Psychological safety
The best-evidenced factor in team performance: the shared sense that you may ask questions, disagree and admit mistakes without being punished for it. Without that foundation, every decentralized organization stays theoretical.
What it is — and what it isn't
The term comes from Harvard professor Amy Edmondson. She defines psychological safety as a team's shared belief that it is safe to take interpersonal risks: to ask a question, offer an idea, admit a mistake, or openly disagree with someone more senior.
❌ Comfort
- Not "being nice to each other at any cost".
- Not low standards, not a cosy culture.
- Not the absence of conflict.
✅ Candour under pressure
- People say what they think — especially when it is uncomfortable.
- High standards AND high safety, together.
- Disagreement about the work is allowed and wanted.
Edmondson stresses the point: only the combination of high standards and high safety produces learning and performance. Safety alone becomes a comfort zone; standards alone become an anxiety zone.
Why this is not an opinion but evidence
Edmondson's field studies
As early as 1999, Edmondson showed in hospital teams that units with high psychological safety reported more errors — not because they made more, but because they spoke about them openly and learned from them. Safety drove the learning behaviour.
Google, Project Aristotle
Google studied 180 of its own teams over several years to find out what makes a high-performing team. The result: not the composition, not the individual talent — by a clear margin the single most important factor was psychological safety.
Deming: "Drive out fear"
W. Edwards Deming had already made it point 8 of his famous 14: drive out fear. His core estimate: roughly 94 % of problems sit in the system, not in the people — and punishing individuals for system faults produces fear and destroys improvement.
Why Beta does not work without it
Decentralized organization means people doing the work decide for themselves, coordinate directly, and are pulled by the market. All of that presupposes that people speak up — otherwise the periphery stays silent and the hierarchy ends up deciding alone after all.
That is why it is one of the five dimensions in the self-check — measured there against two everyday signals: do people dare to disagree openly with someone senior? And: how do you handle mistakes?
How to recognize it day to day
Disagreeing upward
Can you disagree with the boss on the substance, without consequences?
Where rank forces the last word, the knowledge at the edge goes quiet.
How mistakes are handled
Does an error become shared learning — or does it get escalated and sanctioned?
Sanctioned mistakes do not become fewer, they only become hidden.
Questions and not knowing
May someone say "I don't understand this" or "I don't know"?
Where not knowing counts as weakness, people would rather guess than ask.
What you can work on
- Leaders go first: admit your own mistakes and gaps before anyone else does. Safety is created from the top, through vulnerability, not through announcements.
- Curiosity before judgement: respond to concerns with questions, not defence. Whoever disagrees should be thanked for it, not penalized.
- Depersonalize errors: when something goes wrong, ask the system first ("what made this possible?"), not the person ("whose fault was it?").
- Keep standards high: safety is not a retreat from ambition — the two belong together.
Sources
- Amy C. Edmondson: Psychological Safety and Learning Behavior in Work Teams. Administrative Science Quarterly, 1999.
- Amy C. Edmondson: The Fearless Organization. Wiley, 2019.
- Google re:Work / Project Aristotle: Guide: Understand team effectiveness. 2015.
- W. Edwards Deming: Out of the Crisis (the 14 points, point 8 "Drive out fear"). MIT Press, 1986.
More from the Beta Library: betaos.org/library