Last updated 9 March 2020 - see all updates
1. Main facts and figures
- Black offenders had the highest rates of reoffending in 2016/17, at 33.3%
- for the same period, offenders in the Other ethnic group had the lowest rate of reoffending, at 21.3%
- across all ethnic groups, juveniles had a higher rate of reoffending than adults
- in 2016/17, White offenders had the highest average number of reoffences per reoffender at 4.12 reoffences
The ethnic categories used in this data
Ethnicity is decided by the police based on the offender’s appearance and was not confirmed by the offender or the Court. The data is broken down into the following broad groups:
- Other ethnic group
2. By ethnicity
|Ethnicity||%||Number of reoffenders||Average number of reoffences per reoffender|
|Other inc Mixed||21.3||1,136||3.46|
Summary of Proven reoffending By ethnicity Summary
3. By ethnicity and age group
|Ethnicity||Juvenile %||Juvenile Number of reoffenders||Juvenile Average number of reoffences||Adult %||Adult Number of reoffenders||Adult Average number of reoffences|
|Other inc Mixed||39.5||119||3.75||20.2||1,017||3.42|
Summary of Proven reoffending By ethnicity and age group Summary
4. By ethnicity over time
|Asian||Black||White||Other inc Mixed||Unknown|
|Year||Asian %||Asian Number of reoffenders||Black %||Black Number of reoffenders||White %||White Number of reoffenders||Other inc Mixed %||Other inc Mixed Number of reoffenders||Unknown %||Unknown Number of reoffenders|
Summary of Proven reoffending By ethnicity over time Summary
5. Reoffences per reoffender by ethnicity
|Year||Asian||Black||White||Other inc Mixed||Unknown|
Summary of Proven reoffending Reoffences per reoffender by ethnicity Summary
A proven reoffence is defined as any offence committed within 12 months following release from a custodial sentence, a non-custodial conviction or a caution that leads to a subsequent court conviction, caution, reprimand or warning within those 12 months (or within 18 months to allow the offence to be proven in court).
Proven reoffending is measured using a number of data sources including an extract of the Police National Computer (PNC) held by the Ministry of Justice (MOJ).
The process involves matching offender details from the prison and probation data to the personal details recorded on the PNC. Administrative datasets of this size tend to be very high quality, however there may still be inaccuracies due to errors when collecting the data. Some offenders who could not be matched to the PNC have been excluded from the figures.
Proven reoffending statistics are produced using 3-monthly cohorts. This means that data reports on offenders who are released from custody (or received a non-custodial conviction or caution) in a 3-month period and who subsequently went on to reoffend within 12 months. Annual figures are calculated by taking an average of the four preceding three-month offender cohorts and so this means that a reoffender may appear more than once in the annual average figures.
Due to a methodology and data source change in October 2015, users should be careful when using the April 2015 to March 2016 cohort and any subsequent annual figures to compare to the April 2014 to March 2015 cohort and earlier years.
Further information about the change of data source and methodology can be found in MOJ’s publication How the measure of proven reoffending has changed and the effect of these changes (PDF).
The estimates in this measure can be used to calculate a relative rate index of the proportion of offenders who reoffend, as published in the Lammy review. The index provides an estimate of how more likely or less likely offenders from a particular ethnic group are to reoffend, relative to white offenders. To calculate the relative rate index, divide the proportion of offenders who reoffend for the ethnic group of interest by the proportion of offenders who reoffend for the White group.
Suppression rules and disclosure control
Proportions and averages based on fewer than 30 offenders are removed as they make estimates unreliable. Number of reoffenders, reoffences and previous offences based on fewer than five offenders are suppressed. This is to prevent the disclosure of individual information.
Percentages are rounded to 1 decimal place. Reoffences per reoffender are rounded to 2 decimal places.
Further technical information
7. Data sources
Type of data
Type of statistic
Ministry of Justice
Purpose of data source
Reoffending data is used by the government to aid in the development of policy and the delivery of services to reduce re-offending.
8. Download the data
This file contains the following: measure, ethnicity, year, age group, number of offenders, number of reoffenders, number of reoffences, average number of reoffences per reoffender