Return-to-office edicts, election-year rancor, and AI anxiety have a common root: distrust. Employees are tuning out. Global engagement has dropped to 20 percent, the lowest since 2020, and Gallup estimates the productivity loss at about 9 percent of global GDP, roughly 10 trillion dollars (Gallup, 2025). Managers are not immune. Their engagement slid five points in a single year to 22 percent, a sharp fall since 2022 (Gallup, 2025). Meanwhile, seven in ten people report an insular mindset, reluctant to trust those with different values or backgrounds, yet they still look to their employers to mediate trust across divides through training and structured interactions (Edelman, 2026). The message is direct. Trust cannot be left to vibes or charisma. It needs to be a default policy, and disagreement needs a design.
Why default trust is a business decision, not a courtesy
Survey data do not establish causality, but the pattern is hard to ignore. Engagement is in retreat, and the drag on productivity is large and persistent (Gallup, 2025). At the same time, people say they are hesitant to trust those unlike them, and they identify employers as the actor best placed to broker trust at scale by teaching conflict resolution and creating structured contact among people with different views (Edelman, 2026). That combination creates both duty and advantage. Firms that codify trust reduce the friction costs of second-guessing and approval-seeking. They also meet a social need that no other institution appears ready to meet.
The language of policy matters. Trust by default does not mean the absence of checks. It means the presumption of competence and good faith unless evidence shows otherwise, paired with explicit mechanisms to surface and resolve disagreement. Employers hold a unique position here, with the smallest gap between what people expect and what they deliver on trust compared with other institutions, according to the Edelman data (Edelman, 2026). If trust is infrastructure, employers own the construction site.
The mechanism that turns trust into performance: psychological safety
Teams learn when people can speak up, ask for help, and admit error without fear of embarrassment or punishment. That is not a slogan, it is the construct of team psychological safety as defined in a seminal field study of 51 work teams in a manufacturing firm. Edmondson found that psychological safety was positively associated with learning behaviors, and that these learning behaviors mediated the link to performance. Team efficacy alone did not explain learning once psychological safety was included (Edmondson, 1999). The study is cross-sectional and limited to one company, so it does not prove causation across all contexts. Still, it is the foundation of a large literature and a practical guide for managers who want trust that ships.
This is where default trust earns its keep. When the default is to assume good intent and honest error, people share bad news earlier, ask naive questions, and correct course faster. That is learning behavior in action, and it is the only path to sustained performance in uncertain conditions. In low-trust settings, the same mistakes arrive late and cost more. A policy that instructs leaders to respond first with curiosity, not blame, is not soft. It is an operating rule that creates the conditions under which teams do the learning work that predicts performance over time (Edmondson, 1999).
Flatten where problems are complex, and be honest about limits
Hierarchy is good at repetition. It routes decisions through managers who can enforce standardization. The management literature synthesized by Lee and Edmondson shows that this design works for reliable execution of known tasks in stable environments, but it tends to inhibit cross-functional problem solving, create rigidity, and reinforce status differences that stunt development (Lee and Edmondson, 2017). Self-managing organizations represent the opposite bet. They decentralize authority in formal, systematic ways and remove the manager-subordinate relationship as the default. That does not mean a free-for-all. It means clear roles, explicit decision rights, and peer-based governance in place of permission chains (Lee and Edmondson, 2017).
The evidence here is conceptual and case-based rather than causal, and the authors are explicit about the limits. Not every context benefits from radical decentralization. The practical read is narrower and more useful: where work is novel and crosses boundaries, flattening decision rights and clarifying who decides what will usually surface better information faster. That is designing for disagreement. It puts debates at the level where the knowledge lives and removes the career risk of speaking up by defaulting to trust in peers’ intent and competence.
A policy you can ship next quarter
You do not need to copy a self-managing archetype to set a default of trust and a design for disagreement. Four moves fit inside one planning cycle:
- Publish decision rights: For routine work, define who decides and who must be consulted. For novel work, state that the default is that the person closest to the problem decides, with peer review, and that escalation is the exception.
- Install learning rituals: Require short incident reviews and weekly “question logs” that make it safe to surface uncertainties. Calibrate leaders to respond with inquiry first. This aligns with the learning mechanism associated with psychological safety (Edmondson, 1999).
- Broker trust across differences: Invest in conflict resolution training and structured interactions among employees with different values and backgrounds. The Edelman data identify employers as the most credible actors to scale these interventions (Edelman, 2026).
- Change what managers are measured on: Shift from counting attendance to assessing the conditions managers create. Use engagement and learning behaviors as leading indicators. Best-practice organizations show far higher manager engagement, which signals that design choices, not just macro conditions, drive the gap (Gallup, 2025).
None of this denies the need for accountability. Defaults are not absolutes. A default to trust gives teams speed and candor until evidence says a check is needed. A design for disagreement puts arguments where they belong, inside roles and processes rather than inside personalities.
The wider context is bleak. Most people believe the next generation will be worse off, and many are retreating into insular camps that resist difference (Edelman, 2026). Inside firms, engagement is sliding and managers are burning out (Gallup, 2025). Waiting for trust to emerge is not a strategy. Treat it as infrastructure, wire it into how decisions are made, and teach people how to disagree productively. The reward is not only cultural cohesion. It is the operational capacity to learn faster than the problem changes.
Sources
- Edelman Trust Institute (2026). Edelman Trust Barometer 2026. Finds 70% with an insular mindset and identifies employers as the best-positioned trust brokers who can scale conflict-resolution training and cross-values interactions. https://www.edelman.com/trust/2026/trust-barometer
- Gallup (2025). State of the Global Workplace 2025. Reports global engagement at 20%, a productivity cost near $10 trillion or 9% of global GDP, and a sharp decline in manager engagement, highlighting a design gap. https://www.gallup.com/workplace/659610/state-global-workplace.aspx
- Edmondson (1999). Psychological Safety and Learning Behavior in Work Teams. Shows team psychological safety is associated with learning behaviors that mediate performance in a field study of 51 teams. https://journals.sagepub.com/doi/10.2307/2666999
- Lee and Edmondson (2017). Self-managing organizations: Exploring the limits of less-hierarchical organizing. Defines SMOs and synthesizes evidence that hierarchy suits stable tasks but impedes complex cross-boundary problem solving, with limits noted. https://www.sciencedirect.com/science/article/abs/pii/S0191308517300059