Why the most popular idea in boardroom improvement is applied backwards.
Psychological safety has become one of those concepts that nobody disagrees with and nobody quite knows what to do with.
The standard advice is straightforward: create psychological safety, and people will speak up. Lower the temperature. Encourage openness. Make it safe to be vulnerable. Then, supposedly, better decisions will follow.
The intent is good. In governance settings, the sequence is backwards.
Safety is not what you create so that people will challenge. Safety is what you get after the system has proven it can tolerate challenge. It is an output of clarity and trust, not a precondition for them.
That is a meaningful difference. It changes where you intervene, what you invest in, and why most psychological safety initiatives in boardrooms quietly fail.
What the concept actually means
The irony is that the original research supports this argument more than it supports the way the concept is commonly applied.
Amy Edmondson defined psychological safety as a shared belief that the team is safe for interpersonal risk-taking. The emphasis is on risk-taking, not on comfort. Her measurement scale asks whether mistakes are held against people, whether members can bring up problems and tough issues, whether it is safe to take a risk. These are questions about whether the system can tolerate truth under pressure. They are not questions about whether people feel warm.
Somewhere between the research and the current interpretation, psychological safety got conflated with the notion of a safe space. A refuge from feedback, from directness, from anything that might feel harsh. Those are entirely different things. A group with genuine psychological safety is one where people take interpersonal risks precisely because the system has shown it can hold them. A safe space, in the way the term is commonly used, is one where those risks are avoided altogether.
Most boards that say they are working on psychological safety are actually building safe spaces. They are optimising for comfort when the concept was always about tolerance for risk. The original idea was never the problem. The translation was.
The proof that most directors already recognise
Ask experienced directors when boardrooms feel safest for truth, and many will give you a surprising answer: after a crisis.
After a major acquisition, a performance collapse, a reputational event, a shock like a pandemic. These are stressful, high-stakes, uncomfortable moments. They force hard decisions, expose capability gaps, and increase pressure on everyone in the room.
And yet the boardroom often becomes safer for truth after those moments. Not before. After.
The reason is that crises do two things simultaneously. They force clarity, by collapsing ambiguity, making the board prioritise, and surfacing trade-offs that had been deferred. And they test trust, because people see in real time who can tolerate truth, whether challenge is punished or received, whether the chair can hold tension without letting it become personal, whether decisions stick under pressure.
If the system holds through that, trust increases quickly. And when clarity and trust rise together, safety rises as a natural consequence.
The crisis did not create safety by being gentle. It created safety by proving the system could survive truth. Safety followed. It did not lead.
Why clarity and trust are the actual levers
Once you see the crisis pattern, the mechanism becomes clear.
When a board has genuine shared understanding of what it is trying to achieve, challenge becomes contribution rather than politics. A director who questions the strategy is obviously testing it against an agreed direction. The question is interpretable. It is legitimate. It does not need to be softened or caveated because its purpose is self-evident.
When that clarity is missing, the same question feels political, presumptuous, or hostile. Not because the question has changed, but because there is no shared frame to make sense of it. Directors learn this quickly. They soften their language, hedge their concerns, or move the real conversation into the corridor. The room becomes polite and the governance becomes weak.
Trust works the same way. When challenge has been handled well in the past, when hard questions have been received in good faith rather than punished, the board learns that truth is survivable. Each challenge that lands well makes the next one easier. But that trust is earned through experience, not declared through norms. You cannot tell a board to trust that challenge is safe. The board has to live it.
Clarity makes challenge legitimate. Trust makes challenge survivable. When both are present, safety is what the room feels like. When either is absent, no amount of encouragement will make directors speak up on the things that actually matter.
What a lack of safety actually signals
A lack of psychological safety is a genuine red flag. But it is usually pointing to something different from what people assume.
The usual interpretation is that the room is too hard or too confrontational. That can be true. But just as often, the lack of safety is pointing to something structural: unclear direction and weak trust. When purpose is ambiguous, questions feel political. When roles are unclear, challenge feels like interference. When trust is low, truth feels costly. The problem is not that people lack courage. It is that the conditions make courage irrational.
The correct question is not how do we make people feel safer? It is what is this lack of safety pointing to in the system? That reframe turns psychological safety from a vague aspiration into a diagnostic instrument, which is much closer to what Edmondson's original research intended.
What to do instead
Boards do not become effective by pursuing comfort. They become effective by building a system that can tolerate truth.
Start with clarity. Is the board genuinely aligned on what it is trying to achieve? Can directors articulate the direction consistently? Are trade-offs named or deferred? When clarity is strong, challenge has a target and a purpose. It does not need to be coaxed into existence.
Then look at how challenge is handled. When someone asks a hard question, does the room engage or retreat? When someone disagrees, does the chair protect the work or let it become personal? When a decision is made, does it hold? Those behaviours, repeated over time, are what build the trust that makes safety real.
If you want a board where people take genuine interpersonal risks, where hard questions get asked early and received well, do not chase safety directly. Build the conditions that make truth legitimate and survivable, and let safety emerge as the byproduct. That is the reframe. Once you hold it, psychological safety stops being a slogan and starts being useful.