Official Statistics

The Risk of Serious Harm of the prison and probation caseload, in 2018 and 2022

Published 5 May 2023

Applies to England and Wales

1. Background and Coverage

This publication sets out statistics on offenders’ risk of serious harm to others when in the community. It compares the caseloads of what is now His Majesty’s Prison and Probation Service, as of 30 June 2018 and 30 June 2022.

2. Main Body

As part of the public protection process, offenders in custody or on the community probation caseload are assessed to determine the level of risk of serious harm they present to others. These ratings are made by practitioners, observing national risk of serious harm guidance (HM Prison and Probation Service Risk of Serious Harm Guidance 2020, March 2022 v2), and can be recorded through the Offender Assessment System (OASys), or in the caseload management system (nDelius) for those who are outside OASys coverage. While the OASys process involves making several ratings for each offender, considering the offender’s location (in custody or in the community) and different potential victim groups, an individual’s overall “risk of serious harm” is the highest of the “in the community” risks in their most recent OASys assessment^{[footnote 1]}.

It is plausible that the distribution of risk of serious harm ratings (RoSH) could change over time, due to changes in the nature of the caseload, revisions to guidance, changes in management process and/or gradual drift in norms and expectations. This statistical release compares risk of serious harm in 2018 and 2022. As well as overall distributions, it features those convicted of two frequent violent offences - inflicting grievous bodily harm without intent, and common assault and battery - and controls for whether the offender is a known perpetrator of domestic violence. Checking these subgroups provides some assurance that the overall risk of serious harm pattern is not dictated by changes in the composition of the HMPPS caseload.

Table 1 shows the distributions of risk of serious harm overall and by offence.

Between 2018 and 2022, the proportion of offenders at Low RoSH decreased, and the proportion at Medium, High and Very High RoSH increased, overall and for both featured offence types. These two violent offences became less frequent between 2018 and 2022.

Table 1: Risk of serious harm in the caseloads of 30 June 2018 and 30 June 2022, for all offenders and by offence

Offence	Year	Number of offenders [footnote 2]	Risk of Serious Harm: % of offenders [footnote 3]
			Low	Medium	High	Very High
Common assault and battery	2018	22,664	11.9	77.8	10.1	0.3
	2022	13,954	6.7	76.4	16.3	0.6
Wound or inflict grievous bodily harm without intent	2018	4,899	5.6	64.7	29.1	0.5
	2022	4,086	3.1	57.2	38.3	1.4
All other offences	2018	230,312	30.0	48.7	20.7	0.7
	2022	226,638	21.5	52.2	25.3	1.0
All offences	2018	257,875	27.9	51.5	19.9	0.6
	2022	244,678	20.4	53.7	25.0	0.9

Table 2 shows the distributions of risk of serious harm overall and by domestic violence status. The overall recorded prevalence of domestic violence perpetration increased slightly, from 39% in 2018 to 42% in 2022.

An increasing proportion of domestic violence perpetrators were rated as High RoSH, rising from 32% to 36% of perpetrators between 2018 and 2022. This was a greater increase than for nonperpetrators, where the proportion rose from 21% to 23%. Similarly, the increase in the proportion rated Very High RoSH was greater for perpetrators (1.1% to 1.5%) than nonperpetrators (0.6% to 0.7%).

Table 2: Risk of serious harm in the caseloads of 30 June 2018 and 30 June 2022, for all offenders and by domestic violence perpetration status

Domestic violence perpetration [footnote 4]	Year	Number of offenders [footnote 2]	Risk of Serious Harm: % of offenders [footnote 3]
			Low	Medium	High	Very High
Non-perpetrator	2018	114,208	27.7	50.5	21.2	0.6
	2022	119,119	22.6	53.8	22.9	0.7
Perpetrator	2018	71,900	5.2	61.7	32.0	1.1
	2022	87,327	3.4	58.9	36.2	1.5
Not known	2018	71,767	51.0	43.0	5.7	0.2
	2022	38,232	52.4	41.4	5.8	0.4
All offences	2018	257,875	27.9	51.5	19.9	0.6
	2022	244,678	20.4	53.7	25.0	0.9

3. Contact information

Press enquiries should be directed to the Ministry of Justice press office:
Tel: 020 3334 3536
Email: newsdesk@justice.gov.uk

Other enquiries about these statistics should be directed to Data and Analysis at the Ministry of Justice:

Philip Howard
Head of Risk Assessment Data Science, Data and Analysis
Ministry of Justice
10th Floor
102 Petty France
London
SW1H 9AJ

E-mail: statistics.enquiries@justice.gsi.gov.uk

General enquiries about the statistical work of the Ministry of Justice can be emailed to: statistics.enquiries@justice.gsi.gov.uk

General information about the official statistics system of the UK is available from: UK Statistical System – UK Statistics Authority

© Crown copyright

Produced by the Ministry of Justice

Alternative formats are available on request from statistics.enquiries@justice.gsi.gov.uk

1. When an offender has been assessed on OASys during their current sentence and/or during the two years prior to the date of interest, the most recent such assessment is selected; from this assessment, the highest of four community risk of serious harm ratings is selected. If they had no such OASys assessment, the risk of serious harm entry on their current nDelius record is selected. The four OASys ratings represent risk to known adults, the general public, children and staff. (For example, an offender who is rated as presenting High risk to staff, Medium risk to the general public and Low risk to known adults and children has been rated as presenting a High risk of serious harm.) The nDelius rating represents risk to any and all potential victims. ↩

2. Offenders with no risk of serious harm rating, and/or for whom matching between data sources was not complete, are excluded. These numbered 14,183 in 2018, and 14,624 in 2022. ↩ ↩²

3. Percentages may not add to 100 due to rounding. ↩ ↩²

4. OASys data on domestic violence perpetration does not distinguish between violence against partners and against other family/household members. This data is only available when a full (“Layer Three”) OASys assessment has been completed; operational instructions on OASys completion mean that Layer Three assessments are more likely for higher-risk offenders. OASys was not used by some Community Rehabilitation Companies (CRCs), which operated until the reunification of probation services on 26 June 2021. Supervision by CRCs was principally subject to three conditions: these offenders had not been rated as High or Very High Risk of Serious Harm, did not have high actuarial Risk of Serious Recidivism, were not subject to Multi Agency Public Protection Arrangements. ↩