The most consistent finding in the organisational culture pathology literature is that toxic environments are rarely sustained by malicious intent. They are sustained by the ordinary psychology of conformity, avoidance, social proof, and the specific fear of being the person who names what everyone already knows. Understanding these mechanisms is the prerequisite for addressing them — because interventions that treat toxicity as an attitude problem rather than a structural one will consistently fail to change the mechanisms that are actually producing it.

The Schein assumption layer: the toxicity that cannot be seen from inside

Schein’s (1985) three-level model predicts the primary persistence mechanism. Organisational culture operates at three levels: artefacts (the visible structures and behaviours), espoused values (the stated beliefs and principles), and basic underlying assumptions (the taken-for-granted beliefs about how the world works and what produces success and safety). The three levels are not equally visible. Artefacts are fully visible; espoused values are explicitly stated; basic assumptions are largely invisible — they operate as the taken-for-granted background of all behaviour rather than as consciously held beliefs.

Toxic culture is anchored in the assumption layer, which is why it can be recognised as dysfunctional at the artefact and espoused values levels while persisting through the continued operation of the underlying assumptions. The organisation that knows it has a transparency problem, acknowledges it in leadership meetings, and puts transparency in its values statement has recognised the toxicity at the artefact and espoused values levels. If the underlying assumption — that information is power and sharing it is dangerous — has not changed, the transparency problem will persist because the assumption continues to generate the behaviour that produces it.

The assumption layer is invisible precisely because it is so thoroughly internalised that it presents as reality rather than as belief. The assumption “here, raising concerns is career-limiting” does not feel like an assumption to the person who holds it; it feels like an accurate description of how things work. Changing it requires making the invisible visible — which is structurally difficult because the culture’s voice-suppression mechanisms are simultaneously preventing the conversations that would make the assumptions explicit.

The Cialdini descriptive norm cascade: toxic behaviour as social proof

Cialdini, Reno and Kallgren’s (1990) descriptive norm research predicts the self-reinforcing mechanism through which toxic behaviour spreads once it has become the dominant pattern. The descriptive norm — what most people actually do — is the primary social proof signal that tells new entrants and uncertain existing members what is actually expected here. When the toxic behaviour is what most people do, the social proof mechanism ensures it will be adopted as the default by anyone who is uncertain about what the culture requires.

The cascade mechanism explains how one person’s aggressive behaviour, one senior leader’s intimidation, or one tolerated ethics violation becomes the cultural norm. Each instance of the toxic behaviour that goes unaddressed updates the descriptive norm: this is what happens here. Each update increases the social proof weight of the behaviour. Each increase in social proof weight makes the behaviour more likely to be adopted by more people. The cascade does not require anyone to actively promote the toxic behaviour — it requires only that the behaviour be visible and unchallenged.

The new hire who observes that a respected senior colleague regularly dismisses concerns, takes credit for others’ work, or intimidates subordinates and faces no consequence has received accurate information about the descriptive norm. The social learning mechanism predicts they will update their own behaviour toward the observed norm, not toward the stated values. The stated value is the injunctive norm; the observed behaviour is the descriptive norm; culture is the descriptive norm.

The Detert-Edmondson silencing mechanism: voice suppression as self-preservation

Detert and Edmondson’s (2011) implicit voice theories research predicts the specific mechanism through which toxic cultures suppress the corrective information that would allow them to be challenged. The implicit voice theory is the largely untested belief that speaking up to authority is dangerous — that raising concerns will produce retaliation, career damage, or relationship destruction. In genuinely toxic environments, this belief is accurate. The person who raises concerns has observed or learned that doing so produces negative outcomes; their silence is rational self-protection.

The silencing mechanism is the toxic culture’s most effective self-preservation feature because it suppresses exactly the information that the people with power to change the culture need to act on it. The leader who is not receiving honest reporting about the culture’s effects believes the culture is better than it is, because the information environment they are operating in has been filtered by the voice-suppression mechanism. The NHS Mid Staffordshire case documented this at the most consequential possible scale: a culture of fear produced staff silence, silence produced unaddressed patient care problems, and unaddressed problems produced the outcomes that the Francis Report subsequently documented. The staff were not indifferent to patient welfare; they were afraid to report, and the culture had made that fear rational.

The Barsade emotional contagion self-regeneration

Barsade’s (2002) emotional contagion research adds the physiological persistence mechanism. The negative emotional climate of a toxic culture — the anxiety, hostility, and distrust that the culture produces — is continuously regenerated through unconscious mimicry. Team members absorb each other’s negative emotional states automatically, without deliberate transmission or conscious awareness. The toxic emotional environment is therefore self-sustaining at the physiological level independently of any individual’s deliberate choice to sustain it.

This is why toxic cultures that have lost their original source — the aggressive founding leader who left, the specific business pressure that produced the original distress — can persist long after the precipitating cause has been removed. The emotional climate encoded in the culture continues to be transmitted and absorbed through the contagion mechanism, maintaining the affect that the original conditions installed. Changing the culture requires changing the emotional environment, which requires sustained positive affect from the leadership that the contagion mechanism will progressively transmit — not a single culture initiative or a values refresh.

The Uber and Wells Fargo case studies as three-mechanism illustrations

The 2017 Uber culture investigation documented all three persistence mechanisms simultaneously. The assumption layer mechanism: the underlying assumption that performance justifies conduct — that commercial success provides the ethical permission slip for any behaviour that produces it. The descriptive norm cascade: aggressive behaviour spreading as the observed norm as each senior leader’s conduct updated the social proof of what was acceptable. The voice suppression: HR complaint mechanisms actively undermined rather than supported, making the implicit voice theory accurate for the people who experienced it.

The Wells Fargo fake accounts scandal documents the assumption-layer mechanism in its most commercially consequential form. The aggressive, unachievable targets, the fear of job loss, and the implicit norm that survival required compliance were not consciously chosen as a toxic culture design — they were the product of underlying assumptions about performance management that produced the descriptive norm that 3.5 million fraudulent accounts documented. The diffusion of responsibility that Schein’s assumption layer predicts prevented any individual from experiencing themselves as responsible for the collective outcome.

What changes toxic cultures and what does not

The persistence mechanisms together predict what will and will not change toxic culture. Values statements, culture workshops, and leadership rhetoric operate at the artefact and espoused values levels and will not change assumption-layer behaviour. The interventions that produce genuine culture change address the mechanisms: changing the descriptive norm through visible, consistent, publicly-witnessed accountability for the toxic behaviour; reducing the accuracy of the implicit voice theory through demonstrated positive responses to concerns; and making the underlying assumptions explicit through the uncomfortable conversations that name what has previously been taken for granted.

Each of these is difficult, slow, and structurally resisted by the mechanisms they are trying to change. Culture change is difficult primarily because the mechanisms of culture persistence are adaptive — they evolved in human groups to maintain stability and resist disruption. Disrupting them requires sustained, consistent, high-visibility behaviour change from the people with the most authority, maintained long enough for the descriptive norm to update and the implicit voice theory to be disconfirmed.

Books worth reading on this

Snakes in Suits by Paul Babiak and Robert Hare. Babiak and Hare’s account of how specific personality profiles — particularly those characterised by the charm, manipulation, and absence of empathy that produce the most reliably toxic leadership cultures — install the assumption-layer conditions this article describes, and how organisations fail to detect and address them, provides the most psychologically specific available account of how toxic cultures are installed and why they persist despite recognition. Their account of the self-preservation mechanisms that these profiles deploy against corrective intervention maps directly onto the Detert-Edmondson voice suppression and Cialdini descriptive norm cascade mechanisms.

If the dynamics described here are significantly affecting your wellbeing, speaking with a psychologist is the right next step. UK: Samaritans (116 123, free, 24/7). Mind (0300 123 3393). BACP: bacp.co.uk/search/Therapists. Crisis Text Line — text HOME to 741741 (US, UK, Canada, Ireland). International: internationaltherapistdirectory.com.

This article is for educational and informational purposes only. Sources: Schein, E.H. (1985/2017), Organizational Culture and Leadership, Jossey-Bass. Cialdini, R.B., Reno, R.R. & Kallgren, C.A. (1990), A Focus Theory of Normative Conduct, Journal of Personality and Social Psychology, 58(6), 1015–1026. Detert, J.R. & Edmondson, A.C. (2011), Implicit Voice Theories: Taken-for-Granted Rules of Self-Censorship at Work, Academy of Management Journal, 54(3), 461–488. Barsade, S.G. (2002), The Ripple Effect: Emotional Contagion and Its Influence on Group Behavior, Administrative Science Quarterly, 47(4), 644–675. Hofstede, G., Hofstede, G.J. & Minkov, M. (2010), Cultures and Organizations: Software of the Mind, McGraw-Hill. Babiak, P. & Hare, R.D. (2006), Snakes in Suits, HarperCollins.