Why Bullying Endures

By Bea Groves-McDaniel, July 2026
Research Assistant: SAL-9000

More than one in five workers globally will experience workplace bullying or harassment during their careers (International Labour Organization, 2022). Somewhere between one in six and one in seven of all employees in the United Kingdom has experienced bullying at work in the past three years (Chartered Institute of Personnel and Development, 2022). On any given school day in the United Kingdom, it is estimated that around three children in every classroom are being bullied (Department for Education, 2023). These figures are not marginal. They describe the mainstream.

And yet the condition persists, and not only in the statistical sense. It persists culturally, socially, interpersonally — in the comments sections of the very platform on which this essay was conceived. The question this essay asks is not whether bullying exists but why it continues, in the teeth of everything the modern world claims to believe about dignity, rights and mutual respect. The incomprehensibility, as we shall see, is only apparent.

The Grammar of Bullying

Before attempting any explanation, it is worth being clear about what bullying actually is, because precision here matters. The definition most widely adopted in research, refined over several decades by scholars including Smith and colleagues (2019), describes bullying as intentional and repeated aggression in which there is an imbalance of power between perpetrator and target. The power asymmetry is the critical element: it explains why the target cannot simply walk away and why the behaviour, if unchallenged, tends to escalate. The repetition distinguishes bullying from isolated incidents of rudeness or aggression, though both of those things matter and can be precursors.

The definition matters because it locates bullying firmly in the category of social behaviour rather than individual pathology. It is not simply that certain people are bullies and others are not. It is that bullying is a behaviour produced by social conditions and sustained by social processes and can therefore be understood, predicted, and in principle interrupted by changing those conditions and processes. This is not a comfortable conclusion for anyone who would prefer bullying to be an aberration (a thing done by abnormal people) rather than a predictable feature of environments that happen to permit it.

The Machinery of Moral Disengagement

The most psychologically sophisticated account of how normal people come to behave cruelly toward others is Albert Bandura's theory of moral disengagement, developed over several decades from the 1990s onwards. The central insight is that people possess moral standards, and that behaviour contrary to those standards normally activates self-sanctions: the pangs of guilt, shame or remorse that prevent most of us, most of the time, from behaving in ways we know to be wrong. What Bandura argued, and what a large body of research has since confirmed, is that human beings are also capable of selectively deactivating these self-sanctions when the circumstances seem to warrant it. The deactivation is the key move. It is not that the bully lacks a conscience; it is that they have learned to switch it off for specific purposes.

Bandura identified several cognitive mechanisms through which this deactivation operates. These include reframing the behaviour so that it appears less harmful than it is (moral justification), diffusing responsibility so that no single individual feels accountable (splitting of responsibility), blaming the target so that the victim appears to deserve what they get (victim blaming), and dehumanising the target so that normal empathic responses are not triggered (dehumanisation).

Each of these mechanisms is recognisable to anyone who has observed bullying behaviour in any setting, and each provides a kind of psychological permission slip that allows cruelty to proceed without the usual moral friction.

The empirical case for moral disengagement as a predictor of bullying behaviour is now substantial. A meta-analysis by Gini and colleagues (2014), synthesising the results of multiple independent studies, confirmed a robust positive relationship between tendency toward moral disengagement and involvement in bullying behaviour. The association holds across age groups and across different forms of bullying, including the physical, verbal and relational subtypes. The most recent large-scale longitudinal study, published in the Journal of Youth and Adolescence by Romera and colleagues (2021) and following over three thousand adolescents across four waves, found that moral disengagement and the desire for popularity operated as mutually reinforcing psychological forces, each predicting the other over time, and both predicting subsequent bullying behaviour. The desire for status and the capacity for self-excusing are not separate drives; they feed each other.

Dark Traits and Their Distribution

The moral disengagement account explains how ordinarily socialised individuals can behave cruelly under specific social and psychological conditions. But there is a parallel body of research that asks a rather different question: whether some individuals are, by personality disposition, more likely to seek out bullying behaviour regardless of circumstance.

The framework most discussed in recent research is the Dark Triad, comprising Machiavellianism (a tendency toward manipulativeness and strategic interpersonal manipulation), subclinical narcissism (entitlement, grandiosity and need for admiration), and subclinical psychopathy (callousness, impulsivity and shallow emotional responding). All three traits are understood to exist on spectrums in the general population, rather than as discrete categories that people either have or do not have.

A systematic review and meta-analysis published in Frontiers in Psychology by Sui, Malik and colleagues (2026) synthesised the evidence from sixteen independent studies and found that all three Dark Triad traits were positively associated with workplace bullying perpetration. The correlations were not trivial: psychopathy showed the strongest association (r = 0.53), followed by narcissism (r = 0.40) and Machiavellianism (r = 0.35). These figures indicate meaningful effect sizes, and the consistency across studies was notable, even accounting for substantial variation in how different investigations operationalised both the personality traits and the bullying behaviour.

What the meta-analysis also found, however, was that the expression of these traits was moderated by organisational context. A workplace environment characterised by procedural unfairness, weak governance and low psychosocial safety did not merely fail to inhibit dark personality-driven bullying; it actively amplified it. The traits are not merely background vulnerabilities. They interact with conditions.

The Social Ecology of the Bystander

There is a tendency, in popular understanding, to frame bullying as a dyadic transaction between bully and victim. The research literature has moved decisively beyond this framing. The most influential reconceptualisation came from the Finnish psychologist Christina Salmivalli, whose programme of research in the 1990s and 2000s established that bullying is better understood as a group process: a social phenomenon taking place in a specific setting, in which the roles of bystanders are as important as the roles of bully and victim.

Salmivalli's participant roles model distinguished not just bullies and their targets but a range of bystander positions: assistants who actively help the bully, reinforcers who encourage by attending and reacting, defenders who actively support the targeted peer, and outsiders who neither encourage nor intervene. A subsequent study by Waasdorp and colleagues (2011), published in the Journal of Clinical Child and Adolescent Psychology, found that classrooms in which reinforcing and assisting bystander behaviours were more common had significantly higher frequencies of bullying behaviour overall. The implication is striking: the bystander group is not a passive audience. It is an active ecology, and its character determines whether bullying thrives or withers.

This finding has major practical implications because it identifies a leverage point that neither the bully nor the victim alone provides. Interventions that shift the norms and behaviours of the bystander group have been shown, in several large-scale evaluations, to be among the most effective approaches available. The KiVa antibullying programme, developed in Finland and evaluated in multiple randomised controlled trials (Kärnä et al., 2011), specifically targets bystander behaviour, training students to recognise their role in sustaining or interrupting bullying dynamics and providing them with practical strategies for defending targets. The evaluations found meaningful reductions in both bullying and victimisation rates in schools using the programme. The comparable figure for the original Finnish trials was around twenty per cent reductions in self-reported involvement, a finding that has been replicated with varying effect sizes across several national contexts (Hutchings, 2011).

Adults, Workplaces, and the Failure of Systems

The dynamics described above were initially documented in school settings, but the broader patterns generalise. Workplace bullying (defined by the International Labour Organization as repeated unreasonable behaviour that creates risks to health and safety) operates through analogous mechanisms, and the evidence on its prevalence and consequences is now considerable.

The Chartered Institute of Personnel and Development evidence review (2022), drawing on twenty years of research, found that employees reporting higher levels of job demands were four times more likely to report being bullied than those with lower demands. Role ambiguity and role conflict together accounted for over a fifth of cases of workplace harassment. Perceived unfairness (particularly in procedural terms, meaning the sense that processes for handling complaints and allocating resources are applied inconsistently) was identified as the strongest organisational predictor of incivility and bullying.

The consequences for targets are serious and well-documented. Research consistently links experiences of workplace bullying to heightened anxiety, depression, burnout, reduced self-esteem, increased absenteeism and, in the most severe cases, suicidal ideation. Less often acknowledged is the organisational spillover effect: employees who witness incivility and bullying are more likely to replicate that behaviour, particularly if they have experienced it themselves. A workplace culture of incivility does not simply harm its direct targets; it degrades the social environment for everyone in it.

The CIPD evidence review found that the most effective interventions in workplace settings were those combining individual capability-building (particularly what researchers call cognitive rehearsal: structured practice in responding assertively to uncivil behaviour) with systemic changes to organisational culture, leadership accountability and job design. Single-pronged approaches, whether focused exclusively on individual assertiveness training or on policy documents alone, consistently underperformed.

The Question of Evil

It is notable that the academic literature on bullying rarely uses the language of evil. The word appears in the research journals, if at all, only as a deprecated concept, a relic of pre-scientific moralising that has no place in a psychological account. This is understandable as a disciplinary convention, but it is worth pausing over what is lost in the translation.

When someone who has been bullied describes their experience using the word 'evil', they are not making an empirical claim about aetiology. They are making a moral claim about value: this should not be; this is wrong in a way that exceeds the wrongness of, say, an administrative error or an honest difference of opinion. That moral claim is accurate, and the empirical research on causes, however valuable, does not supersede it. To say that bullying is caused by status anxiety, moral disengagement, dark personality traits and institutional failures is not to say it is not evil. It is to say something rather different: that evil is not, or not only, a supernatural force but also a product of identifiable human mechanisms operating in identifiable human conditions. That is a more unsettling conclusion than the supernatural account, in some ways. It suggests that none of us is entirely immune and that the conditions that produce bullying are distributed throughout the social world rather than concentrated in a small class of monstrous individuals.

Understanding and Acting

None of the research literature suggests that bullying is inevitable, ineradicable or beyond the reach of deliberate intervention. What it does suggest is that the problem is more structural and more systemic than the language of individual bad apples tends to imply.

It persists because for the individual bully it often delivers real rewards (status, dominance, relief from status anxiety) and because the social environments in which it occurs frequently fail to name it clearly, sanction it consistently or equip bystanders to act. It persists because organisations routinely prioritise speed and productivity over the quality of social relations and because the costs of bullying fall disproportionately on those least able to absorb them.

The question of whether we understand why bullying occurs is not, ultimately, a theoretical one. If understanding why is a precondition for effective action (and the evidence suggests it is) then the effort of explanation is itself a form of intervention. To name the mechanisms is to identify the leverage points. And the leverage points are real.

The incomprehensibility is only apparent. The reasons are there. The question is what we choose to do with them.

Bibliography

Bandura, A. (1999) 'Moral disengagement in the perpetration of inhumanities', Personality and Social Psychology Review, 3(3), pp. 193–209.

Bandura, A. (2002) 'Selective moral disengagement in the exercise of moral agency', Journal of Moral Education, 31(2), pp. 101–119.

Chartered Institute of Personnel and Development (2022) Bullying and incivility at work: An evidence review. London: CIPD.

Department for Education (2023) Bullying in schools: Evidence review. London: DfE.

Romera, E.M. et al. (2021) 'Bullying perpetration, moral disengagement and need for popularity: Examining reciprocal associations in adolescence', Journal of Youth and Adolescence, 50, pp. 1347–1361.

Gini, G., Pozzoli, T. and Boccato, S. (2014) 'Moral disengagement skills and bullying perpetration: A meta-analysis', Child Development, 85(5), pp. 2004–2021.

Hutchings, J. (2011) 'The Welsh adaptation of the KiVa antibullying programme', Child and Adolescent Mental Health, 16(3), pp. 141–147.

International Labour Organization (2022) World of work report 2022: Building a better future for all. Geneva: ILO.

Kärnä, A. et al. (2011) 'A large-scale evaluation of the KiVa antibullying program', Child Development, 82(1), pp. 311–330.

Waasdorp, T.E. et al. (2011) 'Bystanders matter: Associations between reinforcing, defending, and the frequency of bullying behaviour in classrooms', Journal of Clinical Child and Adolescent Psychology, 40(5), pp. 691–704.

Longpré, N. et al. (2024) 'Dark Tetrad at work: Perceived severity of bullying, harassment, and workplace deviance', International Journal of Offender Therapy and Comparative Criminology, 68(7), pp. 742–765.

Salmivalli, C. (2014) 'Participant roles in bullying: How can peer bystanders be utilized in interventions?', Theory Into Practice, 53(4), pp. 286–292.

Salmivalli, C., Lagerspetz, K. and Björkqvist, K. (1998) 'Bullying as a group process: Participant roles and their relations to social status within the group', Aggressive Behavior, 24(1), pp. 1–8.

Smith, P.K. et al. (2019) 'The definition of bullying', in Smith, P.K. (ed.) Bullying in schools: The power of the group. London: Routledge.

Sui, S.M., Malik, S. et al. (2026) 'Dark Triad traits and workplace bullying: A systematic review and meta-analysis of personality, power, and psychosocial safety', Frontiers in Psychology, 17, article 1738277.

Wang, R. et al. (2017) 'Moral disengagement and bullying perpetration: A longitudinal study', Journal of Youth and Adolescence, 46(8), pp. 1823–1836.