Nobody Felt Safe Saying That Psychological Safety Wasn't Working

Somewhere between Amy Edmondson's 1999 hospital study and your company's last townhall meeting, psychological safety completed a remarkable journey. 

It started as a rigorous empirical finding about why some surgical teams reported more medication errors than others - not because they were more reckless, but because their leaders made it safe to name mistakes out loud. It ended up as EVP gambit, a bullet point on a careers page between "we're a family" and "bring your whole self to work," and - in its most evolved corporate form - the single most effective reason why no one ever told the Chief People Officer that the engagement survey results were being read upside down. 

SPONSORED by Fathom

Your employees know the truth. Does your EVP? At Fathom we measure the "Credibility Gap" between your promise and their reality.

Close the Gap

This is the story of how an idea built to make organisations more sensitive and inclusive turned into the architecture of institutional dishonesty. And what it has quietly done to employer branding in the process. 

What Edmondson Actually Found (Which Was Not What You Think) 

In 1999, Amy Edmondson - then at Harvard - was studying medical teams and expecting to find that the best-performing ones made fewer errors. What she found instead was the opposite. The teams with the strongest leaders and the healthiest dynamics reported more errors, not fewer. Not because they were worse at their jobs. Because they felt safe enough to say so. 

The insight was elegant and counterintuitive: psychological safety, as she defined it, was "a shared belief held by members of a team that the team is safe for interpersonal risk-taking." It was about whether people believed they could speak up - admit a mistake, challenge an assumption, ask a stupid question - without being humiliated, sidelined or quietly managed out. 

Notice what it was not. It was not a promise that work would be comfortable. It was not a guarantee that your ideas would be celebrated, your performance excused, or your feelings centred above operational reality. It was specifically about the interpersonal conditions required for honest, risky communication to happen. A team could be under enormous pressure, facing hard deadlines and genuine stakes, and still be psychologically safe - because safety, as Edmondson meant it, was the precondition for doing difficult things well together, not a holiday from difficulty itself. 

This distinction matters enormously. Because the version of psychological safety that most organisations adopted over the next twenty-five years bore approximately the same relationship to Edmondson's original research as a children's colouring book bears to the Sistine Chapel. 

The Google Effect, or: How a Good Idea Gets Ruined 

For the first decade after Edmondson's paper, psychological safety lived mainly in academic circles. Then, in 2013, Google launched Project Aristotle - an internal study of 180 teams designed to identify what made them perform. The finding that dominated every subsequent summary, keynote and LinkedIn post was this: psychological safety was the single strongest predictor of team effectiveness, outweighing individual talent, resource access, and team composition. 

A New York Times piece in 2016 brought this to the mainstream. Edmondson's 2018 book The Fearless Organization cemented it as a global management imperative. By 2019, "psychological safety" was appearing in job descriptions. By 2020, it was a pandemic-era HR non-negotiable. By 2022, it had completed its transformation from research construct to corporate religion. 

Here is the problem with corporate religions: the devout stop asking whether the doctrine makes sense and start competing to demonstrate their faith most visibly. 

So, HR teams built psychological safety frameworks. Consultants ran psychological safety workshops. Leaders were assessed on their psychological safety scores. Careers pages promised "fearless cultures." LinkedIn began filling with posts about how "I made a mistake last week and my manager said thank you." Which is, of course, wonderful - until you notice that the same organisations posting those stories were also issuing performance improvement plans, freezing salaries, and announcing restructures with two hours' notice. 

The concept had, without anyone quite intending it, completed a mutation. It had shifted from "the conditions under which people can take interpersonal risks" to "the feeling that nothing at work will ever be uncomfortable." These are not the same thing. They are, in several important ways, opposites. 

The Cowardice Mutation: Three Ways It Went Wrong 

The misuse of psychological safety is not random. It takes three fairly consistent forms, each more ironic than the last. 

The first is leadership abdication dressed as care. Psychological safety became, for a certain kind of manager, a useful reason to avoid giving anyone difficult feedback. Saying "I want to create a safe space for you" sounds considerably better than "I find performance conversations uncomfortable and have never been trained to have them." Both sentences, in many organisations, produce the same outcome: the struggling employee hears nothing useful until the PIP lands on their desk, at which point they discover that the safe space had a trapdoor. 

SPONSORED

Helping HR, talent acquisition, employer branding, and company culture professionals find careers worth smiling about.

Find People-First Jobs Now

True psychological safety, as Edmondson designed it, requires a leader to frame work as a learning problem - which means openly naming what is not working, inviting honest challenge, and responding productively when someone brings bad news. It requires more candour, not less. The corporate version somehow produced the opposite: a norm where difficult truths were considered unkind, challenge was reframed as aggression, and the most reliably safe career move was to agree with whoever had the largest budget. 

The second is the policy delusion. In 2024, in the USA, the Rhode Island state senate passed a Workplace Psychological Safety Act. Edmondson's response was withering and worth quoting directly: "Telling people in a company or on a team that they must have psychological safety 'or else' will not produce it. Psychological safety, rather than being created by a policy, is built in a group, interaction by interaction." 

Read that again. The person who invented the concept had to publish a clarification explaining that you cannot mandate it into existence. That organisations had drifted so far from the underlying mechanism - the patient, repeated, relational work of building trust through behaviour - that they had started trying to legislate the outcome instead. This is roughly equivalent to passing a law requiring people to fall in love. 

The third is the most structurally damaging: psychological safety became, in many organisations, a shield for mediocrity. When challenge is culturally coded as threat, when every piece of difficult feedback requires a fifteen-step framework and a HR debrief, when managers have been trained to treat any expression of discomfort as a safety incident - standards quietly collapse. Not dramatically. Not visibly. Just incrementally, in every meeting where no one says the obvious thing, every project review where the problems are parked, every culture survey where the scores look fine because everyone has learned that fine is what you say. 

Wharton's Peter Cappelli and colleagues documented this empirically. Their research found that the relationship between psychological safety and performance is not linear - it is an inverted curve. A moderate level genuinely helps. People who fear ridicule or humiliation perform worse, even in routine jobs. But above a certain threshold, too much psychological safety becomes actively harmful. Accuracy declines. Productivity drops. The hypothesis is straightforward: if the consequences of poor performance are genuinely zero, the incentive to avoid poor performance becomes genuinely minimal. 

Nobody put this in their careers page. 

The Middle Manager Paradox 

Here is perhaps the sharpest irony in the entire story, courtesy of Harvard Business Review research published in late 2025. 

Middle managers - the people most directly responsible for building psychological safety within their teams, the people who attend the workshops and complete the modules and deliver the values - report the lowest levels of psychological safety of any group in organisations. They are squeezed between senior leaders who expect compliance and delivery, and frontline employees they are supposed to protect and develop. They are the transmission belt of culture, and they are the least safe people in the building. 

Free Employer Branding Brief Templates

Two research-grade briefing templates: one for EVP development, one for creative campaigns. Built by EBN in partnership with Fathom. Free to fill in, save, and download.

Employer Branding NewsEBN

So, the concept designed to help people speak truth upward is most absent precisely where the speaking-upward needs to happen most. The managers who most need to challenge the strategy, flag the risks, name the problems - they have the fewest conditions to do so. And yet the employer brand continues to promise a fearless organisation where making mistakes is encouraged. 

The mechanism broke somewhere in the middle. Like it always does. 

Boeing, and What Happens When the Gap Becomes a Canyon 

If you want to understand what it looks like when an organisation perfects the performance of psychological safety while eliminating the practice of it, Boeing is your case study. 

Boeing's public culture narrative for years emphasised safety, voice, and the importance of employees raising concerns. The reality, as investigations following a series of catastrophic incidents revealed, was that employees had raised concerns - about manufacturing processes, about quality controls, about the gap between what was being promised externally and what was being built internally. And they had, with some frequency, concluded that raising those concerns was not, in any meaningful sense, safe. 

The CEO departed in 2024. The legal and reputational consequences are still unfolding. The cost is measured in lives and billions of dollars. 

This is not just an employer branding cautionary tale. It is a catastrophic failure of governance and human safety that should not be reduced to a branding lesson. But it is also the starkest possible illustration of what happens when the language of psychological safety is adopted in a culture deck while the actual conditions for honest challenge are systematically suppressed. The words and the reality diverge so completely that employees learn not to trust either. 

Most organisations are not Boeing. Thankfully. Most gaps between the safety promise and the safety practice are not as catastrophic in consequence. But the gap is almost always there. And in an era when employer reviews are public, when employees talk, and when the candidates EB professionals are trying to attract are more sceptical of brand promises than any generation before them - the gap is increasingly visible. 

The Employer Branding Consequence 

"We are a fearless organisation." "Speak up and be heard." "Psychological safety is core to who we are." 

These phrases, and dozens of variations on them, have spent the last few years migrating from Amy Edmondson's research summaries into careers pages, culture videos, EVP workshops and glassdoor responses. They have become, in the employer branding world, a kind of cultural virtue signal - the HR equivalent of claiming you care about sustainability while turning a blind eye to mass deforestation.  

The problem is not that the aspiration is wrong. The problem is that it became an attraction promise before it became an operational reality. And in the era of the Big Stay - when employees are not moving, when they are watching closely, when they are comparing the promise to the lived experience every single day - that gap is exactly the kind of thing that destroys retention. 

Candidates who joined because you promised psychological safety and then experienced a culture where disagreeing with the strategy was career-limiting, where feedback was weaponised in PIPs, where the most reliable path to promotion was visible agreement - those candidates do not stay quietly. They leave reviews. They talk to their networks. They tell the next cohort what the organisation is actually like. 

The EBN Dispatch Podcast | Employer Branding & Talent

The EBN Dispatch is your no-fluff podcast on employer branding, talent attraction, retention, and company culture served with dry wit and sharper-than-average insight.

Employer Branding NewsEBN

The employer brands that built their EVP on psychological safety without building the underlying conditions are now in an uncomfortable position: the promise is on the wall, the reality is on Glassdoor, and the gap between them is precisely what drives the trust deficit that makes every hiring campaign harder. 

What the Real Version Looks Like 

None of this means the concept is wrong. The original insight - that teams perform better when people believe they can take interpersonal risks - is well-supported, genuinely important, and more relevant than ever in an environment where AI is reshaping roles, uncertainty is high, and the cost of institutional silence (see above) has never been more visible. 

What it means is that the real version of psychological safety looks nothing like the corporate version. And employer brands that want to credibly claim it need to demonstrate it in specifics, not assert it in generalities. 

Edmondson herself offered the sharpest possible corrective in 2025: stop talking about psychological safety and start talking about the goal. When a leader says "we need everyone's input to get this right because we have genuinely never done this before," they are building psychological safety through action. When they say "we're a psychologically safe team," they’re probably missing the point entirely. 

I know I keep harking on about this, but... the employers that will genuinely rebuild trust with talent are the ones that can show what honest, uncomfortable conversations look like in practice. That means openly sharing details of decisions that went wrong and what was learned. Owning it beats being afraid of it or covering it up.  It means that leaders who are visibly challenged in public forums should respond with curiosity and human fallibility rather than defensiveness. It means promotion decisions that reward people who raised the inconvenient truth, not just the people who delivered the comfortable result. 

It does not mean a laminated poster. It does not mean a workshop. And it definitely does not mean a Rhode Island Act of Parliament. 

So, What Now?

There is something almost perfectly circular about the trajectory of psychological safety as a corporate concept. 

It was designed to make it safe to say difficult things. Its widespread adoption made it dangerous to say the one difficult thing it most needed someone to say: this isn't working. The concept became so embedded in the cultural lexicon that challenging its implementation - pointing out that the workshops weren't producing candour, that the frameworks weren't producing honesty, that the promise was outrunning the practice - felt, to many people in HR and employer branding, like the least psychologically safe thing you could do. 

So, nobody said it. The posters stayed on the walls and the social media posts kept rolling. The EVPs kept the language. And the gap between the promise and the reality kept growing in the background, one avoided conversation at a time. 

Amy Edmondson spent 1999 studying why some teams told the truth and others didn't. The answer, it turned out, was whether the people in charge made it genuinely safe to do so - through their behaviour, their responses, and their willingness to hear difficult things without punishing the person who said them. 

Twenty-six years later, the best thing employer brands could do with her research is exactly what most of them have spent twenty-six years failing to demonstrate: say the difficult thing, in public, clearly, and mean it. 

The Trust Recession is here. Edelman shows employer trust falling. This guide gives you the playbook to rebuild it.

Get the white paper

Sources

Takeaways