school suspensions

Do we really have a frightening school to prison pipeline in this country? Only one way to find out

Exclusionary discipline is on the rise in Australian schools, as highlighted by recent research in Queensland and South Australia. This is highly concerning that suspension does not address the reasons underlying behaviour and can instead exacerbate those behaviours. For some students, these experiences devolve into ongoing cycles of repeated suspensions. In the long term, students who experience exclusionary discipline tend to have lower educational outcomes than might have been expected and are far more likely to be involved in the criminal justice system.  

Of even greater concern is the increasing body of research which shows that students from minority and marginalised groups are disproportionately represented. Such research emanates largely from the United States, where there has been decades of research showing that African American students receive suspensions for incidents that White students do not and that they also receive harsher consequences for the same infractions. 

Evidence of the entanglement between racial bias, overrepresentation in exclusionary school discipline, and overrepresentation in prison, prompted significant reforms to education and school discipline policy and practice in the United States. In 2014, the Obama Administration, together with the Office of Civil Rights, acted on the evidence by issuing a set of Guiding Principles. These principles reminded schools of the dangers of direct and indirect racial bias in the use of exclusionary discipline and advocated for the adoption of evidence-based frameworks that aim to improve school climate, student support, and teaching quality.

Despite the overrepresentation of Aboriginal and Torres Strait Islanders in our prison population, rigorous investigations of Australia’s “school-to-prison pipeline” are rare. Indigenous Australians represent 3.3% of the total population, but account for 29.6% of the adult prison population.  It has now been 30 years since the 1991 Royal Commission into Aboriginal Deaths in Custody which found that Aboriginal people were “more likely to die in custody” due to their disproportionate representation in the prison system. 

We cannot afford more decades of research to confirm a link between the use of exclusionary school discipline and involvement in the criminal justice system; a link that has already been established internationally. Rather, we need to urgently identify whether Indigenous students are overrepresented in exclusionary school discipline across Australia, why, and, given its many known ill-effects, how to drastically curtail its use.  

Our recent research (with Associate Professor Kristin Laurens, School of Psychology and Counselling, and Centre for Inclusive Education, Queensland University of Technology and Associate professor Naomi Sweller, School of Psychological Science, Macquarie University) aimed to make a foundational contribution by examining trends in suspension, exclusion, and enrolment cancellation incidents in Queensland state schools, using publicly available data from the years 2013 to 2019. We investigated differences among Indigenous and non-Indigenous students in exclusionary discipline incidents proportionate to enrolments, and whether these trends were increasing or decreasing at different rates over time.

What do we know from these analyses?

We found that the use of exclusionary discipline in Queensland state schools has increased significantly for all students between 2013 and 2019. However, when disaggregating by Indigenous status, we found that this rise was significantly steeper for Indigenous students compared to non-Indigenous. There were also significant increases in exclusion rates and short suspensions for Indigenous students, but not for non-Indigenous.

Analysis of trends by year level showed that in 2019, suspension incidents peaked a year earlier (Grade 8) for Indigenous students compared to non-Indigenous (Grade 9). When considering the reasons for suspension, the highest degree of overrepresentation occurred for disruptive/disengaged behaviours, which includes categories such as ‘refusal to participate in the program of instruction’ and ‘absences’. 

We also considered trends by geographic location, finding that Indigenous students were disproportionately represented in all seven regions around Queensland, but most prominently in Darling Downs South West. In 2019, six regions had between 321.8 and 358.3 suspensions per 1000 students, while Darling Downs South West had 487 suspensions per 1000 students. There was no such diversity for non-Indigenous students with rates across the seven regions ranging between 85.4 and 139.4 per 1000 students.

What DON’T we know?

These analyses provide strong evidence that Indigenous students are disproportionately impacted by the use of exclusionary discipline in Queensland state schools. However, there remains much that we don’t know. 

For instance, the QLD Department of Education publishes the number of exclusionary discipline incidents, and not the number of students involved in those incidents. It is therefore impossible to tell how many suspensions went to the same students, and whether Indigenous students are receiving multiple suspensions at a greater rate than non-Indigenous students. 

Nor do we know whether these Indigenous students might also have a disability or be living in out-of-home care due to the absence of data disaggregated by Indigeneity, disability, living in out-of-home care, and gender.  

Our study points to an overrepresentation of Indigenous students in suspensions for infractions involving disruptive/disengaged behaviour, physical misconduct, and verbal or non-verbal misconduct. More research investigating the extent of potential racial bias in the reasons for issuing a suspension is required. 

Moreover, disruptive/disengaged infractions incorporate reasons such as ‘absences’ – i.e., truancy – meaning that there are students who are being excluded from school as a result of not attending school in the first place.    

Further questions include:

  • How much pressure to suspend are principals facing from other parents, teacher factions and the union? 
  • What is the average short suspension length for Indigenous students? Is it closer to the maximum (10 days) than the minimum and is there a significant difference to non-Indigenous students?
  • Why is the suspension rate significantly higher in Darling Downs South West region?
  • Is the peak in suspensions for Indigenous students occurring in Year 8 due to early school leaving?
  • Where do these students go? What relationship is there to juvenile justice involvement?
  • Most importantly, what do Indigenous students say is the reason they are getting suspended and excluded? What do THEY think needs to change?

Why it is critical to have a national inquiry into this problem and a national solution

The 2014 reforms implemented in the United States have had significant impact. One example is Chicago Public School (CPS) which restricted the use of exclusionary discipline by reducing permissible length and banning suspensions for minor infractions. Critically, CPS did not stop at discipline reform, but also undertook systemic inclusive school reform, by adopting Multi-Tiered Systems of Support (MTSS). MTSS is an evidence-based framework that aims to enhance students’ social-emotional learning, academic, and behavioural outcomes.

In the very same year, Queensland went the other way. 2014 brought about legislative reforms that permitted greater autonomy in the issuing of suspensions by principals and schools, while simultaneously abolishing students’ rights to appeal short suspensions and reducing the requirement to consult with parents and their children. As well, alterations were made to the length of ‘short’ and ‘long’ suspensions, with short suspensions being increased in length (from 1-5 days to 1-10 days), while long suspensions were reduced (5-20 days became 10-20 days).

The graph below shows suspension rates per 100 students before and after 2014, for Chicago Public Schools and Queensland State Schools. Notably, CPS rates of suspensions per 100 students dropped from 24.6 in 2012 to 5.17 in 2019. Conversely, in Queensland, suspension rates rose from 11.71 per 100 students in 2013 to 13.4 in 2019, with a peak in 2018 at 15.04. 

Evidence shows positive outcomes from: 

  • implementation of system-wide Multi-Tiered Systems of Support (MTSS)
  • strict limits on and alternatives to the use of exclusionary discipline through legislative thresholds and safeguards
  • respectful and empathic teacher-student relationships
  • educative responses to discipline enacted within an inclusive school culture
  • systematic implementation of evidence-based practices, programs and interventions to support students’ academic, social-emotional and behavioural development.

Most critically, evidence from the US reforms indicates that the introduction of strict limits on the use of exclusionary discipline, such as banning suspension in the early years (K-3) and for minor reasons, as well as the provision of safeguards for priority equity groups (e.g., Indigenous students, students with disability, children in care) is a necessary first step for effective reform implementation. We cannot wait the decades that the US waited to respond to a problem that was staring them in the face. If some individual states won’t act, then the Australian government must.

Linda J. Graham is Director of The Centre for Inclusive Education (C4IE) and a Professor in the Faculty of Creative Industries, Education and Social Justice at QUT. Her research investigates the role of education policy and schooling practices in the development of disruptive student behaviour and the improvement of responses to children that teachers can find difficult to teach.

Dr Callula Killingly is a Postdoctoral Research Fellow in The Centre for Inclusive Education (C4IE) at QUT and a member of the Accessible Assessment ARC Linkage Team (LP180100830). Her research interests include learning and memory processes, language and literacy development, and music cognition.

Shocking evidence of US-style racial bias in Australian schools

Australian research is almost silent on how disciplinary practices in our schools are affected by racial bias. 

In the United States there is ample evidence that children from minority groups are more likely to labelled as having behaviour disorders. They are also more likely to be diagnosed with having a mild intellectual impairment, learning disabilities or emotional disturbance, and placed in special education classes.

Research from the US consistently shows that African American, American Indian and Hispanic students are more likely to be overrepresented if they are:

  • male,
  • from a low SES background,
  • live in a high-density urban area, and
  • where there is a high proportion of students from minority groups.

Similar trends have been noted in the United Kingdom, New Zealand and Canada.

The lack Australian research on this issue is not because we somehow have escaped the problem but because Australian education systems are remarkably eclectic in the ways in which they report data.

Sophisticated longitudinal and geographical analyses tracking trends in diagnosis and placement are currently impossible. We remain ignorant of magnitude, cause and effect. But there are indicators and if you can see the tip of an iceberg, logic suggests that you would make a serious effort to alter direction.

What DO we know?

The NSW Department of Education and Communities (DEC) publishes an extensive array of educational data. Whilst they are riddled with inconsistencies and blind-alleys, DEC does at least publish some statistics disaggregated by Indigenous status.

These data show that Indigenous students are significantly over-represented in long school suspensions (5-20 days) and in separate special educational settings. DEC doesn’t draw that conclusion themselves but it is clearly evident when Indigenous students make up only 6.3% of total enrolments in NSW government schools in 2012 but account for:

24.4% of long-suspensions,

14.6% of enrolments in primary school support classes,

12.6% of enrolments in secondary school support classes, and

12.8% of enrolments in special schools.

These numbers tell us far less than we need to know. For example, is the disproportionate over-representation of Indigenous students in school suspensions and special educational settings increasing or declining over time? Are there discrepancies that might indicate institutional bias or is Indigenous over-representation in these data simply a reflection of social disadvantage?

I looked for evidence of trends emerging. What I discovered has made me believe urgent attention is needed.

Let’s start with the use of suspension: the strongest predictor of later special education placement and school failure.

Long-school suspensions: 2008-2012

Indigenous students accounted for 6.3% of total enrolments in NSW government schools in 2012, but received 24.4% of long-suspensions (averaging 11.8 days), up from 22% in 2008.

They were 5.1 times more likely to receive a long-suspension than non-Indigenous students (up from a risk of 4.3 in 2008), and 6.1 times more likely than non-Indigenous students to receive a repeat long-suspension (no change from 2008).

There was a 35.1% increase in the number of Indigenous students receiving a long-suspension between 2008 and 2012.

This is almost twice the increase in long-suspensions received by non-Indigenous students.

This is a serious problem because suspension is an ineffective and often harmful response to student disengagement that does nothing to address the underlying causes of disruptive behaviour.

As I mentioned earlier, suspension is also the most robust predictor of special education placement and later school failure. If that’s true, then high rates of suspension may be impacting Indigenous enrolments in special education. Let’s have a closer look at this part of the iceberg…

Enrolments in separate special educational settings

longitudinal analysis  of enrolments in separate special educational settings (1997-2007) found that Indigenous enrolments in support classes and special schools are increasing faster than enrolments of non-Indigenous students, and faster than Indigenous enrolments in mainstream.

In other words, the rise in Indigenous special education placements cannot be explained by Indigenous population growth.

This research also found that Indigenous students were already over-represented in separate settings back in 1997, and that the degree of over-representation has increased significantly since. Particularly worrying was the finding that the disproportionate over-representation of Indigenous students had accelerated in the 6 years since the Review of Indigenous Education (2004).

Enrolments in NSW government special schools

So, we know that Indigenous students are over-represented in special schools and that their enrolments are increasing relative to non-Indigenous students. This doesn’t appear to have stirred much in the way of public outcry, so I investigated whether Indigenous disproportionality differs by special school type.

There are three broad types of special schools in the NSW government school sector:

  • Traditional special schools enrolling students with moderate to severe intellectual impairment, physical and sensory disabilities, and autism;
  • Mental health special schools enrolling students with emotional disturbances, severe psychiatric disorders, or behaviour disorders; and
  • Juvenile justice special schools within juvenile justice detention centres.

In 2009, traditional special schools enrolled just over two thirds of special school students, mental health special schools enrolled almost one quarter, and juvenile justice special schools enrolled just under 10 per cent.

Indigenous representation varied significantly by school type with 1 in 4 kids in mental health special schools and almost 1 in 2 in juvenile justice special schools identifying as Indigenous. Less than 6 from every 100 students in Traditional SSPs were Indigenous.

This means that Indigenous disproportionality in special schooling is explained by over-representation in particular types of special schools; namely mental health special schools and juvenile justice special schools.

Now, I know that still might seem unremarkable to some, so I looked a little more closely at mental health special schools. There are two broad types in this group:

  • special schools for students with verified mental health issues, and
  • special schools for students with disruptive behaviour.

The former requires a confirmation of disability (under the category of mental health problems) prior to entry, the other doesn’t.

Indigenous students accounted for 18.8% of enrolments in the type that requires confirmation of disability and 27.1% of enrolments in the type that doesn’t.

In my field of research, that’s more than the proverbial tip of an iceberg.  It’s the equivalent of a smoking gun.

At the very least, these trends tell us that our school disciplinary practices are affected by racial bias and that we need to more carefully examine how discipline is applied, to whom, what for and in what ways.


Linda GrahamAssociate Professor Linda J. Graham is a Vice Chancellor’s Research Fellow in the Faculty of Education at Queensland University of Technology. She is grateful to have received funding for her research into educational responses to children who are difficult to teach from the Australian Research Council (DP110103093; DP1093020) and the Financial Markets Foundation for Children (2013-030). She is presenting this research at the NSW Aboriginal Education Council’s 50th Anniversary Conference, Saturday August 30, 2014.