that persists across many studies is telling you something about how teams actually work, not about the enthusiasm of a particular author.
The mechanism is trust, not vibes
What makes the finding useful is that it does not stop at the correlation. The relationship between shared leadership and performance was partially mediated by the team's proximal emergent states, things like trust and consensus, rather than being a purely direct effect (Nicolaides and colleagues, 2014). That distinction matters enormously in practice. It means shared leadership does not lift performance by magic. It lifts performance by building trust and shared understanding within the team, and those states then drive results. Get the emergent states and the benefit follows. Fail to build them and shared leadership is just a diffuse chain of command with no advantage.
This is precisely the sort of thing critics miss when they treat distributed authority as wishful thinking. The mechanism is specific and, importantly, buildable. Trust and consensus are not moods that either happen or do not. They are outcomes of how a team communicates, resolves disagreement, and holds itself accountable, which means the performance benefit of shared leadership is something an organization can deliberately engineer.
The same signal from agile research
A second body of evidence points the same way. A systematic literature review of how agile teams reach effectiveness synthesized the enabling factors and put shared leadership among them, alongside autonomy, coordination, and the team's emergent states (Steegh, Van De Voorde and Paauwe, 2025). The review's larger argument reinforces the point: agile effectiveness depends on principles of self-management, not merely on running the ceremonies (Steegh, Van De Voorde and Paauwe, 2025). Teams that adopt the rituals of agile but keep authority concentrated in one person are missing the mechanism that the evidence says actually drives effectiveness. Shared leadership is not decoration on top of self-management. It is one of the load-bearing walls.
Put the two literatures together and the picture is coherent. Distributed authority relates positively to performance (Nicolaides and colleagues, 2014), it does so through trust and consensus (Nicolaides and colleagues, 2014), and it shows up as a named enabler in the practical domain where most companies actually attempt self-management, which is agile teamwork (Steegh, Van De Voorde and Paauwe, 2025). None of that describes a feel-good idea. It describes a measurable practice with a known mechanism.
The practical read
If you are distributing leadership across a team, do not stop at removing the single leader and hope for the best. That gets you the diffuse structure without the benefit. Invest directly in the emergent states the research identifies as the mechanism, building trust and working toward genuine consensus, because those are what convert shared leadership into performance (Nicolaides and colleagues, 2014). Treat shared leadership as a core enabler of team effectiveness rather than a side effect of going flat (Steegh, Van De Voorde and Paauwe, 2025). And drop the assumption that leadership without a designated leader is automatically weaker. The evidence says it can be stronger, provided you build the trust that makes it work. Leadership spread across a team is not the absence of the thing. It is the thing, done by more hands, and measured.
Sources
- Nicolaides, LaPort, Chen, Tomassetti, Weis, Zaccaro and Cortina (2014). The shared leadership of teams: A meta-analysis. Finds shared leadership positively related to team performance, partially mediated by proximal emergent states like trust and consensus. https://doi.org/10.1016/j.leaqua.2014.06.006
- Steegh, Van De Voorde and Paauwe (2025). Understanding how agile teams reach effectiveness. Synthesizes shared leadership among the enablers of agile effectiveness and argues effectiveness rests on self-management principles. https://doi.org/10.1016/j.hrmr.2024.101056
- Lee and Edmondson (2017). Self-managing organizations: Exploring the limits of less-hierarchical organizing. Establishes that in SMOs leadership is contextual and distributed across roles rather than held by individuals. https://www.sciencedirect.com/science/article/abs/pii/S0191308517300059