Analysis of children’s educational, children’s social care, and offending characteristics

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.

Authors: Department for Education and Ministry of Justice

Date: March 2022

Research summary

This project looked at the social care, educational and demographic characteristics of 1.63 million children in England, as well as their interactions with the criminal justice system. Findings show that children cautioned or sentenced for an offence are over-represented by those with certain characteristics, including being known to children’s social care and having special educational needs. The findings have been shared across government.

The research was carried out by the Ministry of Justice (MoJ) and the Department for Education (DfE), as part of the MoJ’s ‘Data First’ innovative data linking programme, funded by ADR UK.

Results were featured in the Special Educational Needs and Disabilities and Alternative Provision Green Paper and the Independent Review of Children’s Social Care. This case study focuses on children’s social care and special educational needs in view of the Special Educational Needs and Disabilities and Alternative Provision Improvement Plan and the Children’s Social Care Implementation Strategy.

Data used

The linked dataset for this project was accessed on the MoJ and DfE’s secure servers, but is also available in the ONS Secure Research Service. Specific component datasets used were:

• School Census Pupil Level
• Pupil Referral Unit Census
• Alternative Provision Census
• Key stage 2 attainment data
• Key stage 4 attainment data
• Pupil absence
• Exclusions
• Children in Need Census
• Children Looked After Census
• Police National Computer.

Before the MoJ & DfE linked dataset was created, analysis on the educational background of children who have offended was not possible as there was no unique identifier across the departments. The two departments worked together to agree match rules, data cleaning rules, and what data would be most beneficial to accredited researchers. A unique person identifier was created to identify the same children across the various justice, social care, and education datasets. The final data are de-identified.

Methods used

The team looked at the demographic, education and children’s social care characteristics of 1.63 million children who completed key stage 2 (ages 7 to 11) in England between 2007-08 and 2009-10. This was linked to data on their interactions with the criminal justice system up to 2017. For this cohort, all records from year 1 to year 13 (or equivalent) were included. They excluded pupils who did not have a key stage 4 record or who attended an independent primary or secondary school.

Using the 2011-12 Children in Need Census data, children were included in the analysis if they had been recorded as a ‘child in need’ between the ages of 12-16. It is important to note that a number of local authorities had issues providing complete data in the first three years of the Children in Need Census, and so the 2011-12 data needs to be interpreted with caution. Using the 2005-6 Children Looked After Census data, available from 2005/06, children were included in the analysis if they had been recorded as a ‘child looked after’ between the ages of 6-16.

The analysis focused on descriptive statistics - methods which summarise the characteristics of a dataset.  This included longitudinal analysis of the timing of justice and education events. It is important to note that the analysis does not imply there is a causal link between the educational outcomes, characteristics and offending.

Research findings

Overall, the research looked at the relationship between serious violence offences, and demographic and education factors such as gender, attendance, suspensions, and exclusions. However, this particular case study focuses on the findings around children’s social care and special educational needs.

Children in need (CIN) and Children looked after (CLA): CIN here refers to children who are designated under a number of different social care classifications, including:

• children on a child in need plan
• children on a child protection plan
• children who are looked after (CLA).

UK official definitions for CIN and CLA can be found here (page 66). 

In this dataset, 32% of children who had been cautioned or sentenced for an offence had been CIN on 31 March in any given year between the ages of 12 and 16. 11% of children who had been cautioned or sentenced for an offence had been CLA on 31 March in any given year between the ages of 6 and 16. This compares to 6% and 1% of the all-pupil population.

The figure below shows that 38% of children who had been cautioned or sentenced for a serious violence offence had been CIN between the ages or 12 and 16. However, 6% of those who had been CIN between the ages or 12 and 16 had also been cautioned or sentenced for a serious violence offence. The figures for CLA were 14% and 10% respectively. This shows that it is not the case that most children in the CIN or CLA categories are also cautioned for a serious violence offence.

Special Educational Needs (SEN): High proportions of the offending groups were recorded as having had SEN, specifically:

• 80% of those who had been cautioned or sentenced for an offence
• 87% of those cautioned or sentenced for a serious violence offence.

In comparison, 45% of the all-pupil population were recorded as having SEN Special Educational Needs up to the end of key stage 4.

More detailed findings can be found in the published reports.

Research impact

Since March 2022 when the research was published, the findings have been shared across government. The report had wide policy impact, with presentations to the Youth Endowment Fund (the What Works Centre for Serious Youth Violence) and the Youth Justice Board.

There are two main targeted policy programmes focusing on reducing serious violence through the education system. These have been using the findings to support decision making across their project strands. The research, and the local data dashboard published in March 2023, have also played an important role in the implementation of the Serious Violence Duty from January 2023. The findings have been available to inform the strategic needs assessments carried out by local areas in the first part of 2023.

Findings have also featured in the Special Educational Needs and Disabilities and Alternative Provision Green Paper and the Independent Review of Children’s Social Care.

ADR UK has funded five Research Fellowships to analyse the Ministry of Justice & Department for Education linked dataset. Some of these focus on understanding links between childhood characteristics, educational outcomes, and offending.

Research outputs

Publications and reports

• Department for Education and Ministry of Justice descriptive statistics, March 2022: Education, children’s social care and offending
• Department for Education research report, July 2023: Education, children’s social care and offending: multi-level modelling
• Department for Education and Ministry of Justice, further publications: Education, children’s social care and offending
• Ministry of Justice – Department for Education linked dataset: Feasibility of evaluating early interventions for violence prevention: Data quality report

Presentations and awards

• British Society of Criminology Conference, June 2022
• Royal Statistical Society Conference, September 2022
• DataConnect22 Conference, September 2022
• Data First quarterly academic seminar, April 2023
• The Stockholm Criminology Symposium, June 2023
• Data Connect 23: Harnessing the potential of linked administrative data

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 ensure you have reported your outputs here: Outputs Reporting Form.