Psychological Safety Theater: Training Without Structural Change

Psychological safety has become the most frequently trained and least structurally addressed topic in large organisations. The training is commissioned, the workshops are delivered, the engagement scores improve, the reports look encouraging. The underlying condition — what actually happens to people who tell the truth when it is inconvenient — is frequently unchanged.

The gap is not accidental. It is produced by a specific misunderstanding of what psychological safety is and where it comes from.

Safety as structural property, not climate attribute

The standard corporate framing treats psychological safety as a climate variable — something you can build through leadership signalling, values communication, and workshops on vulnerability and inclusive behaviour. Climate-variable thinking generates climate-variable interventions: survey items are added, leaders are trained to model the behaviours, the culture deck is updated to name the value, and the metrics move.

This treats safety as a quality that can be installed through signal. It cannot. Psychological safety is a structural property of the environment, not a signalled one. It exists or does not exist based on what actually happens when someone speaks up.

Employees, in my experience, are not credulous about signals. They read consequences with considerable accuracy. The leader who says “we value challenge” in a workshop, and who then quietly sidelines the direct report who challenged them last quarter, has taught the organisation something specific and durable about what actually happens when challenge appears. The workshop is signal. The sidelining is data. Employees attend to the data.

This is why the most common pattern I see in organisations that have invested heavily in psychological safety training is precisely the pattern the investment was meant to address: rising survey scores on safety items, falling retention at senior levels, and a widening gap between what people say in forms and what they say when no one is measuring.

The mechanism of theater

Three specific dynamics compound to produce the theater version of safety work.

The first is asymmetric investment. Training is cheap, visible, and auditable. Structural change is expensive, politically costly, and difficult to demonstrate in quarterly reporting. The organisation’s investment flows toward the visible, the cheap, and the auditable. The structural work that would actually move safety remains unfunded because it is not legible to the reporting apparatus that controls budget.

The second is measurement capture. What gets measured is self-reported psychological safety in surveys — which is responsive to signalling, culture communication, and leadership performance. What is not measured is the actual treatment of people who bring difficult information into the room — because there is no obvious way to track it, and because tracking it would require the organisation to look at specific cases where it behaved badly. The measurement apparatus inadvertently selects for interventions that move survey scores without requiring behavioural change.

The third is leadership incentive misalignment. The leaders responsible for commissioning safety work are also, usually, the leaders whose behaviour has produced the current state. Asking them to fund and sponsor an honest structural audit is asking them to fund an investigation that will, in some cases, end with specific findings about themselves. The political economy of the intervention means that the version commissioned is reliably the version that does not produce uncomfortable findings — which is, usually, the training version.

A specific pattern, observed in three organisations

I have watched this pattern operate at close range in three separate client organisations over the last four years. The details vary. The structure is constant.

Each organisation identified a psychological safety gap through its engagement survey. Each commissioned substantial training — in one case, over eighteen months of rolling workshops across senior levels. Each saw improvement in the survey scores. Each continued to lose, at approximately the same rate as before, the specific population the safety work was meant to support: the reflective, direct, structurally honest leaders whose presence the organisations said they valued but whose contributions they had no structural capacity to metabolise.

In each case, the organisation concluded that the safety work was succeeding — because the scores were moving — and that the attrition was coincidental. In each case, the attrition was the data that told you what the scores could not.

What genuine safety work involves

The intervention is not training. It is structural audit followed by structural change.

The audit asks a specific question, honestly: What actually happens in this organisation when someone is right and inconvenient? Not what the policy says. Not what leadership claims. What happens, concretely, to the specific person who brings difficult information into the room. Are they heard, protected, included in the next conversation? Or are they sidelined, rebranded as difficult, managed toward the exit?

The audit is conducted by identifying specific cases from the last eighteen months in which someone in the organisation raised something inconvenient. Not hypothetical cases. Named cases, with named people, and documented outcomes. The pattern across the cases is the diagnostic. If the people who raised inconvenient things have, on average, left the organisation at materially higher rates than their peers, the safety narrative is theater regardless of what the training schedule shows.

The structural change, when the audit surfaces that the pattern is indeed negative, involves work at several levels. Performance management criteria that no longer reward compliant alignment over productive dissent. Review processes in which the behaviour of leaders toward people who challenged them is part of the evaluation. Exit interviews that are treated as data rather than as closure. And, most difficult, explicit organisational commitment to protect specific individuals whose contributions the organisation has been structurally ejecting.

None of this is in the training brief. All of it is what actually produces safety. Which is why the organisations that have it are rare, and the organisations that have commissioned safety training are common.

The honest question

Before commissioning the next round of psychological safety work, ask one question honestly.

In the last eighteen months, can you name three people in this organisation who brought difficult information into a room and were materially worse off for having done it?

Most senior HR leaders, asked this honestly, can name them. The honest next question is what happened to those people, and whether the pattern across them is one the organisation would be comfortable defending publicly.

If the pattern is defensible, safety work may be appropriate. If it is not, training will not address what is actually happening. And the organisation that commissions training over structural work has, in effect, chosen theater over the thing the theater is meant to represent.

That is a choice. It is worth making consciously.