Organisations talk about psychological safety constantly. It appears in leadership development programmes, in engagement surveys, in DEI frameworks, in the language of progressive HR teams who want to signal that they understand what an inclusive culture requires.
And for the 58% of Black professionals in this research who do not feel safe expressing their opinions at work, the gap between that language and their daily reality is not a small one (Williams, 2026).
This article is about that gap. What psychological safety actually is. Why the version being offered to Black professionals consistently falls short. What the research documents about what genuine safety would need to look like. And what organisations are currently calling safety that is something else entirely.
The Limits of the Original Psychological Safety Framework
The concept of psychological safety originates from Amy Edmondson’s (1999) work at Harvard Business School. She defined it as a “shared belief that the team is safe for interpersonal risk-taking.”
While influential, this framework has significant limits:
- Homogeneous Research: It was built on predominantly white organisational environments.
- Credibility Assumptions: It assumes all team members start with equivalent credibility.
- The Racial Lens: It doesn’t account for “interpersonal risks” where a Black professional’s directness is labelled as “aggressive” rather than “spirited.”
The Cost of Black Excellence research measures what Edmondson’s framework does not: whether a specific professional in a racialised body is safe to challenge status quo without racial consequences.
The Reality of Workplace Silence: What 58% Actually Means
According to Williams (2026), a majority of Black professionals lack the safety to advocate for themselves. Comparative data from the NHS Workforce Race Equality Standard (WRES, 2024) highlights this disparity:
- 71.8% of White staff believe their organisation provides equal opportunities.
- Only 52.7% of Black staff agree—a gap of 19.1 percentage points.
The Credibility Calculation
Black professionals must perform “parallel processing” before every contribution. They must assess:
The “Tool” vs. Human Dynamic: Am I valued for my humanity or just my output?
Stereotype Triggers: Will my passion be labeled as aggression?
The Leadership Paradox: If I hold back, am I seen as lacking leadership potential?
The underlying conditions that produce those numbers are not mysterious. The research identifies the precise mechanism.
White professionals enter conversations with a default credibility assumption. Their contributions are presumed valuable until demonstrated otherwise. They speak without calculating how racial stereotypes might interpret their words. They can be direct, assertive, or challenging without triggering a racialised label. They navigate disagreement without their race being a variable in how the disagreement lands.
Black professionals enter those same conversations knowing their credibility requires constant re-establishment. Their contributions are presumed questionable until proven valuable. Direct communication risks the “aggressive” label. Assertiveness becomes “difficult.” Disagreement confirms “not a team player.” Passion becomes a liability rather than evidence of investment.
The research captured the impossible standard this creates in two participant accounts that deserve to sit next to each other.
“I know what will go wrong but can’t say it without being labelled ‘not a team player’ or ‘too negative.’”
“If I speak with passion, I risk being labelled ‘aggressive,’ but if I hold back, I’m ‘not a leader.’”
Read those together. There is no version of speaking in those conditions that does not carry a cost. The calculation is not about whether to take an interpersonal risk. It is about which of two damaging outcomes to accept. That is not a psychological safety problem in Edmondson’s sense. It is a structural racism problem that psychological safety language has been applied to without the structural redesign that addressing it actually requires.
The Calculation before every Interaction
The research documents what happens cognitively before a Black professional speaks in a professional setting. It does not happen once. It happens across every meeting, every presentation, every conversation where a contribution carries any risk of being heard through a racial lens.
Before speaking, Black professionals assess who is present. What is at stake. What version of self is safe in this particular room, with this particular audience, on this particular day. Whether this opinion will trigger stereotypes. Whether silence will register as a lack of capability. Whether the direct version of this point lands differently from a white colleague saying the same thing.
This parallel processing happens in addition to formulating the actual contribution. The cognitive resources it consumes are the same resources available for strategic thinking, creative problem-solving, and the kind of contribution that careers are built on. White colleagues make contributions without running this parallel process. The starting positions are not equivalent. And an organisation that does not account for the difference is not providing equal conditions, regardless of what its psychological safety framework says.
One participant named the cumulative weight of this in terms that go beyond professional frustration: “It’s soo freaking lonely. And scary sometimes. It’s also exhausting trying to figure out how much of myself will be accepted before it’s too much. Feeling like you’re not useful as a human but only as a tool.”
That is not a description of a team environment that needs better meeting norms or more active listening protocols. It is a description of a professional who has learned, through sustained experience, that the full version of themselves is not welcome. And that the partial version tolerated is not a person. It is a function.
WHAT ORGANISATIONS CALL SAFETY THAT IS NOT SAFETY
The research documents several mechanisms through which organisations perform psychological safety rather than provide it.
The first is the open-door policy without structural consequence. A leader announces that their door is always open, that they want honest feedback, that no question is too challenging. The policy sits in the staff handbook and in the onboarding deck. And when a Black professional walks through that door and names what they are experiencing, one of several things happens.
The problem is reframed as a misunderstanding. They are told the organisation takes these matters very seriously, and pointed toward a process that places the burden of proof entirely on them. They are thanked for their courage, and nothing changes. Or, most commonly, the dynamic in the team around them shifts in ways that are technically deniable but are unmistakable to anyone watching.
The door was open. The organisation was not safe.
The second mechanism is the anonymous survey. Organisations collect staff experience data through surveys specifically designed to allow anonymous disclosure of difficult experiences.
The results are aggregated. The overall score is presented at an all-staff meeting alongside a commitment to improvement. The data is not broken down by race. The specific experiences of Black professionals are dissolved into the average. The commitment to improvement targets the average. The conditions Black professionals navigate remain unchanged because they were never specifically examined.
The survey ran. The organisation was not safe.
The third mechanism is the diversity training programme. Unconscious bias sessions. Inclusive leadership workshops. Cultural awareness modules. These interventions address awareness amongst the people who are not experiencing the harm. They do not change the systems producing it. The research confirms this precisely: HR departments implement unconscious bias training that has no measurable effect on the conditions Black professionals navigate daily (Williams, 2026).
The training happened. The organisation was not safe.
The fourth is the Employee Resource Group without authority. Black professionals build the ERG, resource it themselves, run it in addition to their full professional roles, and use it to create the support infrastructure the organisation refuses to provide. The organisation cites the ERG’s existence as evidence of its commitment to inclusion. The ERG has no budget, no decision-making power, and no ability to change the conditions its members are navigating. It is community resilience mistaken for organisational accountability.
The ERG exists. The organisation is not safe.
Black professionals recognise this pattern with clarity. They observe that diversity statements sit alongside ongoing discrimination. They note that equity commitments are announced and then measured by the existence of the commitment rather than any change in conditions. The gap between the performance and the reality is not subtle. It is sustained, documented, and deeply familiar.
What Genuine Psychological Safety Would Require
The research is specific about what safety would actually mean for Black professionals, and it goes well beyond the interpersonal climate of individual teams.
Safety requires that credibility is conferred equitably. The Performance Burden and Proof Burden documented in the Excellence Tax framework exist because the default credibility assumption does not apply to Black professionals in the way it applies to white colleagues (Williams, 2026). Genuine safety removes the requirement to constantly re-establish competence that should have been permanently established by a track record. It means credentials on the wall are read the same way. It means the same directness from a Black professional and a white colleague produces the same response.
This is not achieved through interpersonal goodwill. It is achieved through redesigned performance evaluation systems, structured promotion criteria, and pay transparency that removes the space for differential treatment to hide in apparently neutral processes.
Safety requires that emotional expression is not racially policed. The emotional regulation tax documented in the research describes the demand to suppress authentic responses to maintain white comfort. Black professionals cannot express frustration with discrimination, or show anger at microaggressions, without risking the “difficult” or “aggressive” label (Williams, 2026). White colleagues express the same emotions without racialised consequences.
Genuine safety does not require Black professionals to perform a narrower range of emotional expression than the range available to white colleagues. It requires organisations to notice when they are holding Black professionals to a different emotional standard, name it explicitly, and hold leaders accountable when it continues.
Safety requires that disagreement does not cost disproportionately. When a Black professional raises a concern, challenges a decision, or names a problem, the response should be engagement with the substance. Not a recalibration of their relationship to the team. Not a shift in how their performance is perceived. Not an informal adjustment in how they are treated that is too small to formally document but too significant to ignore.
Organisations build this through explicit accountability. Not policy statements, but consequences for leaders who respond to legitimate challenge with racialised retaliation, however subtle.
Safety requires culturally competent support structures. 52% of Black professionals lack access to culturally sensitive support, including Black therapists, safe workspaces, and culturally aware leadership (Williams, 2026). A professional navigating the conditions this research documents, then offered access to a generic Employee Assistance Programme staffed by practitioners who do not understand racial trauma, has not been given support. They have been given the appearance of support.
Genuine organisational provision means funding access to practitioners who understand the specific experience being navigated. This is not aspirational. It is the minimum that the documented harm requires.
Safety requires that silence is not mistaken for safety. The research identifies three forms of voice suppression: anticipatory silencing, where Black professionals pre-emptively withhold input to avoid retaliation; strategic silencing, where opinions are shared only in spaces calculated as safe; and complete silencing, where perspectives are no longer offered at all (Williams, 2026). Each operates quietly. None produces visible conflict. All produce invisible loss.
An organisation where Black professionals are silently absorbing the cost of speaking and choosing not to is not a psychologically safe organisation. It is an organisation where the conditions for safety are absent and the professionals most affected have learned to stop trying to demonstrate that.
The measure of psychological safety for Black professionals is not the absence of reported incidents. It is whether the conditions that generate anticipatory silencing have been removed. That requires asking a different question than whether staff feel the organisation takes these matters seriously. It requires asking whether they have any evidence from their own experience that it does.
WHAT THIS LOOKS LIKE IN PRACTICE
One participant named what they wished people understood with words that contain an entire organisational agenda: “That it is not easy. We often get pushed back, our experiences invalidated by those we often think or feel we should be able to trust. It requires a lot from us. And often people only notice when we are burnt out or failing to function in the way we have done before.”
The last sentence is the one organisations consistently fail to act on. They notice when the professional is already depleted. They do not notice the sustained conditions that produced the depletion. And by the time the breakdown is visible, the opportunity for prevention has long passed.
Another participant, a Director in the public sector, named the daily cognitive cost simply: “Constantly being the ‘only one’ in the room and having to judge whether or not to confront or raise ‘race issues.’”
That is a safety problem. Not the kind that a team charter resolves. The kind that requires a fundamental change in who is in the room, what happens when race issues are raised, and whether the professional raising them receives institutional backing or institutional silence.
A third participant named what equity would need to mean at the most basic level: “Being treated ‘the same’ diminishes my unique needs and strengths. I’m always a minority.”
That observation dismantles one of the most common organisational deflections: that treating everyone equally constitutes fairness. It does not. Treating everyone the same in a system built around one group’s cultural norms as default protects the system rather than the people navigating it at a disadvantage.
Genuine psychological safety for Black professionals is not sameness. It is equity. It is conditions that account for the specific disadvantages the system imposes and actively work to reduce them, rather than offering a universal policy and calling the work done.
What the Excellence Tax Audit Measures
The Excellence Tax Audit at costofblackexcellence.com/audits was developed specifically because the standard measurement frameworks available to organisations do not capture what this research documents.
Engagement surveys measure whether staff feel valued. They do not measure the cognitive labour of the calculation before every contribution.
Pulse surveys measure general culture temperature. They do not measure the rate of anticipatory silencing.
Inclusion indices measure representation and policy existence. They do not measure whether the 58% figure in this research is replicated inside a specific organisation’s workforce.
The audit provides the measurement that converts the gap between psychological safety rhetoric and Black professional reality into a baseline that can be tracked, reported, and acted on. It connects directly to the research findings and gives organisations the diagnostic they need before the talent exits and the question becomes why rather than how.
The 58% who do not feel safe speaking are not a DEI problem to be managed. They are a professional population whose full capability the organisation is failing to access, whose health is deteriorating under conditions the organisation has the power to change, and whose silence the organisation is currently mistaking for satisfaction.
That misreading is expensive. The research quantifies exactly how expensive, at £61,140 per Black professional employed annually (Williams, 2026). And it is entirely preventable, once an organisation is willing to measure what psychological safety actually looks like for the people it claims to include.
REFERENCES
Equality Act 2010, c.15. London: HMSO.
Williams, N. (2026) The cost of Black excellence: The excellence tax. Birmingham: The Cost of Black Excellence Research Institute. Available at: https://www.costofblackexcellence.com/report (Accessed: March 2026).
Natasha Williams
