Socio-emotional characteristics in early childhood and offending behaviour in adolescence
Categories: ADR England, Office for National Statistics, Crime & justice, Research using linked data, Research findings, Impact, ADR UK Research Fellows, Policy, Practice, People
21 January 2026
This research used data made available via the Office for National Statistics (ONS) Secure Research Service, which is being expanded and improved with ADR UK funding.
Author: Dr Paul Garcia, ISER (University of Essex)
Date: December 2025
Research summary
This research explores how children’s social and emotional development in early life relates to offending during adolescence, and how school experiences such as exclusion or poor attainment may influence this. The findings have informed discussions on school behaviour and early intervention with stakeholders including the Department for Education, Welsh Government, Youth Endowment Fund, Centre for Justice Innovation, National Children’s Bureau, and British Association of Social Workers. This research has also contributed to policy debates on early risk identification and supported the award of further research funding.
This research is funded by an ADR UK Research Fellowship (ES/Z502601/1).
Data used
This study used linked, de-identified administrative data from the Ministry of Justice and Department for Education (MoJ–DfE). It draws on point-scale indicators of child development from the Early Years Foundation Stage Profile (EYFSP) for three cohorts of pupils aged 4–5 in the 2006/07, 2007/08, and 2008/09 school years. These records are linked to later educational outcomes, including absences, exclusions, and Key Stage 2 assessments, as well as cautions and convictions recorded between ages 11–17 in the Police National Computer (PNC).
Methods used
The analysis focused on identifying patterns in early childhood development and examining how these relate to adolescent offending.
Exploratory Structural Equation Modelling (ESEM) was used, a statistical technique that uncovers underlying dimensions within the Early Years Foundation Stage Profile data, distinguishing, for example, between cognitive abilities and socio-emotional skills. This approach is flexible enough to reflect how real child development often spans multiple areas, while still providing a structured framework for analysis.
The study then used mediation analysis to examine potential pathways: whether early developmental difficulties relate to adolescent offending directly, or whether they work indirectly through educational experiences such as lower attainment, school exclusions, or persistent absence. This helps clarify the mechanisms linking early childhood characteristics to later outcomes.
Research findings
Early childhood development comprises distinct cognitive and socio-emotional domains, with different predictive patterns
Factor analysis of Early Years Foundation Stage Profile data identified two developmental dimensions:
- a cognitive domain capturing numeracy, reading, and writing abilities
- a socio-emotional domain reflecting interpersonal skills, emotional regulation, and communication.
Cognitive difficulties emerge as the dominant predictor of poor academic attainment, while socio-emotional difficulties strongly predict school exclusions.
Exclusions operate as a critical pathway between early socio-emotional difficulties and adolescent offending
Children who are excluded are more than twice as likely to offend in adolescence, making exclusion the strongest school-based predictor.
Socio-emotional difficulties in childhood substantially increase exclusion risk, creating an indirect pathway to offending. However, even after accounting for exclusions, absenteeism, and academic attainment, socio-emotional difficulties retain a direct association with offending, with roughly half of their overall effect operating through school pathways and half operating directly.
In contrast, cognitive difficulties influence offending mainly through their impact on academic attainment, with no direct link once Key Stage 2 performance is considered.
Note: MoJ–DfE linked data. The figure reports marginal effects (ME × 100) from logit models predicting adolescent offending (ever cautioned or convicted between ages 11–17). Models control for demographic, socio-economic, and geographical characteristics. Confidence intervals are shown at the 5% level using school-clustered standard errors. Observation counts are rounded to the nearest 10 for confidentiality.
Socio-emotional difficulties are associated with violent and property offences, while exclusions strongly predict most offence types
Among children who offend, socio-emotional difficulties show meaningful associations with violent offences (violence against the person, sexual offences, robbery) and property offences, as well as the total number of offences committed between ages 11–17. School exclusion emerges as a strong predictor across multiple offence types and offending frequency. Higher academic attainment operates as a protective factor, particularly for less serious offences.
Note: MoJ–DfE linked data. The figure reports marginal effects from logit models predicting offence type and a negative binomial model predicting the number of offences among children who offend (ages 11–17). Models control for demographic, socio-economic, and geographical characteristics. Confidence intervals are shown at the 5% level using school-clustered standard errors. Observation counts are rounded to the nearest 10 for confidentiality.
Research impact
The research has engaged with a diverse range of stakeholders, including policymakers, practitioners in statutory and non-statutory organisations, and voluntary and community groups working on early intervention and youth justice. Findings have been presented at multiple academic conferences and policy roundtables, contributing to ongoing discussions about early risk identification and school behaviour management.
In September 2024, the research was featured at a public policy event hosted by the ESRC Research Centre on Micro-Social Change (MiSoC) titled Schooling, skills and sanctions: Research on children’s development, school behaviour policy, and crime. The event brought together representatives from the Department for Education, Welsh Government, Youth Endowment Fund, Centre for Justice Innovation, National Children’s Bureau, British Association of Social Workers, and other key organisations working in education and youth justice. The presentation examined how early socio-emotional characteristics relate to school exclusions and adolescent offending, sparking discussion on policy approaches to break the school-to-prison pipeline and evidence needs for guiding practice.
The research findings have informed applications for follow-on funding, including a successful grant (starting April 2026) to examine Restorative Relationship Practices in English schools. The work continues to contribute to policy dialogue on early intervention strategies and the role of schools in reducing youth offending risk.
Research outputs
Publications and reports
- Data Explained: Socio-emotional characteristics in early childhood and offending behaviour in adolescence, September 2025
- ADR UK Conference 2025 Proceedings: Socio-Emotional Characteristics in Early Childhood and Offending Behaviour in Adolescence, August 2025
Blogs, news posts, audio-visual
Presentations and awards
- From records to research: Harnessing administrative data to enhance lives, 2025: ADR UK Conference 2025
- Logos of crime and punishment, 2025: EUROCRIM Conference
About the ONS Secure Research Service
The ONS Secure Research Service is an accredited trusted research environment, using the Five Safes Framework to provide secure access to de-identified, unpublished data.
If you use ONS Secure Research Service data and would like to discuss writing a future case study with us, please get in touch at IDS.Impact@ons.gov.uk. Please also report any outputs here: Outputs Reporting Form