You know what a bully looks like. You can even point one out in your classroom. It’s Sean; the big kid, right there, who is constantly in trouble for being physically aggressive and pushing other kids around.

You know how to deal with bullying, too. Your school’s behaviour policy has a clear set of escalating sanctions to apply: first comes detention, then isolation, and finally, exclusion. Meanwhile, the school runs an anti-bullying programme to dissuade pupils from bullying, and to make clear the reporting process if any student is bullied.

And it all seems to work. Last time you checked, this was backed by research, too. Most schools have approached bullying prevention along these lines since 1993, when psychologist Dan Olweus - widely recognised as the pioneer of bullying research - published Bullying at school: What we know and what we can do, which put forward the theory that school bullies are young people from dysfunctional families, who go on to become dysfunctional members of society - and who can be identified and managed as such.

But how well does your anti-bullying programme really work? How closely does that research really match what you see in your classroom, day in, day out?

Because more recent studies suggest the view of bullying in schools could be very wrong. They indicate that bullying is in fact far more complex that we ever thought, that there is no “type” of pupil who bullies - in fact, it can be just about anyone; and that bullying can take many different forms, some of which are invisible.

These latest findings also suggest how we react to bullying in schools is not just likely to be ineffective, it is also almost certainly not directed at everyone it should be and could even make things worse.

So: where does that leave schools?

Professor Ian Rivers has been studying bullying for two decades and even he says this is an area of bewildering complexity. For starters, no one has yet agreed a universal definition of what bullying actually is.

That’s not to say people haven’t tried. Rivers, the professor of education for social change and senior vice-dean for the faculty of humanities and social sciences at the University of Strathclyde, served on a federal expert panel - convened by the US Centers for Disease Control and Prevention (CDC) - that aimed to develop a uniform definition of bullying in 2014. The definition they came up with was based purely on behaviour, rather than individual identities.

“We used to talk about ‘bullies’ and ‘victims’. Now we talk about ‘children who bully others’ and ‘people who are bullied’, so we treat it entirely as a behaviour,” Rivers says.

That behaviour is not always the stereotypical intimidation or physical abuse.

“What makes a bully? This is the problem: I don’t think we’ve ever got a handle on it,” he admits.

For instance, take the example of social isolation, where one student simply decides to stop speaking to another, or decides not to include them in games. Not only can this behaviour be difficult for teachers to see, it can also be difficult to respond to.

“How do you challenge that?” Rivers asks. “Because nothing is said, nothing is done, there’s no bruises. And that becomes very, very difficult. And I always give that example when I’m talking to teachers about what do you do when nothing is done. Because the entire [bullying] narrative is about somebody doing something to someone else. But what about when you withdraw all of the behaviour and you just say that I’m not engaging with that person?”

Anti-bullying week: What is bullying?

This type of bullying is actually more common in schools than the physical intimidation we normally associate with the term. Research by the Department for Education in 2016 found that name calling was the most common form of bullying, making up for just over a quarter of all bullying (26 per cent), followed by exclusion from social groups (18 per cent). Acts of violence, meanwhile, were one of the least common forms of bullying, making up for just 10 per cent.

The trouble is that this 10 per cent is the easiest for teachers to spot. Meanwhile, something like social exclusion can be put down to teenage friendship troubles or social issues - not bullying. It’s also less likely to be reported: children may not perceive name calling or just being “left out” as bullying, as often they too buy into the notion that bullying is a physical act. In those DfE stats, it is likely to be under-represented.

There are likely other forms of bullying not even on the DfE lists, too. In a talk to the American Psychological Association in 2017, Dorothy Espelage, professor of psychology at the University of Florida, who advises members of the US Congress and Senate on bully prevention legislation, pointed out that what counts as bullying is “nebulous” and “subjective” and all children have “different levels of interpersonal sensitivity” to behaviour directed towards them. This makes it a challenge to get an accurate picture of how much bullying might be going on in your classroom and what may constitute bullying for each individual.

A clear, school-level definition of bullying looks all the more needed, then, but all the more complicated to create. And that’s only the beginning of the problem for schools: reacting to bullying is not just hard because defining it is difficult, but because finding the perpetrator is also much more complex than many might imagine. The worst perpetrators can often slip under the radar, says Espelage.

“Developmental psychology literature shows very clearly that there’s two types of kids that engage in high rates of bullying,” Espelage says. “The first group is the Hollywood depiction: the socially ineffective, bigger kids who pick on little kids. What they don’t depict [so much] are those kids that have high social capital. They may be over-represented in the popular group. They are what we call ‘machiavellian bullies’. They know how to turn it on, turn it off; they have heightened social skills; they know who to go after and who’s not going to turn them in.”

These “machiavellian” bullies can often go undetected not only because they have the social skills needed to conceal their behaviour, but because they also have the support of other pupils in the class, who will collude with them or cover for them, which means that they are not the ones who end up in the headteacher’s office.

Schools tend to think of bullying as an individualised behaviour - committed by one or more pupils against another - which it’s not, says Espelage. The research suggests that it is a “group phenomenon”.

“Kids play different roles, whether they are the assister, whether they are the defender - whether they egg things on; they keep things going,” she explains. “And the actual ring-leader bully might not have to do the work because they have their - the kids have all sorts of sayings for this - posse, entourage, bully club.”

Even those not in that “club” can be part of the problem, too. Bystanders can be key to bullying: as Jaana Juvonen, a professor of developmental psychology at UCLA, points out, bullying rarely happens without an audience.

“The audience boosts [the bully’s] status and makes them feel more powerful,” Juvonen says. “Ridiculing a kid alone with no witnesses is only going to give them so much sense of reward, whereas putting someone down in front of a big group of other kids is a totally different story, so the bystanders actually play a huge role in this.”

So, theoretically, an entire class could be playing a role in maintaining a culture of bullying, even if only through bearing witness.

Challenging convention

All this raises questions about the typical approach to both prevention and reaction to bullying in schools. You have a much more complex web of relationships to try to influence and address.

On the prevention side, you would think that whole school anti-bullying programmes may address some of the above complexity by ensuring students are less complicit: either by not participating directly in the bullying or not standing by and watching it happen. Unfortunately, that is not the case, says Rivers.

“Existing interventions don’t seem to have much effect simply because what we don’t know is what types of bullying are immune to the generic interventions that are available,” he says. “Generic, off-the-peg interventions will not work if the issue that you’re dealing with is particular to your geographical or socio-economic or cultural context.”

While there are schemes out there that have been proven to reduce bullying (such as Second Step and Olweus Bullying Prevention), even these schemes do no go far enough, Rivers argues.

“At the very best, some of the interventions are claiming [a reduction of] 50 per cent, but that’s still 50 per cent that’s not being tackled. And I’m saying that what we need to do is stop looking at what has been done and start understanding the 50 per cent that seems to be immune to every intervention we throw at it,” he says. “So, what aren’t we actually addressing?”

On the reaction side, there are problems, too. Luke Roberts is a University of Cambridge PhD candidate whose research focuses on bullying in schools. He has begun to explore why the ways that schools typically respond to bullying might not be working.

“I looked at the exclusion data for bullying this year and permanent exclusions are lower than last year, but fixed-term exclusions are going through the roof for bullying. So that tells you that actually the default setting of most schools is being punitive,” he says.

That’s problematic in light of the research: you are unlikely to have the key perpetrator, nor everyone involved, as the focus of the sanction, so it’s unlikely to have the impact on preventing further bullying you might expect. And Roberts says there is a further issue with this approach.

“Bullying is an ongoing relationship dynamic. So, to try and do a punishment, such as a detention or an isolation room, doesn’t disrupt the ongoing dynamic and I think that’s the hardest thing for schools to get their heads around,” he says.

Any child can be a bully

You may be beginning to think that the situation is hopeless, but while the research brings problems, it is also trying to bring solutions, too.

The main idea being mooted is that we need to treat every child as a potential bully and to start looking at the possible environmental and social factors in schools that can turn them into one, rather than just relying on broad messages for prevention.

“I always say bullying is an equal opportunity behaviour; that given the right conditions, [all] kids can engage in bullying behaviour, and so the conditions really matter,” says Susan Swearer, professor of educational psychology and co-director of the Bullying Research Network at the University of Nebraska-Lincoln.

“Kids move in and out of these roles over time. I think that’s really important as we think about interventions, particularly punitive interventions. Often our strategies that are used in schools don’t take that into account and so they’re really missing the complexity of this type of social behaviour. It’s not a static problem: ‘once a bully, always a bully.’”

This holds true for Juvonen, also. “Could we create conditions where any individual would bully another? I tend to think we could,” she says.

With this in mind, the best form of intervention, it is argued, is to identify and tackle the conditions that are most fertile for bullying. Unfortunately, no one has yet found a definitive answer as to what those conditions might be. But Roberts has some ideas from his recent research.

“If schools don’t have cultures that are supportive, nurturing and enabling, then young people will find ways to create their own hierarchy,” Roberts says. “They’re trying to figure out how to structure the world in a way adults have already learned, because we have job titles and organisational charts etc. Whereas, for young people, we don’t explain that. We just kind of say ‘over to you’ sometimes.”

For Roberts, it is crucial that teachers address the underlying social structures that exist within peer groups, through talking explicitly about “power” with pupils and teaching them about leadership - something he believes that schools do not currently do enough.

But for those discussions to be worthwhile, school leaders need to first consider the wider culture of their schools, including the behaviour that teachers are modelling.

“There’s no point having that conversation with students and then you walk down the corridor and you’ve got the deputy head who’s in charge of bullying just screaming at everyone,” Roberts explains. “It’s so easy to look at individuals and go ‘there’s something wrong with them’, rather than asking, ‘What is the environment we’ve created for that situation to occur?’”

One environmental factor that Roberts is highly critical of is taking a “zero-tolerance” approach to behaviour. Such strict, punitive methods encourage staff to rely too heavily on the school’s system of sanctions, he suggests, rather than on having outcome-focused discussions with pupils about issues of behaviour. A restorative approach, he argues, is more effective in tackling bullying.

Espelage also believes a restorative culture in schools may be useful. In her talk to the APA, she explains that when she first started applying anti-bullying strategies in schools, she was working under a “zero tolerance” model, which took a “three strikes and you’re out” approach to bullying that could result in offending pupils being suspended. But this didn’t seem to solve the problem.

“Remember who’s going to be suspended; who’s ending up in the principal’s office. You’re actually not getting those kids that are driving this group phenomenon of bullying. So the punishment route has not worked,” she says.

And while Espelage concedes that school leaders may feel under pressure to take punitive approaches - particularly from parents of the pupils who are targets of bullying - she says that, in the long-term, different approaches are probably needed, which take into account how the “climate” of a school can promote bullying behaviour.

So restorative approaches are tentatively being touted as a route not just to bullying resolution but modelling relationships to prevent bullying, as well.

Juvonen has another cultural element to add to the mix, too. She has a theory, which she is currently testing, that schools with more diverse intakes are less likely to experience widespread problems with bullying.

She concedes that it is too early in her research to draw firm conclusions, but she speculates that homogeneity in schools may be responsible for creating conditions where a single social hierarchy, established through “mean behaviour” and where there is only “one way to fit in”, will emerge. This, in turn, leads to stronger enforcement of group norms.

Pupils who are different, because they wear the “wrong” trainers, or have a different skin colour, or have a disability, really stand out in homogeneous environments, making them easy targets.

“What I’m proposing is in those kinds of environments we’re going to get more of a ‘policing’ kind of climate,” Juvonen says. “And this is one functioning of bullying: to attack these individuals who do not fit in with the group norm.”

On the other hand, she says, in schools with more diverse student populations, there are “multiple norms operating”, making it harder to determine who “fits in” and who doesn’t, leading to a safer environment for all.

Schools have little control over their catchment, however. That said, inclusion - at least from a SEND perspective, is something schools have a degree of influence upon. And Juvonen suggests that, in the future, anti-bullying strategies could help teachers to consider the mix of pupils in their classrooms, posing the question: “Is this a mix that is likely to generate somebody standing out in the wrong way or is this a good mix where we can see that even those who would typically not fit in have a better chance of being included and accepted by others?”

Of course, this kind of “social engineering” won’t be possible in all classrooms, Juvonen admits. But it is “even more important” in these instances that teachers receive training on specific strategies that will help to minimise the differences, she says, adding that cooperative learning practices have been proven to be particularly helpful here.

“When kids with disabilities, kids of different races, are working together towards shared goals through those kinds of activities, in all likelihood they come to share more and get to know one another better and, in fact, those have been shown to reduce prejudice.”

If you are after hard answers on bullying, the current research is perhaps offering concrete problems with few concrete solutions. But knowing that the path many schools are taking is likely to be flawed is at least a start, and Juvonen says the tentative suggestions from the research are as much of a safe bet as the things schools are already relying upon.

“I think that we have actually a nice range of options for school personnel to rely on at this moment, some still experimental,” she says. “But I think it should be as much relied on as these big and sometimes very bulky curricula that we use in schools to do anti-bullying programmes.”

Helen Amass is deputy commissioning editor for Tes. She tweets @Helen_Amass

Further reading

• Gladden, RM et al (2014) “Bullying surveillance among youths: uniform definitions for public health and recommended data elements”

• Juvonen, J and Schacter, HL (2017) “Bullying in school and online contexts: Social dominance, bystander compliance, and social pain of victims”, Handbook on Intergroup Relations in Children and Adolescents (315-332)

• Juvonen, J, Kogachi, K and Graham, S (2017) “When and how do students benefit from ethnic diversity in middle school?”, Child Development, 89 (1268-1282)

• Olweus, D (1993) Bullying at school: What we know and what we can do

• Rivers, I, Duncan, N and Besag, VE (2007) Bullying: A handbook for educators and parents