What is psychological safety, and does your board have it?

Two months into their first term, a new trustee is reading the treasurer’s report and cannot tell whether the deficit on page three is planned or a problem. The question is fully formed. It’s a good question — half the room probably has it too. They glance up, see eleven people nodding along, and make a small, silent calculation: what does asking cost here? Then they decide to look it up later.

They won’t look it up later. And next month the calculation will be a little faster, because now there are two meetings’ worth of things they should already know.

That silent calculation is what psychological safety is about. The term comes from Amy Edmondson at Harvard, and it means something specific: a shared belief, across a group, that this is a room where you can take an interpersonal risk — ask the basic question, admit the mistake, voice the doubt, be wrong out loud — without being embarrassed or quietly marked down for it. Not a personality trait, and not a policy. A property of the room, built or eroded one response at a time.

Whether your board has it is worth knowing, because the research behind the idea is one of the more startling stories in the study of teams — and it started with a finding that looked like a failure.

The nurses who reported more mistakes

In the 1990s, Edmondson was a doctoral researcher studying medication errors in hospitals. Her hypothesis was the obvious one: better nursing teams — better led, better coordinated, more effective by every measure — would make fewer drug errors. She gathered the data, ran the numbers, and got the reverse. The better teams were reporting more errors, not fewer.

For a while this looked like a career-ending result. Then she considered another explanation: maybe the strong teams weren’t making more mistakes. Maybe they were admitting them — catching errors, flagging them, talking about them — while the weaker teams were making just as many and burying them. Further investigation bore that out. The best teams had a climate where a nurse could say “I think I gave the wrong dose” and expect help rather than humiliation. The worst had a climate where the sensible move was silence.

The errors were never the variable. The willingness to say them out loud was.

Google asks the same question, twenty years later

Google spent two years and studied more than 180 of its own teams trying to figure out what separated the great ones from the ordinary ones — the effort was called Project Aristotle. The expectation, more or less, was a recipe: the right mix of stars, skills, and personalities. What they found instead was that who was on the team mattered far less than how the team treated each other — and of the five dynamics that distinguished the effective teams, psychological safety came first. Teams whose members felt safe to take risks in front of each other outperformed teams that, on paper, had better people.

A hospital ward and a software company are not a volunteer board. But notice what both findings have in common: groups of competent people, doing consequential work, where the difference-maker wasn’t competence. It was whether the room punished honesty.

What psychological safety is not

The term gets misread as a synonym for comfort, so it’s worth being precise about what it doesn’t mean.

It is not niceness. A board can be warm, friendly, and deeply unsafe — a room where everyone is pleasant precisely because nobody would dare be anything else. Politeness that prevents honesty isn’t safety; it’s the symptom of its absence.

It is not comfort. A psychologically safe board is often less comfortable, in the moment, than an unsafe one. Real questions get asked. Assumptions get examined. The safety isn’t freedom from challenge — it’s confidence that challenge won’t be punished.

It is not the absence of standards. Edmondson’s high-performing teams weren’t casual about errors; they were rigorous about them, which is exactly what the reporting made possible. Safety and standards aren’t a trade-off. Safety is what lets a group hold standards honestly instead of theatrically.

The tells

You can’t survey your way to this answer in one evening, but a board wears its climate openly if you know where to look.

Who hasn’t spoken? Not tonight — this quarter. Most boards have a member or two whose voice has gone missing, and the room has stopped noticing. Sustained silence from a capable adult is rarely a personality; it’s a verdict on the room.

What happened the last time someone was wrong out loud? Somebody misread the financials, made a prediction that missed, backed the wrong candidate. Replay how the room responded — the actual sentences, the actual faces. That moment taught everyone present what being wrong costs here, and they have all priced it in since.

Has the newest member asked a question yet? New trustees arrive with the most valuable perspective a board gets — the ability to see what everyone else has stopped seeing — and a strong incentive to keep it to themselves. A board where new members go quiet for a year isn’t onboarding them. It’s training them.

Where does the real conversation happen? If the candid assessments happen in the parking lot, on the drive home, in the two-person phone call after — the board is having its real meeting without a quorum. The distance between what’s said in the room and out of it is a fair measure of how safe the room is.

How chairs build it — small moves, consistently

The good news buried in Edmondson’s work is that this climate is made of small, ordinary responses — which means a chair can build it without a retreat or a consultant. Three moves matter most.

Respond to the first bad idea generously. Someone will eventually float a suggestion that doesn’t survive contact with the budget. That moment is worth more than the idea. “Walk me through it — what would that solve?” costs the chair nothing and tells the room that thinking out loud is allowed here. A wince, a joke at the idea’s expense, a too-quick “we’ve tried that” — each one teaches everyone watching to pre-screen their thoughts, and the board loses every idea that wasn’t a sure thing.

Admit your own errors first, and specifically. “I called that one wrong in March — I pushed for the earlier date and it cost us” is a sentence that takes a chair ten seconds and repays the board for months. Interpersonal risk works like any other kind: people watch what happens to the first ones in. When the person with the most standing spends some of it on candor, the price drops for everyone.

Thank the question, not just the answer. Boards reliably praise good answers and quietly resent good questions, because questions slow things down. Reverse it, out loud: “I’m glad you asked that — I’d guess three of us were wondering.” Said in the moment, in front of the room, it converts the new trustee’s silent calculation from what will asking cost me? to asking is what we do here.

The condition underneath the mechanics

None of this replaces the mechanics of good discussion — sending the packet early, bringing real options, having the chair speak last. Those techniques matter, and they’re covered elsewhere. But mechanics run on a condition. A round-robin in an unsafe room produces twelve rehearsed sentences; the same round-robin in a safe one produces the reservation that saves the project. Build the condition and the techniques work. Skip it and they’re furniture.

Somewhere at your table, a trustee is doing the silent calculation right now. Every response the board makes — to a question, a mistake, a half-formed idea — adjusts the price. Set it low enough, and you’ll finally hear what your board has been thinking all along.