What Makes People Speak Up at Work - and Why Most Leaders Get This Wrong

The silence problem

Most organizations have a silence problem. People in meetings know the plan has a flaw, but they do not say so. Nurses notice a physician is making an error, but they do not intervene. A team lead sees that the project is heading in the wrong direction three weeks before the deadline, but raises it only once, quietly, to someone who is not in a position to act on it. Morrison and Milliken (2000) called this organizational silence: the withholding of potentially important information, concerns, and opinions out of a rational calculation that speaking up is not safe.

The word "rational" is important here. Silence is not a personality defect or a character failure. It is a sensible response to an environment where speaking up has been observed, directly or indirectly, to carry risk. Someone tried to raise a concern and was dismissed. Someone surfaced a problem and became associated with that problem. Someone disagreed with a senior person in a meeting and was frozen out of subsequent conversations. These events do not have to happen to any individual personally. Watching them happen to others is sufficient to teach the lesson.

Edmondson (1999) named the variable that determines whether people speak up: psychological safety. It is defined precisely - not as comfort, or friendliness, or the absence of conflict, but as the shared belief among team members that the team is safe for interpersonal risk-taking. This distinction matters. Psychologically safe environments are not necessarily pleasant or consensus-driven. They are environments where honest disagreement, difficult questions, and the admission of uncertainty are treated as valuable rather than threatening.

What the research actually shows

Edmondson's original research was conducted in hospital settings and found that teams with higher psychological safety reported more errors - which initially seemed like a bad thing, until further analysis showed that the teams reporting more errors were actually making fewer of them. They were catching and surfacing mistakes rather than hiding them. This finding established a pattern that has since been replicated across industries: psychological safety predicts not just whether errors get caught, but whether teams learn, innovate, adapt, and perform at the level they are capable of.

Frazier, Fainshmidt, Klinger, Pezeshkan, and Vracheva (2017) meta-analyzed 136 studies and found that psychological safety consistently predicts knowledge-sharing, creative performance, learning behavior, and engagement. Google's Project Aristotle, which studied what distinguished its highest-performing teams from the rest, identified psychological safety as the single most important factor - more important than individual talent, technical skill, or team composition.

The mechanism is not complicated. Teams that feel safe share information more freely, which means they catch problems earlier. They challenge assumptions, which means they make fewer bad decisions. They admit uncertainty, which means they seek help sooner. They surface what is actually happening rather than what people want to hear, which means leadership gets the information it needs to make good decisions. The absence of psychological safety does not just suppress voice. It degrades the entire information environment that organizations depend on to function well.

Why leaders underestimate how unsafe their environments feel

One of the most consistent findings in this literature is the gap between how safe leaders believe their environments are and how safe team members actually experience them. Leaders almost universally rate their environments as more psychologically safe than their team members do. This gap is not random. It reflects a structural feature of organizational life: the people with the most power to affect psychological safety are the same people least likely to experience its absence.

Leaders can speak in meetings without calibrating the risk first, because there is no risk for them. They can disagree with someone senior to them - their peers - without the same career consequences a junior person faces disagreeing upward. They can ask naive questions without worrying that the question will define how they are seen. The result is that leaders genuinely experience their environments as safe, while the people around them are making constant calculations about what is and is not worth risking.

Detert and Edmondson (2011) found that most employees have a highly developed implicit theory about when speaking up is safe, who it is safe with, and about what topics. This calculation happens quickly and largely below conscious awareness. A leader who talks about wanting candor but visibly tightens when challenged is sending clear signals, even if they never explicitly punish anyone for disagreeing. People read these signals with precision.

The specific behaviors that create and destroy psychological safety

Psychological safety is created and destroyed through the accumulation of specific behavioral evidence. Owens and Hekman (2012) found that leaders who model intellectual humility - who acknowledge what they do not know, invite challenge, and visibly incorporate others' input into their thinking - create significantly safer environments than those who project certainty and authority. The act of a senior person saying "I am not sure" or "help me understand why you see it differently" sends a powerful signal that uncertainty and disagreement are welcome.

Edmondson (2018) identified several specific leader behaviors as the most reliable predictors of team psychological safety. Framing work as a learning problem rather than an execution problem, acknowledging one's own fallibility, and responding to concerns with curiosity rather than defensiveness all significantly increase the likelihood that people will speak up. Conversely, dismissing concerns as unimportant, attributing problems to individual failures rather than system factors, and visibly rewarding conformity over candor reliably decrease it.

Hannah, Avolio, and Walumbwa (2011) found that moral courage - the willingness to act on values even at personal cost - is among the most powerful leader behaviors for building trust and safety. This is the hardest one. It requires leaders to occasionally say things in front of powerful people that those people do not want to hear, to acknowledge mistakes publicly rather than quietly, and to protect the people who raise difficult issues rather than distancing themselves from those issues. None of this is comfortable. All of it is what psychologically safe environments are actually built from.

From theory to practice

The research is consistent enough that organizations can make specific, targeted investments in building psychological safety rather than hoping it emerges from culture initiatives. This starts with measurement: assessing safety at the team level, not just the organizational level, because safety is highly local and varies dramatically across teams in the same organization. It continues with leader-specific feedback that surfaces the gap between how leaders think they are showing up and how they are actually experienced. And it requires accountability structures that make psychologically safe behavior an explicit expectation of leadership, not an optional quality.

The organizations that consistently outperform on innovation, execution, and talent retention are not the ones where everyone gets along. They are the ones where the work of raising hard questions, admitting uncertainty, and challenging assumptions is treated as leadership - regardless of who is doing it, and regardless of what level they are at.