Abstract
Addressing school bullying is a matter of social justice, but prevention and intervention efforts worldwide have achieved only modest results. Bullying has traditionally been conceptualised within psychology in terms of individuals’ dysfunctionality, but research now also encompasses bullying based on bias rather than personal characteristics, with identity at its centre. Scholars from various disciplines increasingly propose bullying to be a functional, group-based phenomenon concerned with social positioning, but the traditional definition still predominates. Furthermore, there is no widely-accepted definition of bias-based bullying, and a lack of theoretical guidance for practice. This conceptual paper focuses on the application of the social identity approach and related theories to bias-based bullying, examining the relevance of the traditional criteria. Taking separately the perspectives of targets and perpetrators, a new definition is proposed that draws upon social identity principles, including perceived group memberships and stereotyping. Comparison is made with the proposed new UNESCO definition of bullying. The theoretical perspectives explored offer an alternative conceptualisation of bias-based bullying for guiding research and practice, potentially leading to more effective prevention and intervention programmes aimed at creating greater safety and equity in our schools.
Introduction
School bullying is a matter of social justice (Polanin & Vera, 2013). Its immediate and long-term harmfulness, including its association with lower school engagement and academic achievement, is well-established (e.g. Arseneault, 2017; Ladd et al., 2017), and young people report that feeling safe from bullying is a key aspect of their wellbeing (Powell et al., 2018). Despite concerted prevention and intervention efforts around the world (e.g. Richardson & Hiu, 2018), a recent meta-analysis found that, overall, anti-bullying programmes reduce perpetration by a modest 18% to 19% and being victimised by about 15% to 16% (Gaffney et al., 2021). As welcome as these successes are, there is clearly considerable room for improvement. One suggested reason for the generally weak results is that the predominant psychological research tradition (e.g. Olweus, 1978, 1993) has focussed on individuals’ characteristics, to the neglect of the wider peer group context (e.g. Duncan, 1999; Horton, 2019; Schott & Søndergaard, 2014). Increasingly, however, psychological research is broadening its purview by considering bullying based on perceived membership of specific social groups, dubbed ‘bias-based bullying’ or BBB (e.g. Kim et al., 2023; Mulvey et al., 2018).
Bullying research has largely been descriptive (Bradshaw & Johnson, 2011), with empirical findings outstripping theory development (e.g. Horton, 2019; Maunder & Crafter, 2018). Regarding BBB specifically, a paucity of both theoretical guidance for practice (Earnshaw et al., 2018) and rigorously tested programmes (Ramirez et al., 2023) has been identified. As bullying research widens its scope, and the success of interventions seems to have stalled, it is timely to take a fresh look at the nature of BBB.
While Olweus’ definition of bullying (e.g. Olweus, 1993) has been widely accepted for many years, there is no generally agreed definition of BBB specifically. Proposals vary in how far they adopt the Olweus criteria, but the matter of ‘identity’ is introduced as the core feature. Theoretical consideration of the term is lacking, however. The aim of the present paper is to interrogate proposed definitions of BBB and suggest an alternative based on the social identity approach and related theories. In so doing, a brief overview is presented of evidence that the social identity approach is applicable to children and adolescents, including in the context of school bullying. Candidate criteria for incorporation into a definition are discussed, including consideration of the vexed question of ‘intention’ from a moral decision-making perspective. The proposed new definition draws attention to group processes, and not just individual factors, as crucial for understanding BBB.
The Olweus Tradition
Most bullying research, prevention and intervention efforts, and related teacher education, have adopted Olweus’s definition or some variant (e.g. Andreou et al., 2020): that bullying is intentional, repeated aggression by a more powerful person or persons towards a less powerful individual (‘the Olweus definition’). Whether to include intentionality and repetition has often been debated (e.g. Skrzypiec et al., 2019), but the original criteria are still widely cited and continued to be used by UNESCO’s International Bureau of Education until October 2023 (UNESCO, 2024). Similar definitions are found on many informational websites (e.g. Stopbullying.gov, n.d., in the USA and Anti-bullying Alliance, n.d., UK).
Olweus’s view was that a dysfunctional family background or personal characteristics predispose individuals towards bullying involvement, whether as perpetrators, victims, or both (bully-victims; Olweus, 1993). Subsequent research did confirm that certain individual propensities are risk factors (e.g. Arseneault et al., 2010; Crick & Dodge, 1994; Halpern et al., 2015), but contextual factors for understanding bullying’s aetiology and consequences remained neglected (Cook et al., 2010). Indeed, Olweus regarded it as a myth that belonging to certain social groups was a common reason for being bullied (Horton, 2019). While the Olweus definition can stand independently of his views on causality, the early focus on individual dysfunction was maintained in bullying research for many years.
Definitions of Bias-Based Bullying
Today, bullying researchers recognise that many fall victim not because of personal characteristics but because of who they are perceived to be, in relation to their social groups (e.g. Dixon & Smith, 2011; Earnshaw et al., 2018; Price-Feeney et al., 2018; Ramirez et al., 2023), a phenomenon long studied outside psychology as ‘harassment’ (Meyer, 2014). Various definitions of BBB have been proposed, including or excluding various aspects of the Olweus definition. Significantly, they also generally introduce the word ‘identity’, referring to, for example, a person’s gender, ethnicity, or religion. In this paper, I refer to these as ‘identifications’, which together contribute to overall identity, which is multifaceted (Turner et al., 1987); if ‘identity’ is in quotes it refers to others’ usage.
One definition is that BBB is ‘the intentional, or perceived, use of claimed or perceived identities, such as sexual orientation, race, immigration status, or gender, to target an individual or group’ (Price-Feeney et al., 2018, p. 681: ‘the Price-Feeney definition’). The verb ‘target’ implies aggression, and intentionality is included. Otherwise, this definition deviates considerably from the Olweus tradition, with no reference to repetition or power asymmetry, and ‘identity’ becoming the core feature, covering both self-perceptions (‘claimed identity’) and perceptions of others’ ‘identities’. Although consistent with one sociological definition of identity (Jenkins, 2004), this differs from the social identity approach considered here.
Another, much more discursive, definition that draws on Earnshaw et al. (2018) explicitly makes links with the Olweus definition (Ramirez et al., 2023: ‘the Ramirez definition’). It similarly regards BBB as a form of aggression, but with ‘stigmatized identities and characteristics’ (p. 508) at its centre. Unlike the Price-Feeney definition, it also accepts the Olweus criteria of repetition and a power imbalance, though with the latter specified as social, rather than personal. It also broadly accepts the (controversial) Olweus position that bullying is intentional, though allowing that, in the specific case of BBB, it may not always be so. Finally, these authors move towards creating a definition that is not just descriptive, but theoretical, by reference to BBB as stemming from ‘social dominance, stereotypes, and prejudice’ (p. 508). Social dominance, generally considered as an individual characteristic (Halpern et al., 2015), also plays a part in group-based status-seeking in schools (Adler & Adler, 1996; Duncan & Owens, 2011). Social status, prejudice and stereotypes are all issues that can be considered from the perspective of the social identity approach.
The Social Identity Approach and Children’s Development
Social psychological theories concern the normal behaviour of individuals within social contexts and are crucial for the development and evaluation of interventions aimed at ameliorating societal challenges through behavioural change (Sassenberg & Vliek, 2019). The social identity approach to human social phenomena (Tajfel & Turner, 1979; Turner et al., 1987) was first applied to school bullying some 20 years ago (Ojala & Nesdale, 2004), but has remained rather separate from the mainstream bullying literature (Brenick & Halgunseth, 2017).
What is the Social Identity Approach?
Human beings have evolved a group-based, socially interconnected way of living (e.g. Dunbar, 1998; Fuentes, 2004; Smaldino, 2019), and having a sense of belonging is seen as a fundamental human need (Baumeister & Leary, 1995). Human cognitive processes are based on categorisation (Harnad, 2017) and the major psychological theories of identity recognise this: they are social identity theory (Tajfel & Turner, 1979) and self-categorisation theory (Turner et al., 1987), together sometimes called ‘the social identity approach’. This holds that humans categorise the social world in terms of their ‘ingroups’ and ‘outgroups’ and have a fundamental need for self-esteem based on social status. Individuals fulfil this need by identifying with their ingroups, which they contrast favourably against relevant outgroups. Social comparisons are made on the basis of norms and stereotypes, the latter being recognised in the Ramirez BBB definition. Although stereotypes are generally seen as fixed, biased beliefs about social groups (Britannica, 2023), they need some flexibility to enable adaptation to changing social conditions (Turner & Reynolds, 2012), a phenomenon shown to occur in children (Sani et al., 2003). Norms are informal, often implicit, expectations about how one should behave, and they promote, and are promoted by, stereotypes (Stewart et al., 2021). The social identity approach posits that much self-understanding depends on perceptions of social similarity and difference. People’s values, attitudes, beliefs and behaviours frequently reflect the norms of their ingroups, and someone who deviates is likely to face social repercussions (e.g. Ditrich et al., 2022). Tajfel’s original social identity theory arguably over-emphasised self-esteem and has since become so elaborated that it is less a specific theory than a broad framework, one that has successfully stimulated research highlighting the role of groups in social life (Brown, 2020).
Is the Social Identity Approach Applicable to Children and Adolescents?
Research on the development of identity in children and adolescents has largely focussed on individual identity (Kroger, 2004), but there is also a corpus of research on the development of social identity. Children demonstrate relevant abilities from a very early age, for example, being able to socially categorise and stereotype others by age, gender, native/foreign language, and ‘race’, 1 and becoming increasingly able to self-categorise over the preschool years (Bennett, 2011; Martin & Ruble, 2010). They favour their ingroups, often on the basis of behaviours and other observable characteristics and, by late childhood, understand group identifications in terms of beliefs as well (Sani & Bennett, 2004).
Nesdale (2004, 2017) differentiated his theory of social identity development from the adult-based theory by noting that, at least initially, children are mainly concerned with being accepted and having friends, regardless of group social status. In both primary (elementary) and secondary (high) schools, friendship groups or cliques form which are often hierarchical; while members of lower-status groups do seem content simply to have friends, members of the most ‘popular’ group in a class certainly defend their status and members of ‘second-tier’ cliques desperately seek higher status (Adler & Adler, 1996; Duncan & Owens, 2011). In adolescence, maintaining the status of one’s ingroups becomes very important (e.g. Duncan, 1999; Minton, 2012). Status based on social groups is therefore a prominent feature of school life at all ages.
Social Identity, Prejudice, and School Bullying
Further evidence that the social identity approach is applicable to children and adolescents, including in relation to bullying, comes from developmental research on prejudice (Arthur et al., 2008; Killen et al., 2013; Nesdale, 2004, 2012). ‘Prejudice-related bullying’ is an alternative term for BBB (Menesini & Salmivalli, 2017) and prejudice is mentioned in the Ramirez definition. It covers phenomena such as stereotyping, discrimination and intolerance, and the associated negative emotions (Paluck & Green, 2009). Nesdale (2004) initially studied the development of ethnic prejudice, and social identity principles have now been applied to school bullying, aggression, and social exclusion (e.g. Charters et al., 2013; Nesdale, 2012). Even preschoolers are very conscious of feelings of being excluded (Peltola et al., 2023). Children are sensitive to ingroup norms from 5 years and while they prefer their ingroups, this does not mean they hate outgroups; if this occurs it is unlikely before 6 or 7 years (Nesdale, 2017). Across middle childhood social acumen increases, enhancing children’s ability to manage peer group belonging including, potentially, both engagement in, and avoidance of, bullying (McGuire et al., 2015). An explicit school norm of inclusiveness promotes more positive attitudes to outgroups, even in the face of peer ingroup norms favouring exclusion, but this effect is counteracted if children are accountable to peers for upholding peer group norms (McGuire et al., 2015). Young people who are vulnerable by reason of having a learning disability are protected from victimisation if they feel integrated into the peer group (Marini et al., 2023), and middle-school students with more complex ‘identities’ (identifications and their interrelationships) hold more positive attitudes towards ethnic outgroups (Knifsend & Juvonen, 2014). Children’s developing understanding of groups, identities and norms is also important for determining bystander responses to BBB (Palmer & Abbott, 2018).
Self-Categorisation Theory
Also applied to bullying in a few studies is self-categorisation theory, the second arm of the social identity approach. It addresses how the salience of particular identifications depends on context and influences behaviour accordingly (Turner et al., 1987). Children’s self-categorisations are affected by social context by 4 to 6 years (Banerjee & Lintern, 2000), and play a part in adolescents’ gendered bullying (Page et al., 2015), and in cyberbullying (at least, by university age; Trifiletti et al., 2021).
Similarity/Difference: A Theme Common to Social Identity Theory and Bullying
Ingroup membership is based on similarity while outgroups are those considered different. The notion of difference is central not only to the social identity approach but to bullying (Thornberg, 2015), as reflected in Minton’s (2012) neologism ‘alterophobia’—literally fear of the other. This is evident in the recent phenomenon of ‘furrydom’ (young people harmlessly identifying with cartoon-like animals and sometimes wearing furry costumes); they have attracted widespread social media attacks labelling them as sexual deviants (Roberts, 2022). According to the social identity approach, individuals define themselves as both a unique person (different from everyone else) and in terms of their ingroups, which are contrasted against relevant outgroups. Even when someone from an outgroup is bullied because they are perceived as ‘too similar’ to the ingroup, the underlying motive is to preserve the difference (exclusivity) of the ingroup (Ojala & Nesdale, 2004). Links are obvious with the sociological concept of ‘the other’, whereby an individual or group deemed different is accorded less worth (Guttormsen, 2018).
A theory related to the social identity approach offers a further way of conceptualising the role of difference in bullying: communication accommodation theory (CAT). This holds that communication involves accommodating to one another by dynamically shifting interactional behaviour. This may be variously driven by personal or group-based considerations, depending on one’s goals concerning identity, message-sending, and relationship to the other person(s); Pitts & Harwood, 2015). To associate more closely with someone a person converges (behaves more like them or their perceived group) but to create distance, they diverge (e.g. Soliz & Giles, 2014). Although successfully applied for decades to a wide range of contexts, CAT does not seem to have been applied to school bullying, and in fact has almost exclusively focussed on adults. However, communicative accommodation has been conceptualised as a competence that develops over the lifespan (Pitts & Harwood, 2015), and its relevance for understanding racism and social justice has been acknowledged (Shrikant et al., 2022). Therefore, it has potential for understanding BBB (as well as interpersonal bullying). The signals a person sends (e.g. language, accent, dress, or nonverbal behaviour) are used by others to make judgements about the kind of person someone is and how to relate to them. To take dress as an example, clothing is one way in which people signal ‘who they are’ (Minton, 2012; Safdar et al., 2020), and a young person’s attire may be critiqued because it is interpreted as a signal of belonging to a marginalised group such as immigrants, poor people, or a certain youth sub-culture. An immigrant girl in one qualitative study, whose difference in dress was ‘innocently’ asked about, understood perfectly well that this was a device to marginalise her (Haavind, 2014).
CAT can cast light on the use of descriptors such as ‘stinking’, commonly used in Swedish primary school children’s bullying (Thornberg, 2018). It exemplifies the sociological concept of an outcast person as ‘abject’—thrown aside like smelly rubbish (Duschinsky, 2013). This epithet is also reportedly used against Black youth in the USA as a form of hate speech (Chavez, 2021), suggesting that it is often based on outgroup stereotyping. At other times, though, such derogatory language may be used personally or against an ingroup member who has fallen afoul of ingroup norms, with a fine line between friends’ banter and bullying (Dixon & Smith, 2011; Skrzypiec et al., 2019). Such utterances can be seen as linguistic manoeuvres to create accommodative divergence, in pointing out some unacceptable ‘difference’.
Another theory related to social identity theory, optimal distinctiveness theory, holds that the needs for distinctiveness and similarity are always in tension, and that people work to manage group belonging in order to uphold their self-concept (Leonardelli et al., 2010). Numbers of authors suggest that bullying is goal-directed behaviour (e.g. Volk et al., 2014), and there is robust evidence that the use of bullying is effective in raising one’s social status in terms of being popular and high-profile (Wiertsema et al., 2023); young people themselves report that this is the reason for bullying (Thornberg & Delby, 2019). However, those who bully are also more rejected (Wiertsema et al., 2023), and it has been suggested that while they may be disliked by victims and the wider peer group, they are more liked by their friends (Pozzoli & Gini, 2021), which is rewarding and provides a reason to keep bullying. An example of how someone might actively work to reinforce their ingroup belonging is Ioverno et al.’s (2021) finding that sixth graders (especially boys) who experienced homophobic name-calling were more likely to later bully gender-non-conforming peers.
Summary: The Applicability of the Social Identity Approach to School Bullying
When children begin school, they already perceive their social world in terms of ingroups and outgroups, about which they hold stereotypes. While their understanding of groups and norms becomes increasingly sophisticated, and prejudices may develop, belonging to the ‘right’ groups is a strong feature of school life at all ages. Matters of identity, norms, and similarity/difference that are central to the social identity approach and related theories are highly relevant for bullying.
The importance of these as features of bullying has also been emphasised by researchers from disciplines other than psychology. For example, from a philosophical perspective, Johnston (2015) has discussed the importance of identity and the relational self, while Thornberg’s (2015) sociological research found identity to be shifting and co-constructed with peers. There is therefore evidence from several disciplines that understanding of school bullying is enhanced by considering it as a dynamic group-based phenomenon concerned with attaining and maintaining group belongingness that underpins identity; as such, bullying can be seen as a matter of function rather than dysfunction 2 (Shute & Slee, 2022).
Candidate Criteria for Defining Bias-Based Bullying
Identity
Returning to the Price-Feeney two-pronged (perpetrator and target) definition of BBB, let us first consider the perspective of a target, in social identity terms. They may perceive that their victimisation is based on an identification that they hold, either because they understand that members of that group are particularly subject to being picked on, or because a perpetrator has explicitly attributed the victimisation to their membership of that group, such as through online comments or name-calling (Espelage et al., 2019). Mulvey et al. (2018) examined victims’ beliefs about whether or not bullying they had experienced was based on their ‘identity’ (identifications), finding evidence that the negative effects of bullying perceived as bias-based run especially deeply and are hard to mitigate; this is exacerbated by intersectionality—that is, if a person believes they have been targeted on the basis of more than one identification (e.g. both gender and ethnicity).
Turning to the perpetrator’s viewpoint, the Price-Feeney BBB definition refers to the intentional victimisation of another based on an ‘identity’ (identification) they are perceived to have. This is problematic from the perspective of the social identity approach: someone cannot ‘perceive the identity’ of another since this theoretical framework defines social identity in terms of a person’s own self-concept based on the groups to which they feel a sense of belonging. This can be resolved by replacing the term ‘perceived identity’ with stereotype, as included by Earnshaw et al. (2018) and Ramirez et al. (2023). It should also be borne in mind that sometimes a term relating to a marginalised group (e.g. ‘faggot’) may be used abusively or in a teasing manner towards someone who is not genuinely perceived as belonging to that group (Espelage et al., 2019; Skrzypiec et al., 2021).
Intentionality
The Price-Feeney BBB definition accepts the Olweus criterion of intentionality on the part of the perpetrator, one advantage being that it differentiates bullying from harm caused unintentionally, which could lead to accusations against an innocent party. Another advantage is that intentionality differentiates bullying from aggression more generally (Jia & Mikami, 2018), such as reactive aggression, an anger-fuelled response to provocation (this is nevertheless relevant to bullying, being displayed more often by so-called bully-victims; Runions et al., 2018). For parents, intentionality is a key consideration in bullying (Stives et al., 2023). In practice, the difficulty of assessing intentionality means that although researchers frequently include it in defining bullying, they rarely address it in practice (Jia & Mikami, 2018). The ‘fuzziness’ of the concept and practical difficulty of determining intentionality is such that, despite its potential advantages, some choose to omit it from definitions of bullying (e.g. Skrzypiec et al., 2019; Thornberg & Delby, 2019), or acknowledge that while intentionality is often present, it may not be, and thus is not a necessary criterion for determining that an incident constitutes bullying (as in the Ramirez definition of BBB).
While the impracticality of determining intentionality is the usual reason for not including it in definitions, a further, theoretical, argument is advanced here. In seeking to determine whether a person has deliberately sought to harm another, it is important to distinguish between ‘cause’ and ‘reason’ (Saltzstein & Kasachkoff, 2004). Although, according to social identity theory, the underlying cause for BBB may be a threat to one’s group-based identifications and their associated advantages, this is likely to be opaque to the perpetrator, as social identity processes are generally unconscious, although threats to social identity are detectable by neuropsychological methods (Scheepers & Derks, 2016). Young people nevertheless become increasingly able to reason about bullying (McGuire et al., 2015). Bullying is a moral issue (e.g. Levasseur et al., 2017), and moral judgements are reached through both reasoned thinking and intuitive, emotional processes that may become automatic over time (Saltzstein & Kasachkoff, 2004). Similarly, stereotypes, including negative ones, may develop in young people (Nesdale, 2004) and be used as mental shortcuts (Kahneman, 2011). ‘Implicit bias’, or the unconscious stereotyping of groups, drives responses to them (Greenwald & Banaji, 1995). Someone who experiences a threat to their identity may therefore respond to their emotional discomfort automatically by drawing on an available negative stereotype to select a target to bully, without engaging in much moral reasoning, if any; this possibility is supported by Yeager et al.’s (2011) demonstration that adolescents’ implicit theories underlie a desire for vengeance in response to peer conflict.
If such unconscious processes can play a role in bullying, then this seems a good reason not to include ‘intentionality’ in bullying definitions, despite there also being good arguments for its inclusion. A lack of perpetrator insight may also create difficulty in determining whether they selected a target because of bias. However, they may sometimes consciously use moral reasoning to guide their actions (e.g. deciding whether to bully, selecting whom to bully), and are not then really ‘morally disengaged’ (a term commonly used in bullying research, e.g. Thornberg, 2023), even if reaching a decision others judge as ‘immoral’. In the case of the automatic application of a negative stereotype, ‘moral disengagement’ seems more applicable, although post-hoc justification of such bullying may occur. 3 Relevant too is Horton’s (2019) view that bullying involves ‘the manufacturing of contempt’ (p. 1448; italics added), whereby the ostensible reason for bullying hides the cause—a central one being, according to social identity and other group-based approaches, the perpetrator’s need to reinforce their group-based identifications and the advantages that go with them, although CAT reminds us that there may also be personal causes, depending on the goals of the perpetrator.
Repetition and Harm
In the Olweus definition, aggressive behaviour must be repeated in order to be considered as bullying, although this is another criterion ignored in most research (Jia & Mikami, 2018) and by parents (Stives et al., 2023). Although students have often been exposed to the traditional definition through anti-bullying programmes, they nevertheless express varying opinions about whether single incidents count as bullying, or are just quarrels, depending on whether the target appears to feel they have been subjected to harmful bullying, or just treat it as normal social interaction that they have to deal with (Forsberg, 2019).
The ‘repetition’ criterion is generally taken to refer to a situation whereby a particular individual (or group) selects the same target person more than once. Although adults may accept this as a necessary criterion, students themselves may regard repeated incidents as ‘not bullying’ if the target does not appear harmed by it and still has friends (Forsberg, 2019).
However, repetition may take a different form, that of ‘floating’ bullying, in which a person or group does not repeatedly target a specific victim, but a range of different people. While this might occasionally be based on personal characteristics (e.g. a gang picking on weaker kids to steal their lunch money), in the everyday business of schools various researchers have identified bias as central to victim selection, as young people strive for social positioning (Horton, 2019; Schott, 2014a; Schott & Søndergaard, 2014; Thornberg, 2018; Thornberg & Delby, 2019). This is in accord with the social identity approach outlined here. Since BBB is especially harmful (Mulvey et al., 2018), ‘floating’ bullying calls for intervention even if any particular person has only been targeted once.
What about other seemingly one-off incidents that are bias-based, such as using a racialised or gendered epithet? The impact needs to be understood in the context of the victim’s broader experiences as a member of a marginalised group (e.g. Haavind, 2014). They are likely to have experienced many negative attacks on their ‘identity’ (identification), whether at school, in the community more broadly, or via various media (Matheson et al., 2019). Even a single bias-based incident therefore has a high likelihood of further harming the victim, while also adding to the impact on other marginalised young people who observe it (the ‘potentiality of being bullied’—Matthiassen, 2014), as well as contributing to the normalisation of such behaviour if it is allowed to pass unchecked. A further complication is that a single act such as posting a nasty video online may have its impact many times over as it is shared electronically.
For bullying in general (i.e. not necessarily bias-based), some researchers have argued that the important point is not whether an activity happens more than once (a perpetrator-focussed view), but whether the target is harmed. The inclusion of ‘harm’ in definitions (rather than repetition) allows for a single event to count as bullying, as well as helping to differentiate bullying from banter (Skrzypiec et al., 2021). For research purposes, respondents can be asked directly about their experiences of harm. In response to a specific incident at school, how ‘harm’ is decided is an interesting question. We have seen that, even in the case of repeated incidents, peers may not perceive harm, which may be hidden under a victim’s stiff upper lip, or denied by classmates who lack empathy or feel the person ‘deserved it’. A young victim themself may not have insight into future harm (e.g. educational harm through avoiding school). While the perceptions of all involved are important to consider (Forsberg, 2019), trained school personnel may have a broad overview of what harm has occurred, or is likely to do so. This cannot be taken for granted, however, as there is much evidence that gender-based bullying is often ignored or minimised by teachers (e.g. Rawlings, 2019), and teachers’, parents’, and students’ perceptions of bullying and cyberbullying are heavily influenced by gender stereotyping (Mishna et al., 2021). Worse, some teachers themselves have been known to employ racial and gendered slurs against their students (Simon & Olson, 2014).
A counter-argument to replacing ‘repetition’ with ‘harm’ is that calling a single incident ‘bullying’ could trigger an over-the-top institutional reaction. This perhaps carries less force in the case of bias-based attacks, given their known harmfulness and the need to address them urgently for the safety of all in the school community. Limber (cited in Simon & Olson, 2014) has said that teachers should not wait for a pattern to emerge before addressing an issue. Addressing problematic incidents early may also prevent an escalation of seriousness (Forsberg, 2019).
While there are understandable concerns about over-reacting to incidents, it is not clear that drawing a line between a single and a second incident will solve the problem. Identifying whether a particular incident is ‘bullying’ is rarely as simple as ticking off criteria on a list, given the various perceptions of those involved (such as teachers, parents and students), and the known difficulty of identifying ‘what happened’ and ‘who is responsible’ in a given situation (e.g. Schott, 2014b). A definition that is helpful for research purposes is often difficult to apply in practice, when efforts are being made to distinguish between ‘bullying’ and ‘not bullying’. One interesting variation on the traditional criterion of repetition states that the behaviour may be repeated ‘. . . or is highly likely to be repeated’ (Gladden et al., 2014, p. 7). This would involve a good deal of judgement on the part of school personnel.
Such judgement will inevitably be exercised in seeking to enable responses to incidents that are proportionate. Policies and practices that might assist this include: not adopting zero-tolerance policies that do not consider gravity or context (APA Zero Policy Taskforce, 2008); refraining from referring to ‘bullying’ in initial discussions with students, or even avoiding it altogether if an intervention such as the Support Group Method is used (Young, 1998). The Scottish anti-bullying organisation respectme, endorsed by the Scottish government, maintains that the terms ‘victim’ and ‘bully’ are unhelpful, and that the focus should be on the behaviour and its impact, and changing the behaviour; it explicitly rules out repetition as a necessary criterion for bullying, since single incidents can create an ongoing threatening atmosphere (‘latent bullying’) that is not detected by surveys based on actual events.
Power Difference
A less contentious criterion is that bullying involves a power difference between the perpetrator and target, as accepted by the Ramirez BBB definition. In Olweus’s terms, a perpetrator selects someone whose personality or physical presentation indicates that they will not be able to defend themself. Someone may also have low social power, lacking friends to support them (e.g. Bollmer et al., 2005), and sometimes this may be because they have previously suffered exclusion, not because they are socially ‘dysfunctional’ (Johnston, 2015). Alternatively, they may have become vulnerable by breaking an ingroup norm, or attempting to infiltrate an outgroup, thus risking having a powerful leader or a whole group against them.
There is therefore a range of factors that can affect power asymmetry. In the case of BBB, the focus is on the victim’s lower power by virtue of belonging to a lower-status or marginalised group in comparison with the perpetrator, or at least a judgement is made that they will be unable to defend themself against such an accusation. The Ramirez definition of BBB captures this by reference to stigmatisation: the victim’s vulnerable status fits the sociological view that power is not something an individual ‘has’, but something that stems from group memberships (Schott, 2014b). Such groups may be widely stigmatised in certain societies (e.g. those with disabilities) or they may be viewed with contempt within a more limited context, such as a school or classroom (e.g. if members of the most ‘popular’ clique look down on the ‘nerds’). Group-derived power is also recognised by social identity theory, with ‘banding together’ a strategy for achieving social change, as in #MeToo and Black Lives Matter (Shrikant et al., 2022). In BBB a perpetrator is able to wield greater power than the target by taking advantage of the latter’s minority or oppressed status, presumably anticipating little trouble finding supporters and justifying the bullying to themselves and others; for example, the person ‘deserved it’, on the grounds of the negatively stereotyped characteristics of their perceived or alleged social group membership. Also relevant is Thornberg’s (2018) conclusion that bullying (in general) concerns the selective application of oppressive normativities.
Redefining Bias-Based Bullying
There is no easy answer to the question of which criteria should be necessary for defining BBB. Bullying is a complex phenomenon (Shute & Slee, 2022; Slee & Shute, 2025; UNESCO, 2024), and differences of opinion will continue to exist even between experts (Thornberg & Delby, 2019). Considering the above arguments, the following decisions have been reached here:
although identity lies at the heart of the social identity approach, the word ‘identity’ itself is problematic in defining BBB, and best replaced by reference to stereotyping;
intentionality is difficult to determine, on both practical and theoretical grounds and so should not be a necessary criterion;
harmfulness to the target is a preferable criterion to repetition;
a power difference, however derived, should be included.
An alternative succinct definition to others that have been proposed might therefore run as follows: Bias-based bullying is aggressive behaviour directed towards a less powerful person on the basis of a stereotype of a group to which the target is perceived or alleged to belong, and that causes harm to the target. It is also bias-based bullying if a person believes they have been targeted by such behaviour because another person attributes to them membership of a marginalised social group.
This bears some similarity to yet another definition of BBB: that it is aggression directed towards an individual ‘based on perceived membership of a historically marginalized group, including race/ethnicity, sexual or gender identity, or immigrant status’ (Kim et al., 2023, n.p.). This includes the problematic word ‘identity’, and the term ‘historically’ could be somewhat limiting, in not covering bullying resulting from prejudice against newly-emerging groups, such as peer group cliques or youth subcultures (‘alterophobic bullying’—Minton, 2012). It is also rather ambiguous about whether it is inclusive of a victim’s own perceived group membership; the previously suggested definition is clear about this.
Bias-Based and ‘Other’ Bullying
The label BBB implies that it is distinct from bullying of some other kind, although this is generally not discussed explicitly. The term ‘general bullying’ has been used, although it may not be defined, as in Mulvey et al. (2018). The alternative term ‘interpersonal bullying’ has also been used to differentiate it from BBB (Palmer & Abbott, 2018). These terms could be taken to refer to bullying based on unique characteristics of the target, such as those identified by Olweus (1993), for example, appearing physically weak, or displaying antisocial behaviours. However, experimental evidence suggests that, especially as they grow older, children tend to explain peer exclusion in group, rather than individual, terms (Abrams et al., 2009), highlighting the importance of norms and stereotypes. Nevertheless, it may not be straightforward to determine whether bullying is interpersonal or based on bias.
Suppose, for example, it is an unusual aspect of a person’s physical presentation that has drawn negative attention, such as a craniofacial anomaly (Carroll & Shute, 2005). Is this the interpersonal bullying of an individual, or are they being placed in a stereotypical group such as ‘ugly people’, ‘weirdos’, or ‘disabled people’? This is unclear if a perpetrator is not explicit, though the target (or others) may well form their own opinion about the perpetrator’s reasons. As another example, suppose peers criticise someone’s dress. It could either be that their personal fashion sense is being censured, or that their dress is seen as a symbol of belonging to an outgroup against which the critic is negatively biased. Therefore, while it is possible to suggest (hopefully) useful definitions of BBB, where the line is to be drawn between ‘bias-based’ and ‘not bias-based’ bullying can be elusive. Nevertheless, the presently suggested definition, that has been carefully considered in both theoretical and practical terms, serves to highlight the crucial importance of the group-based psychological processes of perpetrators and targets that have been insufficiently considered in the past, and lays the groundwork for further research that is theoretically based.
It is also noteworthy that, since this paper was submitted, the UNESCO Chair on Bullying and Cyberbullying observed at the World Anti-Bullying Forum in late 2023 that, while the Olweus definition has served well for many years, understanding of bullying has advanced and a new definition (of bullying in general) is now called for. A working group presented a new proposed definition, which takes a holistic, school community view, and according to the Chair, was ‘largely welcomed by delegates at the forum’ (UNESCO, 2024) though there were certainly dissenting voices. The UNESCO proposal shares some commonalities with the views presented here on BBB specifically. Note the references to bullying as a social process, involving a power imbalance driven by norms, as causing harm, with repetition as a frequent, but not necessary, feature, and with intentionality not included: School bullying is a damaging social process that is characterized by an imbalance of power driven by social (societal) and institutional norms. It is often repeated and manifests as unwanted interpersonal behaviour among students or school personnel that causes physical, social, and emotional harm to the targeted individuals or groups, and the wider school community. (UNESCO, 2024).
It is noteworthy that this definition assumes that bullying is always driven by social norms, and does not take any account of possible personal (non-group-based) reasons for bullying. The present discussion makes it clear that either type of motivation may contribute to bullying, but that norms, stereotypes etc. relate specifically to bullying that is based on bias.
Conclusion
Unlike the traditional individualist approach, theories emanating from several disciplines conceptualise bullying as a group-based dynamic process concerned with social positioning. The applicability of the social identity approach to school bullying is well-established, but its potential for informing practice remains unfulfilled. Here, I have used the social identity approach and related theories to reassess definitions of bias-based bullying, suggesting that these need to consider the psychological processes of both perpetrators and victims in relation to the multiple social groups with which they identify, and which they perceive others as belonging to. A new definition of bias-based bullying is proposed, although differentiating bias-based bullying from ‘other’ (interpersonal) bullying is not necessarily straightforward, as a perpetrator’s motives may sometimes be obscure, even to themself. The importance of group belonging, social identity and status, similarity/difference, norms, and stereotyping (including the latter’s potential role in moral decision-making) are all discussed.
Future theoretical research could examine more closely the relationship between bias-based bullying and interpersonal bullying. Focus group studies (e.g. with researchers, teachers, parents, and students) could also be carried out to gain their views about the relative merits and challenges in practice of using the Olweus definition, the proposed new UNESCO definition and the bias-based bullying definition put forward here.
Numbers of authors have suggested that prevention and intervention programmes in schools may prove more effective when social identity is considered (Brenick & Halgunseth, 2017; Reynolds et al., 2020; Turner & Reynolds, 2012). I hope that the present theoretical and definitional suggestions regarding bias-based bullying in particular will encourage new directions for research and practice to address school bullying. This would entail the adoption of a more theoretically-driven mind-set towards bullying research and practice, that takes greater account of dynamic group processes that are central to human social functioning and serve a purpose for the perpetrators of bullying. As an example of how social identity thinking might be applied in schools, Wójcik and Hełka (2019) have presented an innovative proposal for how a newly-formed class might be managed for social harmony.
The proposed changes presented here regarding the theory and definition of bias-based bullying are in line with the recent UNESCO proposal that a new conceptualisation of bullying is needed that is more holistic and inclusive, ‘with the potential to break down academic and professional barriers, encouraging cooperation’ (between diverse stakeholders), placing a particular focus on bullying of the marginalised (UNESCO, 2024), although the UNESCO definition (of bullying in general) fails to take account of personal motives that are not group-based. An in-depth consideration of ways to bring together those working under different paradigms for understanding bullying can be found in Shute and Slee (2022).
Footnotes
Funding
The author received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
