Annick Rogiers  Founder, Sustainable Wellbeing Solutions - Connecting Psychological Safety, Psychosocial Risk and Organisational Practice

Toxic Leadership & Trust: What Should Organisations Do When Leadership Becomes Harmful?

The same event looks different depending on where you sit and when no function sees the whole picture, nobody can act on it.

Over the past months, The Organisational Capability Series™ has taken me into some difficult territory. Leadership and accountability. Toxic leadership. Trust. Speaking up. What happens when the organisation already knows.

And the hardest one: what happens when leadership itself becomes part of the concern?

I proposed a round table to explore that question with HR, HSE, Prevention and Wellbeing in the same room. The table didn't fill. But people responded anyway in comments, in messages, and by sending me articles they said I needed to read.

What came back was not feedback on an invitation. It was a body of insight that deserves to be tested rather than just collected.

So this is where I test it. Against the research. Against the framework. And against the one question that keeps returning: if the person with the authority to act is part of the problem, what holds?

What did people actually say?

Five things, and they cross every professional boundary.

Behaviour is a system signal, not a personality clash. One respondent, a behavioural performance consultant, put it most directly: "Harmful leadership isn't a personality clash, it's a system signal. If the system tolerates it, it teaches everyone else what is acceptable."

Organisations usually already know. Another said it in one line: "When leadership becomes harmful, organisations often already know. The question is what they do with what they know."

More pathways are not the same as independent pathways. Someone raised that trust breaks hardest not when people are invited to speak, but when they need to speak "about something that involves power, relationships or hierarchy."

Having a system and being able to use it are different things. "The hardest part may not be having policies but using them when power is involved."

And the question almost nobody asks out loud: what happens when the mechanisms designed to protect people are themselves affected by the conditions they are meant to challenge?

Five observations, from five different professional positions. The question is whether they survive contact with the evidence.

Do they? What does the research say?

Largely yes and in one case, more forcefully than the respondents claimed.

Is harmful leadership really an organisational condition, and not just an individual?

Yes. That is now the dominant position in the research, not a minority view.

Padilla, Hogan and Kaiser introduced what they called the toxic triangle in 2007, and their central argument is a direct challenge to the personality-first explanation. They argued that the definition of destructive leadership "should emphasize negative outcomes... rather than focusing on the characteristics of individual leaders."

Their model requires three elements to coexist: a destructive leader, susceptible followers, and a conducive environment. Remove the third and the other two have nowhere to operate.

Later work in the same tradition states the consequence plainly: where governance systems impose no consequence when misbehaviour is signalled, "destructive leaders have free space."

That is not a metaphor. It is a description of an organisational condition.

And how large is the actual damage?

Large enough that no organisation can treat this as a cultural preference.

A meta-analysis by Mackey and colleagues drew on 418 empirical studies covering 123,511 participants. The pattern is consistent: destructive leadership correlates negatively with how people feel about their leader, with wellbeing and with individual performance — and positively with turnover intention, resistance and counterproductive work behaviour.

A longitudinal meta-analysis confirmed the direction holds over time, not just in a single measurement: destructive leadership was positively related to turnover, emotional exhaustion, burnout and deviant behaviour.

This is why "toxic leadership" is not a strong word used loosely. The harm is measurable, it accumulates, and it shows up in the outcomes organisations already track.

Why doesn't speaking up fix it?

Because the interventions that encourage voice appear to underperform and the reason is structural, not personal.

A systematic review of employee voice and silence in healthcare found that "the accumulated evidence concerning speaking-up interventions... indicates that they achieve disappointing outcomes because of a professional and organizational culture which is not supportive."

Read that again. The interventions aren't failing because people lack courage, or because the training was poor quality. They are failing because the surrounding culture doesn't support what the intervention asks people to do.

This is the empirical version of the respondent's point about independence. A speak-up channel that routes a concern back into the same power structure it questions isn't a channel. It's a formality.

What does leadership have to do with psychological safety?

Everything  and the effect sizes are not marginal.

Frazier and colleagues conducted a meta-analysis covering 136 independent samples, more than 22,000 individuals and nearly 5,000 groups. Trust in leadership correlates meaningfully with psychological safety, and leadership behaviour ranks among the strongest antecedents of whether people feel able to speak.

A 2025 study of psychological safety in healthcare put the structural conclusion starkly: psychological safety is "not a cultural accessory but a structural requirement for excellence."

Which is precisely the tension the respondents identified. If leadership behaviour is the strongest driver of psychological safety, then the moment leadership itself becomes the concern, you cannot rely on the very relationship that has become unsafe. Something else has to hold.

And when the protective mechanisms are themselves compromised?

Here the research is quieter and that silence is itself informative.

There is no large evidence base on what happens when the functions meant to protect employees are captured by the conditions they are meant to challenge. But the logic of the toxic triangle points one way: if a conducive environment is what allows destructive leadership to operate, then a compromised safeguard is part of that environment.

The respondent was asking a question the literature has not answered. That is worth saying plainly.

So what should an organisation actually do?

Build infrastructure, not intention.

The evidence points at conditions, not at individuals. Which means the practical question is not "how do we develop better leaders?" but "what conditions make harmful leadership possible here and which of those can we change?"

That reframes the work:

• Recognise what happened without reducing it to one person.

• Connect signals across functions instead of examining them separately.

• Examine how power influenced what people could say and do.

• Test whether escalation routes are genuinely independent.

• Change something in the conditions, not only the personnel.

• Verify afterwards that the change actually worked.

• Learn from the organisation's own response, not only from the event.

Because a lodged complaint is not yet learning. A coached leader is not necessarily adaptation.

Why does this belong inside The Organisational Capability Framework™?

Because three of its domains are tested by exactly this moment — and the research tells us which one fails first.

Organisational Signals asks: what is the organisation already telling us?

Behaviour is a signal. Silence is a signal. Turnover intention — which the meta-analytic evidence links directly to destructive leadership — is a signal. The organisation was reporting the problem the entire time. The question is whether anyone was in a position to connect what different parts of it were saying.

Leadership & Accountability asks: who takes responsibility, who has the authority to act, and can accountability still function when circumstances become difficult?

When responsibility, independence and authority land in different places, accountability breaks — not because nobody cares, but because no single person holds all three. The systematic review of voice interventions found exactly this: even well-designed mechanisms fail in an unsupportive culture. The gap isn't between knowing and caring. It's between caring and being able to act.

Learning & Adaptation asks: what changes because of what the organisation has learned?

And here the respondents added something the framework can name. People weren't only learning about the case. They were learning what happens next time you speak up. Whether values hold when power is involved. Whether accountability applies equally.

The organisational response is itself a learning event for everyone who watched it. That is how capability either strengthens or erodes, one case at a time.

What the round table was actually for. Not to find an answer. To put five partial views in one room.

HR may see an employee relations concern. HSE or Prevention may see a psychosocial hazard. Wellbeing may see the impact on individuals. Operations may see absence, performance or turnover. Governance may see organisational risk.

None of those views is wrong. And none is sufficient alone.

That is the whole argument for bringing them together and it is also the finding: if each function learns something different from the same event and those lessons never meet, the organisation may not learn at all.

I'm still working out what shape that conversation should take. If this question sits inside your work  in HR, HSE, prevention, wellbeing or governance  I'd like to hear what you see from where you sit.

Read the rest of The Organisational Capability Series™ →