Psychological Safety Is Not Comfort: What Leaders Get Wrong
Psychological safety is the shared belief that a team is safe for interpersonal risk-taking — that members can ask questions, admit mistakes, and challenge decisions without fear of humiliation or reprisal. It is frequently mistaken for comfort, niceness, or lowered standards. The evidence points the other way: psychological safety improves performance only when it is paired with demanding accountability, not when it replaces it.
The concept has an unusually precise origin. In a study published in Administrative Science Quarterly in 1999, the Harvard organisational behaviourist Amy Edmondson set out to test whether better hospital teams made fewer medication errors. The data showed the opposite: the higher-performing nursing units reported more errors, not fewer. The finding was initially baffling until Edmondson realised she was not measuring error rates at all, but reporting rates. Strong teams were not more careless; they were more willing to surface mistakes, because doing so carried no social penalty. Weaker teams buried their errors to protect themselves. That single reversal reframed a generation of research. It established that what distinguishes effective teams is often not the absence of problems but the presence of a climate in which problems can be named. For leaders, the implication is uncomfortable: a quiet, agreeable meeting in which nothing is questioned is not evidence of a healthy team. It may be evidence of a frightened one.
The evidence base has since become substantial. A meta-analytic review by Frazier and colleagues, published in Personnel Psychology in 2017, aggregated 136 independent samples covering more than 22,000 individuals and nearly 5,000 groups, and found consistent positive associations between psychological safety and voice, learning behaviour, information sharing, creativity and task performance. A parallel systematic review by Newman, Donohue and Eva in the Human Resource Management Review reached similar conclusions and identified the principal antecedents: leader behaviour, supportive relationships and organisational context. Google's own internal study of team effectiveness, Project Aristotle, arrived at the same place by a different route. Having examined roughly 180 of its teams expecting to find that the right mix of talent or personality mattered most, Google concluded that psychological safety was the single most important of the five dynamics separating high-performing teams from the rest. The convergence of academic meta-analysis, systematic review and large-scale corporate data is rare in organisational science, and it is the reason evidence-based leadership development now treats psychological safety as a foundational variable rather than a soft preference.
Yet the popular reading of that evidence has drifted into error. Psychological safety is routinely translated, in corporate practice, into an injunction to be kind, to avoid conflict, and to shield people from discomfort. This is a category mistake. Edmondson herself has been explicit that safety is not the same as comfort, and that the two can pull in opposite directions. Her framework maps teams along two axes — psychological safety and accountability to demanding standards — producing four quadrants. High safety with low standards yields a comfort zone, where people feel free to speak but are not pushed to deliver, and complacency sets in. High standards with low safety produces an anxiety zone, where people are held to account but are too frightened to raise concerns, so preventable failures accumulate in silence. Low safety and low standards give apathy. Only the combination of high safety and high standards produces what Edmondson calls the learning zone, where teams can take interpersonal risks precisely because they are also expected to perform. The boardroom that mistakes safety for comfort has not made itself high-performing; it has made itself pleasant, which is a different and lesser thing.
The distinction is easiest to see in the anxiety zone, where the costs of getting it wrong are most visible. Consider a leadership team under intense pressure to hit a quarterly target, led by a chief executive known to react badly to bad news. The standards are exacting and everyone understands them; what is missing is the safety to report a slipping forecast, a faltering integration or a compliance concern while there is still time to act. The information does not disappear. It simply travels slowly, arriving as a fait accompli rather than an early warning, by which point the options have narrowed and the cost has multiplied. The same pattern recurs wherever authority is steep and candour is unrewarded — a political adviser who declines to tell a principal that a policy is unravelling, a founder whose senior team privately doubts a product decision but publicly defers. In each case the organisation is not short of intelligence; it is short of routes for that intelligence to reach the decision-maker in time. This is why the quiet meeting should worry a leader rather than reassure them: silence is not the sound of alignment but, frequently, the sound of a team managing its own exposure.
What the evidence does and does not support
The maturing literature has also begun to map the limits of the idea, and leaders should attend to this as carefully as they attend to its benefits. A study by Eldor, Hodor and Cappelli, published in Organizational Behavior and Human Decision Processes in 2023, examined the relationship across five independent settings and found that it is not linear. For routine, well-defined tasks, performance rose with psychological safety only up to a point and then declined: beyond a threshold, a very high-safety climate appeared to reduce the diligence with which repetitive work was executed. Crucially, the researchers found that this decline was offset by collective accountability. In other words, safety without accountability can erode standards on predictable work, while safety with accountability sustains them. The Chartered Institute of Personnel and Development reached a compatible conclusion in its 2024 evidence review, cautioning that psychological safety is powerful but frequently oversold as a universal remedy, and that its effects depend heavily on context, task type and the standards a team is held to. The honest reading of the evidence is therefore not that more safety is always better, but that safety is a necessary condition for candour and learning which becomes damaging when it is decoupled from expectation.
Understanding why people stay silent explains why this matters so acutely in high-stakes environments. The decision to speak up in a meeting is, at root, a rapid social risk calculation: the potential benefit of raising a concern is diffuse and future, while the potential cost — looking ignorant, negative or disloyal in front of powerful people — is immediate and personal. Human beings are systematically biased towards avoiding that immediate cost, which is why silence is the default in most hierarchies. The dynamic intensifies as the stakes rise and the audience grows more senior. A junior analyst who spots a flaw in a fundraising model during a board presentation, a civil servant who doubts the premise of a minister's favoured policy, an executive who suspects an acquisition is being driven by ego rather than logic — each faces the same asymmetry, and each will tend to resolve it in favour of self-protection unless the leader has done deliberate work to change the calculus. This is the practical heart of leadership under pressure. The most consequential information in an organisation is often held by the people least incentivised to volunteer it, and it is the leader's behaviour, far more than any stated value, that determines whether it surfaces in time to be useful.
What that behaviour looks like is well specified by the research, and it is more demanding than the cheerful version implies. Because leader conduct is among the strongest antecedents of psychological safety, the burden falls disproportionately on the person with the most authority in the room. It begins with framing the work honestly as complex and uncertain, so that questions and errors are understood as inevitable features of difficult problems rather than as personal failings. It requires the leader to model fallibility — to acknowledge the limits of their own knowledge and to state plainly that they need to be told when they are wrong, which lowers the perceived cost of dissent for everyone below them. It requires actively soliciting the contrary view rather than waiting for it, because the absence of objection in a hierarchy is rarely the same as agreement. And it depends, above all, on how the leader responds the first time someone delivers unwelcome news: a single punitive reaction can extinguish candour that took months to build, while a productive one compounds it. None of this lowers the bar. A leader can hold a team to exacting standards and still make it safe to fail intelligently; indeed, the evidence suggests that only the combination produces sustained high-performing board dynamics. Emotional discipline on the part of the leader — the capacity to receive challenge without defensiveness — is the mechanism that makes the two compatible.
The enduring misunderstanding, then, is to treat psychological safety as an end in itself, a matter of making people feel good, when the evidence frames it as an instrument for making people perform. It is the condition under which a team's real intelligence — its doubts, its dissent, its early warnings — becomes available to the person who has to decide. Stripped of accountability it decays into comfort; welded to it, it becomes the substrate of candour, learning and resilience under pressure. Leaders operating in environments where a single unspoken concern can prove catastrophic may wish to treat the climate of candour in their teams as seriously as they treat strategy or capital allocation, and to examine honestly whether the quiet in their own meetings reflects agreement or fear.
Frequently Asked Questions
What is psychological safety in a leadership context?
Psychological safety is the shared belief that a team is safe for interpersonal risk-taking. In leadership terms, it means people can ask questions, admit mistakes, disagree with senior colleagues and raise concerns without fearing humiliation or punishment. It enables candour, learning and early warning of problems that would otherwise stay hidden.
Is psychological safety the same as being nice or comfortable?
No. This is the most common misconception. Psychological safety is not comfort, niceness or lowered standards. Research by Amy Edmondson shows it must be paired with high accountability to produce performance. Safety without demanding standards creates complacency; safety combined with high standards creates a genuine learning environment.
Does psychological safety actually improve team performance?
The evidence is strong but conditional. A 2017 meta-analysis of over 22,000 individuals links psychological safety to learning, voice and task performance, and Google's Project Aristotle found it the top predictor of team effectiveness. However, 2023 research shows the effect is non-linear and depends on being paired with accountability.
Why do people stay silent in high-stakes meetings?
Speaking up is a social risk calculation. The cost of appearing ignorant, negative or disloyal in front of powerful people is immediate and personal, while the benefit of raising a concern is diffuse and future. People default to self-protection unless a leader deliberately lowers that cost through their own behaviour.
