Working with offenders
The impact of corrections on re-offending
The Home Office commissioned this report to review and update knowledge of 'what works' in corrections to reduce re-offending. It focuses on correctional services and interventions with adult offenders, which aim to reduce re-offending. It builds on a review, which assessed the evidence available in the mid-1990s on ways to reduce offending (Goldblatt & Lewis, 1998).
Title: The impact of corrections on re-offending
Author: Gemma Harper and Chloë Chitty
Series: Home Office Research Study 291
Date published: December 2004
Number of pages: 128
Availability: Download full report 
PDF 475 Kb
Developments in sentencing
Sentencing has become more severe over the past decade. This is most likely due to:
a perception among sentencers that cases have become more serious
changes in sentencing policy
demands from the media, public and politicians for tough punishment.
Courts are increasingly using custody and longer sentence lengths and as a result the prison population has continued to rise since the early 1990s and now stands at over 75,000. Community sentences are also being used more, at the expense of fines.
The National Offender Management Service (NOMS) is trying to reverse this sentencing trend, encourage fines, and improve the credibility of community sentences. The new sentencing provisions of the Criminal Justice Act 2003 introduce significant changes to the custodial and community sentences and should help NOMS achieve this aim.
Factors associated with offending
To reduce re-offending, programmes and initiatives must target those risk factors (or problems) that can influence criminal activity. For example:
employment
education and training
substance mis-use
money
accommodation.
Recent Offender Assessment System (OASys) data suggests that offenders have on average 4 problems which contribute to them committing a crime. Therefore, the most effective interventions will tackle a range of problems.
Evidence does not prove that poor basic skills cause offending, but they do have an effect on unemployment which is related to offending. However, it is unlikely that unemployment is a direct cause of crime because simply having a job does not stop a person offending. It depends on other social and relating factors such as job stability and satisfaction.
Research also suggests that:
stable accommodation both directly and indirectly reduces re-offending
substance mis-use is associated with offending and can have wider implications for offenders, for example, on their job prospects
maintaining good family relationships are important when resettling prisoners and can reduce re-offending
on average girls and women offend less frequently than boys and men, start offending later and generally stop earlier. When women do offend they tend to commit acquisitive rather than violent crime.
Treating offenders
The most effective way to treat offenders is to combine interventions and initiatives. This research identified several management principles that encouraged the offender to participate. These include:
acknowledging offenders' experiences and needs
providing continuity of contact
openness, flexibility and support
'What works' principles
Programmes and interventions are more likely to reduce re-offending if:
they are based on an explicit model of the causes of crime
they have a risk classification
they target offenders needs and problems that led to the offending
they are responsive
the treatment method is based on cognitive- behavioural techniques
they have programme integrity.
Offending behaviour programmes in prison and probation
International evidence suggests cognitive-behavioural offending programmes and interventions are the best for working with offenders.
Findings from some offending behaviour programmes in prisons are below:
Accredited Enhanced Thinking Skills (ETS) and Reasoning & Rehabilitation (R&R) - No differences in 2-year reconviction rates between adult male prisoners who participated in cognitive skills programmes and those who had not.
Pre-accredited ETS and R&R - reconviction rates were 14 percentage points lower for medium-low risk and 11 percentage points for medium-high risk adult male offenders.
ETS as part of intensive regime for Young Offenders - reconviction rates were 10 percentage points lower after one year but not after two.
Findings from offending behaviour programmes for those on probation are as follows:
Pre-accredited Sex Offender Treatment (SOTP) in prisons - reconviction rates fell by 3.5 percentage points over 2 years for those who were convicted of sexual or violent offences.
Pre-accredited SOTP for those on probation - lower reconviction rates of 8.1 percentage points for child sex abusers.
Two programmes for domestic violence - the frequency of violence was significantly lower 12 months later.
Pre-accredited Think First Probation - significantly higher 2-year reconviction rate of 24 percentage points for adult males on community-based programme compared to adult males sentenced to custody without programme.
Aggression Replacement Training (ART) - reconviction rates were 9.2 percentage points lower compared to those that didn't take part.
Implementing these programmes
Three main problems were identified in delivering these interventions effectively. These included
the rapid expansion of programmes
targeting programmes ineffectively
higher than expected number of people dropping out.
It is therefore important to monitor the programmes quality and its impact at all times.
Alternative approaches to integrate offenders into the community
Participating in programmes while in custody is only part of the rehabilitation process. Prisons need to provide aftercare after release to help offenders gain stable accommodation and employment to reduce the likelihood of re-offending.
Some intervention programmes to reintegrate offenders back into the community include:
Amity Therapeutic Community - re-imprisonment rates were lower over a 5-year period for those taking part.
Post-Release Employment Project - those who worked in prison were less likely to be arrested 12 months after release.
Last update: Wednesday, September 17, 2008