Findings

Most schools are working towards a bullying‑free environment to some extent or more

Of the secondary and composite schools we visited, 32 percent were doing working towards a bullying-free environment to a great extent, a further 48 percent to were working to some extent, and 19 percent to a limited extent. The spread was a little more positive for primary schools, where 40 percent were doing this to a great extent, 44 percent to some extent and 16 percent to a limited extent.

Figure 2: Most schools are working towards a bullying-free environment[1]

This bar graph is called Most schools are working towards a bullying-free environment.  There are two horizontal bars the top one is called Secondary/Composite on the y-axis and the second lower one is Primary.  The x-axis goes from 0% -100% at intervals of 10% jumps. The colours represent the following, orange to a limited extent, yellow to some extent and green to a great extent. The top bar is coloured orange for the first 19%, the second portion is yellow and represents the next 48% with the last portion being green and representing the last 32%.  The second lower bar is coloured orange for the first 16% the second portion is yellow and represents the next 44% with the last portion being coloured green and being the last 40%

Smaller schools were somewhat over-represented toward the lower end of the spectrum, and so looking at the numbers of students enrolled in schools of each category yields a slightly more positive picture. This distribution also reflects the higher proportion of secondary schools, which tend to be larger than primary schools, in the some extent category.

Figure 3: Number of students in schools by overall judgment

This bar graph is called umber of students in schools by overall judgment. It is one long bar. The y-axis is called number of students the x-axis is defined by colour orange represents to a limited extent this number is 4226, yellow represents to some extent this number is 26010 and green represents to a great extent this number is 19425.

The strongest schools are marked by consistency and coherence

In almost all schools, ERO found some degree of commitment from leaders to preventing and responding to bullying, as well as relevant written policies. The most effective schools were effective across the nine domains in the Bullying‑Free NZ School Framework, consistently implementing a whole‑school approach. They had strong universal approaches, and targeted more intensive support where monitoring and evaluation indicated it was needed. The schools working to some extent had some domains of strength but were also weaker in other domains. The smaller number of schools working to a limited extent had significant weaknesses across the domain.

Figure 4 below shows the range of practice across the nine domains.

Figure 4: Performance in each of the Bullying‑Free Framework domains[2]

This is a horizontal bar graph. The title is Performance in each domain of the bullying-free NZ Framework (% of schools).  The bars are coloured orange for unsatisfactory, yellow for satisfactory and green for Good.  The y-axis has 9 points they are from top to bottom, Data, Student agency, Targeted support, Policies/Procedures, PLD, Climate, Leadership, Whanau and Universal approach.  The x-axis ranges from 0-100% at jumps of 10.  The bars from the top are Data - orange 28% - yellow 54% and green 19%, Student agency orange 26% - yellow 51% and green 24%, Targated support orange 20% - yellow 34% and gren 46%, Policies/procedures orange 17% - yellow 34% and green 49%, PLD orange 14% - yellow 46% and green 40%, Climate orange 14% - yellow 47% and green 39%, Leadership orange 13% - yellow 43% and green 44%, Whanau orange 12% - yellow 50% and green 38% and lastly Universal approach orange 10% - yellow 48% an green 42%.

Evaluation and supporting student agency could be improved

ERO found the biggest gap in schools’ approaches to bullying prevention and response was a lack of effective evaluation and monitoring. Schools often did not have clear and robust data on the incidence of bullying, and were unsure how effective their prevention and response strategies were. Only around a fifth of schools were performing well in this aspect, which relates to the data domain in the Bullying‑Free NZ Framework. Supporting student agency was another area of relative weakness, and targeted support was sometimes less effective due to a lack of data on areas of greatest need.

The challenge that school leaders most commonly cited in conversation with ERO was difficulty in developing effective partnerships with parents and whānau. ERO’s rubric gave a ‘satisfactory’ judgment for those schools who were involving whānau as appropriate in response to specific bullying incidents, and ‘good’ judgment to those whose involvement of parents and whānau was more proactive.

Other challenges cited by schools included students’ reluctance to report bullying, which complicated efforts to understand the nature and extent of the issue. Additionally, leaders and teachers believed cyberbullying was increasing in prevalence, but could often happen undetected, or offsite and outside of school time, making it more difficult to address.

ERO also found nearly a fifth of students had not learned what to do when encountering bullying, and some student comments indicated that they thought the strategies they had learned were not helpful. Finally, in some schools, leaders indicated that accessing or making time for training staff in bullying prevention was a challenge, with other competing priorities.

Bullying experience varied by gender

ERO’s findings, both from the student survey, and from review officers’ onsite discussions with students, confirm that bullying does occur to at least some extent in almost all schools visited.  ERO’s companion report on student survey results includes more detail about students’ reported bullying experiences.

A smaller percentage of students reported experiencing bullying behaviour more frequently. Table 1 below shows different types of bullying behaviour students reported experiencing ‘almost every day’ or ‘1 or 2 times a week’, broken down by gender.[3]

Table 1. Bullying behaviours reported often by students, broken down by gender

Bullying behaviour experienced Percentage of respondents indicating experiencing this behaviour weekly or more often
Male Female Gender‑diverse[4]
Called names, put down or teased 21 12 34
Left out or ignored by other students 14 13 33
Been threatened 9 4 27
Hit, pushed, kicked, punched, choked 11 4 19
Personal things damaged or stolen 7 5 23
Lies or bad stories spread 9 8 20
Nasty messages on phone or computer 4 3 18
Made to do something didn’t want to do 9 5 20

Male students were more likely to report experiencing every kind of bullying behaviour than female students, but the gap was especially wide with respect to being called names, put down or teased, and physical forms of bullying. 

Comparing the proportion of gender‑diverse and cisgender students,[5] both male and female, who identified that they experience at least one of the bullying behaviours almost every day yields an observable difference, with gender-diverse students reporting higher rates of bullying behaviours experienced. Only a small number of respondents identified as gender‑diverse, so some caution is warranted in interpreting these results.[6] However, ERO’s findings do align with The University of Auckland’s Youth’12 health and wellbeing survey findings that nearly one in five transgender[7] students reported experiencing bullying at least a weekly. At minimum, ERO’s findings support other research indicating gender‑diverse young people are more likely to experience bullying than their cisgender peers.  ERO has previously published Promoting Wellbeing Through Sexuality Education, a report focusing in part on how schools can improve their inclusion of gender-diverse students.

Students in more effective schools reported less bullying

ERO compared the difference in reported prevalence[8] of bullying across schools by overall judgment. Students in the bottom group of schools, those working to a limited extent, were more likely to report being bullied, or seeing others being bullied at their school. There was, however, almost no difference in reported prevalence between the schools working to some extent, and those working to a great extent. This is likely due to the input‑focused nature of most of the Bullying‑Free NZ Framework domains. Comparing reported prevalence based on ERO’s judgment against the more outcome‑focused school climate domain, however, shows that in the schools ERO identified as having a better school climate, students reported that they experienced and witnessed less bullying.

Table 2: Reported prevalence of bullying by school climate judgment

School climate judgment Mean percentage of students reporting they had been bullied at school Mean percentage of students reporting they had witnessed bullying at their school
Unsatisfactory 56 77
Satisfactory 47 67
Good 38 55

Almost all schools have some policies but effective schools are more consistent

Appropriate charter values and written policies are a necessary, but not sufficient, condition for effective bullying prevention and response. What matters most in working towards a bullying‑free environment are the deliberate actions undertaken by trustees, leaders, teachers, and students.

ERO found the vast majority of schools recognised bullying as an issue and had explicit policies and procedures for responding to incidents. The implementation of strategies aimed at bullying prevention was somewhat more variable. At a minimum, all schools visited recognised and accepted their legal and moral responsibility to provide a safe physical and emotional learning environment for their students. School charters espoused values related to this – commonly cited values included respect, responsibility, acceptance, and compassion. Often, school values were expressed as Māori concepts like whakamana,[9] whanaungatanga,[10] and manaakitanga.[11]

The most effective schools in this evaluation were distinguished by the commitment of their leadership, the consistency of their approach, and robust internal evaluation and monitoring. Schools with sound internal evaluation practice drew on a range of evidence to make sure they had a good sense of patterns of bullying incidents, and how well their prevention and response strategies were impacting on student safety and wellbeing. They used this information to continually improve their practice, targeting areas that most needed attention.

Leadership was crucial but the level of trustee involvement was variable

School leaders have a pivotal role in promoting a bullying‑free environment. Leaders model inclusive practice, set guidelines and expectations for how teachers manage behaviour, and how they respond to specific instances of bullying. The whole‑school climate is greatly influenced by how leaders demonstrate their commitment to enacting school values and treating others with respect and integrity. Most leaders espoused commitment to bullying prevention, although the level of implementation was somewhat more variable.

ERO found effective leaders were taking a deliberate and strategic approach to bullying prevention and response. They were engaged in the school community and, through both formal reporting and more informal channels, were aware of, and responded promptly to issues and incidents as they arose. Some leaders emphasised the importance of knowing their community and their students’ family circumstances to inform the school’s approach to bullying prevention and response. A few principals also had expertise specifically in restorative practices or behaviour management that informed their leadership and the broader school approach.

Effective leaders were also discerning in their use of specific bullying prevention programmes, employing these when evidence suggested they would support the school’s overall approach, or address issues of immediate concern. Leadership was also crucial in supporting staff capability by prioritising opportunities for relevant professional development.

ERO found trustees’ level of involvement in bullying prevention and response varied from school to school. All boards were involved in developing school vision, values and bullying prevention and response policies. They were also responsible for creating the enabling conditions for bullying‑free environments through their resourcing decisions. This included initiatives such as purchasing bullying prevention programmes for the school, funding additional guidance counsellor positions, or resourcing professional development initiatives.[12]  

More engaged boards of trustees went beyond this to request regular detailed reporting on bullying and other student wellbeing matters, and used this information well as part of strategic and annual planning cycles. ERO found this was more likely in schools with well‑developed cultures of internal evaluation. In a few schools, trustees and leaders talked about the board having a pro‑active role around bullying. Examples included: board members and the senior leadership team developing a bullying prevention and response document with a strengths‑based approach; surveying parents as part of a consultation series on bullying; developing a code of conduct for parents regarding bullying attitudes and approaches. One source also talked about the board and neighbouring marae proactively promoting the school’s bullying prevention messages as part of engaging families and whānau.

In schools with less engaged stewardship, trustees were more passive and relied on leaders to decide what to inform them about regarding what was happening in the school with respect to bullying.

The most common approaches were PB4L and restorative practice

Most schools were implementing a named approach or programme, to at least some extent, as part of their approach to bullying response and/or prevention.

The most commonly implemented approach was some variety of Positive Behaviour for Learning (PB4L). PB4L is an expansive initiative, with several components and stages, so the level of implementation and specific actions undertaken varied considerably from school to school. Just over 40 percent of the schools ERO visited were using PB4L School‑Wide, which is intended to be implemented over three to five years, and comprises three tiers. The tiers move from more general, looking at improving systems, processes and expectations across the whole school, to focusing more specifically on interventions for students presenting with more challenging behaviours, and requiring more intensive support. ERO found schools implementing PB4L School‑Wide were often adapting the programme to fit their context and using tools from the programme to support internal evaluation and monitoring of bullying prevention and response. 

However, we found no clear link between participation in PB4L School‑Wide and the overall quality of schools’ evaluation practice in the schools ERO visited. ERO did find slightly higher rates of reported bullying prevalence in those schools implementing PB4L, but this could be due to a selection effect whereby some schools with higher rates of challenging behaviour implement PB4L as a response. Participation in PB4L is likely to result in improved monitoring of behaviour, which could also drive up reported rates of bullying.

The other approach schools were commonly using was restorative practice, although the formality of this and the level of implementation fidelity varied. Restorative practice specifies a set of responses to incidences of bullying, focused on restoring relationships. It contrasts with more punitive responses. There is a restorative practice programme under the PB4L umbrella, but some schools appeared to espouse restorative practice without necessarily being part of the PB4L initiative. Successfully implementing restorative practice requires staff capability building, and it is important schools implement processes with a good understanding of the rationale and philosophy behind restorative practice. In one school, ERO found that a lack of understanding meant restorative principles were inappropriately applied, leading to an unsatisfactory outcome.  

Schools also selected from a large variety of other bullying‑related, or generally pro‑social programmes, including: Kia Kaha, TravellersKeeping Ourselves Safe and KiVa. Additionally, some schools brought in external speakers or short‑term programmes for specific kinds of bullying or problematic behaviour. Examples include the Police’s Loves‑Me‑Not for abusive behaviour in relationships, and a variety of different programmes or seminars on cyberbullying. It is beyond the scope of this evaluation to assess the effectiveness of these programmes, and due to small numbers, it would be inappropriate to draw any strong conclusions about their contribution to the overall effectiveness of schools’ bullying prevention and response strategies. ERO also acknowledges that approaches and programmes aimed at addressing school climate can take some time to embed and show impact, which highlights the importance of ongoing monitoring and formative evaluation.

Outside of specific programmes, ERO found schools also taught bullying prevention through the curriculum. Most commonly in secondary years this was in Year 9 and 10 health classes, although special character schools also commonly discussed aspects of bullying prevention in religious education classes. In primary schools, social‑emotional learning was more likely to be woven through the regular classroom programme. Many teachers focused on relationships and interpersonal skills, developing empathy, resilience, and how to respond to bullying, both as a victim, and as a bystander. In a few schools, ERO found explicit opportunities for learning about bullying were spread across the breadth of the curriculum. Many teachers also used classroom pedagogy as an opportunity to reinforce bullying prevention messages by modelling respectful behaviour and reminding students of relevant school values.

The level of professional learning and development was variable

ERO found a variable level of explicit professional learning and development (PLD) in schools. In schools that were doing this well, leaders used their internal‑evaluation processes to inform the choice of relevant PLD linked to a coherent whole‑school approach to bullying prevention. Where schools were involved in PB4L, or restorative practice approaches, they often accessed PLD related to these programmes. Where this was done less well, PLD was accessed in a more ad hoc fashion, without necessarily being clearly linked to identified needs and priorities. Many schools had not recently accessed PLD relevant to bullying prevention at the time of ERO’s visit, either because it was not seen as necessary, or because of barriers related to resourcing or finding relevant opportunities.

Often, schools mobilised internal expertise from guidance counsellors or other qualified staff to share effective practice and build consistency of approach in the school. They identified staff members who had a strength in bullying prevention and provided deliberate opportunities for them to upskill their colleagues, as part of professional learning groups, or staff meetings.

Cyberbullying was the most common area where schools brought in external expertise, whether from Netsafe, or a variety of other independent providers. In a handful of cases, parents and whānau were included in cyberbullying PLD, to promote consistent messaging in both home and school environments

Parent and whānau engagement was mostly reactive

More than half of schools directly involved parents and whānau only when a severe incident occurred. Evidence suggests some schools involve the parents and whānau only of the student/s who is/are doing the bullying, while others involve the parents of both the bully and bullied students.

Around a third of schools were taking a more proactive approach to involving parents and whānau in their prevention approach. Typically, schools that demonstrated strong involvement with parents and whānau (and which ERO judge to be very effective against this criterion in the rubric) were also driven by strong leadership and took a school‑wide approach to implementing bullying prevention and response strategies.

Proactive involvement usually took the form of communication from the school to the whānau, whether through information evenings, newsletters, blogs, and invitations to relevant assemblies or seminars on topics like cyberbullying. These schools often used surveys to gather whānau views on bullying or broader wellbeing issues. A few schools did involve whānau in a more genuinely collaborative way, providing opportunities for whānau to have meaningful input into the development of bullying prevention and response strategies and approaches. 

In a few schools, leaders and teachers talked about regular informal contact with parents, for example, at the school gate, via an open‑door policy for conversations with the principal or teachers, which allowed opportunities for parents or teachers to raise any behavioural or wellbeing concerns about students.

A few leaders talked about active involvement with iwi and the local marae, or with parents of different cultures. Two schools specifically leveraged off active engagement with the marae and board of trustee members to promote bullying prevention messages and help build community support for the schools’ approach.

Effective schools made a point of involving students and promoting student agency

School climate and culture are improved by student ownership of bullying prevention and response strategies.

ERO found that the most effective secondary schools provided many opportunities for students to exercise agency and leadership around bullying prevention. Many of these schools had some form of peer mentoring structure in place, where senior students were paired with junior students to help with transitions and provide guidance and support. Students told ERO this helped provide a more welcoming school environment, and having role models reinforced school values of inclusion and respect.

Student‑led groups were the other main way in which many of the secondary schools supported students to contribute to bullying prevention. ERO found many groups focused on sex‑, gender‑ and sexuality‑diversity. These provided safe spaces and support for a population of students often more at risk of bullying. Other groups included health committees, and a ‘Caring and Kindness’ club. The existence of these groups and their activities in the school provided a protective factor that contributed to moving towards a bullying‑free environment.

Student groups from nine primary schools talked about having specific student leadership roles in their school as part of the schools’ approach to bullying prevention. Most were referred to as peer mediators, other titles included peer support, student leaders, student counsellor, and PB4L ambassadors. In general, the student leadership role was to uphold the school’s values and expectations of student behaviour, and to support and be role models for other students.  In addition, peer mediators helped identify and defuse problems in the playground. Some also kept a record of incidents to discuss with a dedicated teacher or school leader at regular meetings where the focus was on incidents and patterns of bullying behaviour in the playground and seeking solutions. Most students believed the peer mediators in their schools did a good job and were sought out for support and problem solving in the playground.

Less effective schools did not provide the same level of support for student agency. There were often still leadership opportunities in student councils or prefect roles, although these tended to be driven by adults not students.

Schools usually responded to incidents of bullying appropriately

In most of the schools we visited, most students were confident that teachers and leaders in their schools would, and did respond effectively to bullying incidents. In a few schools, students expressed a lack of confidence in their school’s response to bullying, saying their concerns were minimised, nothing happened, or teachers ‘gave advice, but didn’t solve the issue’. In a few schools, students said things got worse, because telling a teacher was perceived as ‘snitching’ and teachers had not adequately addressed the issue.

ERO found schools’ response to bullying incidents depended on the severity of the incident and other contextual factors. The BPAG guide provides a bullying assessment matrix to help schools decide on an appropriate response to incidents, but ERO found only a few schools where teachers or leaders explicitly mentioned using this matrix.

As indicated, ERO found many schools took a restorative approach to incidents of bullying. Restorative practice specifies different levels of response depending on the nature of the incident, but all focus on:

  • what has happened and who has been affected
  • holding those who have caused harm accountable
  • providing support to those who have been harmed, and others involved.

(adapted from Te Kete Ipurangi website)

Students were generally aware if their school employed restorative practice, even if they had not themselves been involved in any formal restorative conferences. Some students spoke positively about the use of restorative practice, while a small number favoured a more punitive response. In one school, leaders told ERO they had found it necessary to work with parents and whānau to counter a perception that restorative practice was ‘all talk’. Implementing restorative practice takes time, and a few leaders indicated challenges in shifting teacher practice away from more traditional behaviour management strategies focused on punishment and reward.

More generally, and independent of whether they were implementing restorative practices, schools responded to bullying initially by talking with both bullies and victims to understand what had happened. Students ERO spoke to, and those who completed the survey, agreed that timeliness was very important. They wanted allegations of bullying to be taken seriously, to get all sides of a story, and delay any substantive response until the incident or incidents were well understood.

Students wanted, and expected, teachers or other staff would stop the bullying behaviour from happening, and would provide support to both the bully and the victim. Many students took a compassionate approach, and expressed the idea that teachers should check on the bully’s wellbeing, recognising there could be underlying causes of the behaviour that would need to be addressed. A few students did say they would like to see bullies punished, particularly for more severe incidents.

Most students have learned what to do but do not always put it into practice

Three‑quarters of secondary‑age students completing the survey indicated they had learned, at their current school, what to do when experiencing or witnessing bullying. Eighty‑nine percent of primary students said they had learned what to do at their current school, which suggests that primary schools have somewhat more of a focus on explicitly teaching students how to respond to bullying. In speaking with ERO, students identified three main strategies:

  • reporting incidents to teachers/pastoral staff/other adults
  • walking away/ignoring the bullying
  • or non‑violently confronting the bully, either on their own or others’ behalf.

However, in responding to our student survey, secondary students were much less likely than primary students to tell a teacher if they were being bullied (26 percent secondary, 74 percent primary). By contrast, 66 percent of secondary students would tell their friends. These findings were reflected in onsite student interviews as well. Students told ERO they would first try to ignore the bullying, then escalate to another response if it persisted or worsened.

Having learned what to do when experiencing or witnessing bullying was positively associated with students’ confidence to respond to bullying, although ERO’s onsite discussions with students made clear their level of confidence was heavily influenced by the specific context of the bullying.

In the student survey, ERO asked about the outcomes for those students who identified they had used bullying response strategies they had learned at school. Thirty‑five percent of students who responded to this question said the bullying stopped. However, the most commonly identified outcome (forty-four percent) was that bullying stopped for a while and then started again. Sixteen percent of respondents indicated the bullying continued, and five percent said the bullying got worse as a result of what they had tried.

Schools could do a better job of using data for monitoring and evaluation

The biggest weakness across the schools ERO visited was the level and quality of internal evaluation. Schools were better at monitoring incidents than evaluating their initiatives. Most schools collected administrative data on bullying incidents, stored and shared it through their student management system (SMS). The other major sources of data on bullying were student surveys with a wellbeing or school climate focus. Just under a half of the schools visited specifically used the New Zealand Council for Education Research (NZCER) Wellbeing@School survey, while others developed their own. The schools with well‑developed use of data and evaluation also spoke to students more directly, for example, through focus groups, or otherwise making sure students had opportunities to contribute their views in whatever form they were most comfortable.

Some of the less effective schools relied overly on anecdotal data for monitoring bullying prevalence. This is of concern, as student reluctance to report bullying, and the prevalence of less visible forms, such as cyberbullying, could contribute to schools’ systematically underestimating the extent of bullying behaviours in their school.

ERO found only few schools were doing a good job of evaluating their bullying prevention and response policies, programmes and procedures. Most schools used their monitoring data to some extent to target support where it was needed. However, analysis and sense‑making were limited, and consequently schools did not know enough about how their actions were contributing to a bullying‑free environment. Attributing some level of causality was also a challenge for those schools whose approach was less coherent, as pursuing a variety of disconnected programmes or initiatives could make it more difficult to know what specific actions were making a difference.



[1] Not all percentages sum to 100 due to rounding.

[2] Not all percentages sum to 100 due to rounding.

[3] These are unweighted percentages, as population figures for gender-diverse students were not available. See Bullying Prevention and Response: Student Voice for more detail.
[4] ERO’s survey allowed students an open‑response question to indicate their gender identity. 

[5] Meaning students who identify with the gender they were assigned at birth.

[6] The percentages listed above have a margin of error up to /- 10.2%. For gender-diverse students indicating being left out or ignored, we would have 90% confidence that the true range lies between 14.2% and 34.6%. With the exception of the comparison between gender-diverse and male students reporting being called names, put down or teased, these differences remain after accounting for the margins of error.

[7] The Youth’12 survey used ‘transgender’ as an umbrella term. This is not entirely synonymous with ERO’s preferred term ‘gender‑diverse’, which includes for example intersex and gender‑nonconforming youth. See the reference below for notes on Youth’12 usage.

[8] In those schools for which there were sufficient student survey responses.

[9] To confer mana; respect.

[10] Kinship; sense of family connection or other close reciprocal relationships.

[11] Kindness; generosity; support; hospitality.

[12] ERO’s evaluation was not able to isolate the contribution of resourcing decisions to overall effectiveness.