Return‑to‑office edicts, AI whiplash, and rolling supply shocks have made one thing obvious: decisions that wait for head office often arrive after the moment has passed. The cost shows up in people as much as in P&L. Global employee engagement slid to 20 percent in 2025, with managers’ engagement falling from 27 to 22 percent in a single year, and Gallup estimates lost productivity at about 9 percent of global GDP, or 10 trillion dollars (Gallup, 2025). A design that traps choices at the top is not only slow. It drains energy from the layer expected to translate strategy into action.
Why distributed calls win in turbulence
The strongest field evidence we have comes from the last major shock. Firms that had delegated more decision authority from headquarters to local plant managers before the Great Recession fared better in the sectors hardest hit by the crisis (Aghion, Bloom, Lucking, Sadun and Van Reenen, 2021). Using two large micro datasets from the United States and other OECD countries, the authors show that decentralization was associated with stronger performance during bad times. The mechanism matters. In turbulence, local information becomes more valuable than centralized coordination, a pattern the authors corroborate with measures such as product churn and stock market volatility. The results are robust to alternative explanations like fear of bankruptcy or generic coordination changes, which strengthens the case that information proximity was the edge.
Read that carefully. The advantage did not come from a general love of autonomy, and it did not say decentralization always wins. It was contingent on the environment. When shocks raise uncertainty and shorten the half‑life of facts, the place closest to the facts is the right place to decide. When conditions are stable, standardization and central control can drive consistency and scale.
What to centralize, what to push out
This is where theory helps draw the line. Self‑managing organizations decentralize authority in a radical, formal, and systematic way, eliminating the manager‑subordinate relationship across the enterprise (Lee and Edmondson, 2017). Few firms need to go that far. The broader synthesis in the same work is the useful guide: managerial hierarchy performs well for reliable execution of known tasks in stable environments, but it inhibits complex, nonroutine problem solving that crosses boundaries. That gives leaders a practical rule. Centralize the known, repeatable, and regulated. Distribute the novel, cross‑functional, and fast‑moving.
Treat this as a design choice, not a culture poster. The empirical base for fully self‑managing systems is still limited, and much of it is case‑based rather than experimental (Lee and Edmondson, 2017). The causal evidence we do have is strongest on the specific claim that decentralization helps in downturns because of local information advantages (Aghion et al., 2021). A dual design follows from these two strands without overreaching: keep a spine of centralized standards, and build a switch that shifts more authority to the edge when turbulence rises.
The human precondition: psychological safety
Distributing decisions without changing interpersonal climate is a recipe for quiet fear. Edmondson’s field study with 51 teams in a manufacturing company established the construct of team psychological safety as a shared belief that a team is safe for interpersonal risk taking (Edmondson, 1999). In the data, psychological safety correlated with learning behaviors such as speaking up about errors and seeking feedback, while team efficacy did not predict learning when safety was accounted for. Learning behavior mediated the link between safety and performance. The implication for distributed decision‑making is straightforward. Local teams will only surface weak signals, admit uncertainty, and iterate in the open if they believe doing so will not be punished.
In practice, that means senior leaders must pair any shift of decision rights with explicit protection for candor and experimentation. Absent that, decentralization can harden into local silos that hide problems rather than solve them. Psychological safety is not a soft add‑on. It is the rate limiter on how fast and how far you can push decisions to the edge.
A playbook you can run next quarter
Leaders do not need a full rewrite to act. Start with a switch, not a revolution.
- Map decision types by information half‑life. Classify recurring, compliance‑laden, and brand‑critical calls as centralized. Classify environment‑facing, novel, or cross‑functional calls as distributed. This mirrors the evidence that hierarchy fits stable execution while decentralization fits nonroutine problem solving (Lee and Edmondson, 2017).
- Pre‑commit turbulence triggers. Define in advance the metrics that shift authority outward, inspired by the turbulence measures used in the Great Recession research. Product churn and market volatility were informative in that work. Use analogous indicators in your sector to toggle to local decision rights when uncertainty spikes (Aghion et al., 2021).
- Write decision rights, not job titles. Specify who decides, who must be consulted, and who is informed for the distributed category. Headquarters sets the frame and guardrails, local units make the call within them. Clear rights keep speed from turning into chaos.
- Build the safety net for learning. Ask teams to run short learning cycles and report what they tried, what they learned, and what they will change next. Then model the response you want by rewarding disclosure of missteps. This aligns with evidence that team learning behaviors, enabled by psychological safety, are the path to performance (Edmondson, 1999).
- Redefine the manager’s job. The Gallup data show managers are less engaged than before, with a sharp drop since 2024 (Gallup, 2025). Treat this as a design signal. Shift the role from gatekeeping decisions to coaching clarity, enabling information flow, and removing obstacles. This is an interpretive step, but it follows from moving authority toward the information and from the engagement gap as a warning light.
- Test in one product, plant, or region. The strongest evidence comes from firms that had already decentralized before the shock hit (Aghion et al., 2021). Pilot now, so that when the next surprise arrives, the switch is already wired.
The return shows up where boards look. In bad times, firms that had delegated more before the downturn outperformed centralized peers in the sectors that took the biggest hits (Aghion et al., 2021). The people side points the same way. Engagement is at a low ebb globally, and manager engagement in particular is lagging badly, which signals a need to redesign how decisions and accountability travel through the organization (Gallup, 2025). The theory behind self‑management clarifies the boundary conditions. Use hierarchy to run the known play well. Use distributed judgment to find the next play when the field shifts (Lee and Edmondson, 2017).
You do not need another slogan. You need a switch. Install it before the weather turns again, pair it with psychological safety, and make decisions where the information lives.
Sources
- Aghion, Bloom, Lucking, Sadun and Van Reenen (2021). Turbulence, Firm Decentralization, and Growth in Bad Times. Firms that delegated more before the Great Recession outperformed in hard‑hit sectors, with the mechanism tied to the value of local information under turbulence. https://www.aeaweb.org/articles?id=10.1257%2Fapp.20180752
- Lee and Edmondson (2017). Self-managing organizations: Exploring the limits of less-hierarchical organizing. Defines SMOs and synthesizes when hierarchy suits stable execution versus when decentralization aids complex, nonroutine problem solving. https://www.sciencedirect.com/science/article/abs/pii/S0191308517300059
- Edmondson (1999). Psychological Safety and Learning Behavior in Work Teams. Establishes psychological safety and shows that learning behaviors mediate its link to performance, a key precondition for effective distributed decisions. https://journals.sagepub.com/doi/10.2307/2666999
- Gallup (2025). State of the Global Workplace 2025. Reports global engagement at 20 percent, a sharp drop in manager engagement, and an estimated 10 trillion dollars in lost productivity, underscoring the cost of current organizational designs. https://www.gallup.com/workplace/659610/state-global-workplace.aspx