Early life predictors of suspensions from primary school: a longitudinal multi-agency record-linkage study

A/Prof. Kristin Laurens1,2, Prof. Kimberlie Dean2,3, Dr. Tyson Whitten2,4, Dr. Stacy Tzoumakis2,5, Ms. Felicity Harris2, Prof. Vaughan Carr2,6,7, Prof. Melissa Green2,7

1Queensland University of Technology, Brisbane, Australia, 2University of New South Wales, Sydney, Australia, 3Justice Health and Forensic Mental Health Network, Matraville, Australia, 4University of Adelaide, Adelaide, Australia, 5Griffith University, Southport, Australia, 6Monash University, Melbourne, Australia, 7Neuroscience Research Australia, Sydney, Australia


Out-of-school suspension is a key component of many school discipline policies, but is associated with adverse educational, justice, health, and welfare outcomes for excluded students. Exclusion at earlier grades and an over-representation of particular groups of disadvantaged students may intensify the inequalities of outcomes for suspended students. Previous studies have focused predominantly on secondary school exclusions and on sociodemographic predictors, with few investigations of suspension during primary school. This study aimed to identify early life predictors of primary school suspensions using multi-agency data from the New South Wales Child Development Study.


Linked data were available for 34,855 children and their parents, including the 2009 Australian Early Development Census, education, health, child protection, and criminal justice records. Multinomial logistic regression was used to examine associations between sociodemographic, pregnancy/birth, child, and parent factors (identified by the 2nd grade) and any suspension issued during 3rd to 6th grade.


In a fully adjusted model, 18 of 26 factors had significant independent associations with suspension. The strongest associations (odds ratios [ORs] ≥4.0) were observed for male gender, out-of-home care placement, and substantiated child maltreatment, with moderate or large associations (ORs 1.5 to <4.0) for unsubstantiated and subthreshold child protection notifications, aggressive behaviour, Indigenous status, socioeconomic disadvantage, exposure to smoking in utero, parental criminal offending, ≥2 emergency department presentations for physical injury, and problems in school-based responsibility and respect. Small associations (ORs >1.0 to <1.5) were identified for having an emotional/behavioural problem, poor language and cognitive skills (school-based), hyperactivity-inattention, young motherhood (<26 years), and parental mental illness. Lack of English proficiency reduced the risk of suspension.


Using multi-sector information available at the time of entry to school to facilitate the identification of students at risk of early exclusion may help educators and policy makers to deliver targeted preventative interventions.


Associate Professor Kristin Laurens is an Australian Research Council Future Fellow in the School of Psychology and Counselling at Queensland University of Technology, and Adjunct Associate Professor in the School of Psychiatry at the University of New South Wales. Her primary research interests are in child and youth mental health. She has expertise in developmental psychology, epidemiology, and cognitive neuroscience.