Outcome evidence on offender rehabilitation: the role of probation programmes

Similar documents
Corrections, Public Safety and Policing

Young people in custody learning thinking skills: Experiences; Skills and Developments Directorate of Public Sector Prisons

Research on transition management: What works in re-entry?

Best Practices for Effective Correctional Programs

TABLE OF CONTENT INTRODUCTION, HISTORIC OVERVIEW, NATIONAL AND INTERNATIONAL RESEARCH ON OFFENDER NEEDS AND RISK ASSESSMENT

New Me Coping UK. Type of intervention. Target group/s, level/s of prevention and sub-group/s: Target population. Delivery organisation

Impact evaluation of the prison-based Core Sex Offender Treatment Programme: a success story. Laura Di Bella, Mark Purver, and Aidan Mews

Utah Cost of Crime. Cognitive-Behavioral Therapy (Juveniles): Technical Report. December 2012

Risk assessment principle and Risk management

The economic case for and against prison

Getting To Desired Outcomes:

EPICS. Effective Practices in Community Supervision. Brought to you by the Multco. EPICS Training team

Maximizing the Impact of Interventions for Youth: The Importance of Risk/Needs Assessment

Use of Structured Risk/Need Assessments to Improve Outcomes for Juvenile Offenders

Juvenile Justice Vision 20/20 Fall Conference November 13, 2014 Grand Valley State University

I understand that the Royal Commission is particularly interested in:

GOVERNMENT OF BERMUDA Ministry of Culture and Social Rehabilitation THE BERMUDA DRUG TREATMENT COURT PROGRAMME

Over the last several years, the importance of the risk principle has been

TURNING POINT ASSESSMENT/TREATMENT WOMAN ABUSE PROTOCOL DEPARTMENT OF JUSTICE AND PUBLIC SAFETY

Assessing ACE: The Probation Board s Use of Risk Assessment Tools to Reduce Reoffending

PRISON MENTAL HEALTH TRAINING WORKSHOP JUNE Social problem-solving PENAL REFORM INTERNATIONAL. University of Liverpool. James McGuire.

NATIONAL INSTITUTE FOR HEALTH AND CLINICAL EXCELLENCE SCOPE

ACCREDITED PROGRAMMES AND INTERVENTIONS

Recent thinking and results from OASys

2016 Annual Meeting Conference

Interventions. I think it s time I started to make something of my life now and I m fed up [with] hurting people around me, it s time to stop it.

GOOD PRACTICE GUIDELINES Training in Forensic Clinical Psychology

Psychometric qualities of the Dutch Risk Assessment Scales (RISc)

JUSTICE REINVESTMENT: FOUNDATIONAL REQUIREMENTS FOR EFFECTIVE COMMUNITY-CENTERED OFFENDER REHABILITATION. Hon. Frank L. Racek

Civil Commitment: If It Is Used, It Should Be Only One Element of a Comprehensive Approach for the Management of Individuals Who Have Sexually Abused

For its rehabilitative programmes, the Department of Corrections has adopted the

AN OVERVIEW OF WHAT WORKS IN CORRECTIONAL INTERVENTIONS

Becoming New Me UK. Type of intervention. Target group, level of prevention and sub-groups: Target population. Delivery organisation

MSc Forensic Psychology. Joining Instructions 2018/2019

Interventions for High Risk Sexual Offenders

Recognising Dangerousness Thames Valley Partnership.

Dealing with Feelings: The Effectiveness of Cognitive Behavioural Group Treatment for Women in Secure Settings

Integrating Evidence via Systematic Reviews

Understanding the Risk Need, Responsivity (RNR) model and crime desistance perspective and integrating them into correctional practice

First Do No Harm. Avoiding adverse outcomes in personality disorder treatments. Mary McMurran PhD.

GUIDE TO INTERVENTIONS

The Technical Background of the Risk, Need, Responsivity (RNR) Simulation Tool

Targeting and Delivering Offender Management in Custody. Practice Guidance for Offender Supervisors

The Violence Against Women and Domestic Abuse

Domestic Violence Inventory (DVI) Reliability and Validity Study Risk & Needs Assessment, Inc.

THE CASE OF NORWAY: A RELAPSE

Clinical guideline Published: 28 January 2009 nice.org.uk/guidance/cg77

OFFENDING BEHAVIOUR PROGRAMMES: HISTORY AND DEVELOPMENT

ELIZABETH K. DRAKE, PH.D. CANDIDATE

What works in substance misuse treatments for offenders?

RESEARCH & THEORY ON FAMILY VIOLENCE Chapter 3 DR GINNA BABCOCK

Maximizing the Impact of Juvenile Justice Interventions: The Importance of Risk/Needs Assessment

The positive effects of cognitive behavioral programs for offenders: A meta-analysis of factors associated with effective treatment

Christina M BSC (Hons.), MSC., CPsychol., AFBPsS

Barnsley Youth Justice Plan 2017/18. Introduction

A Foundation for Evidence-Based Justice Decisions

Working together to reduce reoffending. BeNCH CRC PROSPECTUS. A leading provider of innovative justice services that change people s lives

Contents Opioid Treatment Program Core Program Standards... 2

SOCIAL WORK COURSE DESCRIPTIONS

Risk-Need-Responsivity: Managing Risk & Mental Health For Juvenile Justice-Involved Youth

probation, number of parole revocations, DVI Alcohol Scale scores, DVI Control Scale scores, and DVI Stress Coping Abilities Scale scores.

Nanaimo Correctional Centre Therapeutic Community

Research Summary 7/09

MORE TREATMENT, BETTER TREATMENT AND THE RIGHT TREATMENT

The Offender Assessment System (OASys): Development, validation and use in practice

Berks County Treatment Courts

ACE! The Risk-Need- Responsivity Simulation Tool. The Center for Advancing Correctional Excellence. Solutions For Justice Professionals.

CONTENT OUTLINES AND KSAS

A Risk Assessment and Risk Management Approach to Sexual Offending for the Probation Service

Prison Population Reduction Strategies Through the Use of Offender Assessment: A Path Toward Enhanced Public Safety

The Risk-Need-Responsivity Model of Assessing Justice Involved Clients. Roberta C. Churchill M.A., LMHC ACJS, Inc.

CONTENT OUTLINES AND KSAS

270 COLLEGES AND SCHOOLS. SS 430 High School Teaching Methods (2). See ECI 430. SS 702 Seminar: Social Science Teaching Methodologies (3).

Programme Specification. MSc/PGDip Forensic and Legal Psychology

Evidence-based interventions in forensic mental health and correctional settings

Mental Health Treatment Requirement Denise Butt

Program Accreditation and Correctional Treatment

Reducing Prisoner Reoffending

Two, contrasting, models of offender rehabilitation evident, each with distinct normative and etiological assumptions:

CRIMINAL JUSTICE (CRIMLJUS)

Work Package 1. Mapping the evidence base: a descriptive analysis of the WP1 Systematic Review Database. Authors: Kate Bowers and Lisa Tompson

Report-back on the Alcohol and Other Drug Treatment Court Pilot and other AOD-related Initiatives

Evidence-Based Sentencing to Reduce Recidivism

Violence Prevention A Strategy for Reducing Health Inequalities

Criminal & Addictive Thinking Part I

2.2 Development/Collation of a suite of tools for use when working with young people on a 1:1 basis.

Effectiveness of Interventions with Adult Male Violent Offenders

Measuring Attitudinal Change: An Action Research Project

What works in policing domestic abuse?

POST-SENTENCE INITIATIVES FOR SEX OFFENDERS IN THE COMMUNITY: A PSYCHOLOGIST S PERSPECTIVE

Community-based sanctions

UNIVERSITY OF MALTA FACULTY FOR SOCIAL WELLBEING DEPARTMENT OF CRIMINOLOGY LECTURE PROGRAMME CRI1017 CRIME SCENE TO COURT

Justice Data Lab Re offending Analysis: Prisoners Education Trust

REGINA AUTO THEFT STRATEGY: PROCESS EVALUATION

A Quasi Experimental Evaluation of Thinking for a Change: A Real-World" Application

Assessing the effectiveness of the correctional sex offender treatment program

Ask the Doctor Webinar Series:

The Standardized Program Evaluation Protocol (SPEP): Using Meta-analytic Evidence to Assess Program Effectiveness

SAQ-Adult Probation III: Normative Study

Transcription:

Fundación Paz Ciudadana Seminar, Santiago de Chile October 2010 Outcome evidence on offender rehabilitation: the role of probation programmes James McGuire University of Liverpool, UK merc@liv.ac.uk 1

Objectives To present a general overview of the current evidence base for reducing offender recidivism: Large scale outcome research/meta analytic review Some practical examples To describe recent practice in probation services in England and Wales: The range of programmes used and theoretical models Outcome effects and implications 2

Where does crime come from? Levels of explanation in criminological theory Level: Focus: Objective: Illustrative theories: 1 Society To explain crime as a 2 Local areas, communities large scale social phenomenon To account for geographical variations in crime Conflict theory, Strain theory Sociological control theories Feminist theory Ecological theories Differential opportunity theory 3 Proximate To understand the roles Sub cultural delinquency theory Differential association theory social groups of socialisation and group influence 4 Individual To analyse and account criminal acts 5 Individual offenders for patterns and types of crime events To examine intraindividual factors e.g. thoughts, feelings and behaviour Routine activity theory Rational choice theory Neutralization theory Psychological control theory Cognitive social learning theory Mental disorder 3

Longitudinal studies of human development elopment Conducted in: Finland, Sweden, Denmark, UK, USA, Canada, New Zealand Cohorts followed for periods up to 40 years Large samples, wide range of data collected Retrospective versus prospective designs Multiple l l sources of evidence Accelerated / parallel cohorts in some studies Construct models of change over time 4

How far can the criminal i justice system affect the crime rate? Numbers of offences Numbers of individuals 5

Risk factors: variables recurrently associated with serious and/or persistent offending Poor parental supervision, low attachment to families Difficulties i in school and employment Network of delinquent associates Manifestation of anti social attitudes Distorted / biased patterns of information processing Poor personal and social skills Low levels of self control; impulsiveness Negative emotions and low level of constraint 6

Approaches to crime reduction Primary prevention situational prevention (preventing crime at places), policing long term developmental prevention, system level investment Secondary prevention work with at risk groups Tertiary y prevention work with adjudicated offenders, criminal justice/penal system 7

The evidence base on reducing recidivism There are many hundreds of basic studies evaluating interventions to reduce criminal recidivism There were 75 meta analyticanalytic reviews as of mid 2008 Positive (and significant, though modest) mean effect size Some interventions have negative effects (...they make people worse!) There is a consistent patterns in combination of features that can increase effect sizes considerably These can also result in significant cost savings The principles of effective intervention are well understood (e.g. the Risk Need Responsivity model) 8

The Risk Need Responsivity (RNR) model (Andrews, Bonta & Wormith 2006) Identify general risk factors for offending At an individual level those most reliably supported include: criminal associates low level of self control / self management poor self regulation of negative emotion limitations in cognitive or social skills anti social attitudes Use offence analysis and case formulation to identify combination for an individual Deliver interventions taking account of contextual variations and situational ti requirements 9

Meta analysis of interventions for violence (Dowden & Andrews, 2000) 34 studies, 52 effect sizes 70% of studies with adult samples Mean effect size: +0.07 Range from 0.22 to +0.63 Effect size for punitive sanctions: 0.01 Effect size for human service principles : +0.12 Correlation of 0.69 between the number of criminogenic needs targeted and outcome effect size (p<.001) 10

Meta analysis analysis of interventions to reduce sexual offending (Lösel & Schmucker, 2005) 69 studies Total sample n = 22,181 80 tests Randomization: 6; statistical controls: 7 Mean effect sizes (odds ratios): For sexual recidivism: +1.70 (37% reduction) For violent recidivism: i +1.90 (44% reduction) For general recidivism: +1.67 (31% reduction) 11

Meta analysis of interventions for serious violent and sexual offending by young people: EFFECT SIZES FOR SPECIFIC INTERVENTIONS (Lipsey & Wilson 1998; Garrido & Morales 2007) COMMUNITY (k=117) INSTITUTION (k=83) Intervention Effect size Intervention Effect size Positive effects, consistent Individual counselling 0.46 Interpersonal skills 0.39 Interpersonal skills training 0.44 Teaching family home 0.34 Behavioural programmes 042 0.42 Positive effects, less consistent Multiple services 0.29 Behavioural programmes 0.33 Restitution, probation/ 0.15 Community residential 0.28 parole Multiple utpeservices 0.20 0 12

Is offender treatment of practical significance? Comparative effects sizes for selected interventions Intervention Treatment target Effect size Aspirin Risk of myocardial infarction 0.034 Chemotherapy Breast cancer 0.08 0.11 Bypass surgery Coronary heart disease 0.15 AZT HIV/AIDS 0.23 Psychological therapy Mental health problems 0.32 Treatment of offenders Recidivism: overall effect 0.10 Recidivism: appropriate service 0.29 13

Guidelines for effective programmes Adapted from Andrews, Gendreau, Lipsey, Hollin, Lösel, and others Employ a sound theoretical base using evidence from psychology and criminology Use a personality and social learning approach Avoid retribution and deterrence strategies Focus on dynamic risk factors & targeted interventions Use evidence based d risk assessment and allocation Use best validated methods Use community based services where possible Use methods that activate behavioural and cognitive change

Guidelines for effective programmes (...continued) Use multi modal approaches focusing on a range of needs General responsivity: learning styles and high quality relationships Specific responsivity: adapt interventions to groups Identify areas for use of personal discretion by staff Monitor programme and treatment integrity Develop an agency plan Develop staff skills Develop management framework Establish wider community and societal links

Variations in effect sizes (McGuire, 2004) The figure shows effect sizes for selected types of interventions from meta analytic reviews, to illustrate the range of findings obtained. In the graph, the vertical line represents zero change (no difference between experimental and control samples). Bars extending to the right represent increases in recidivism amongst experimental samples relative to controls. Bars to the left represent reductions in recidivism. 16

Effectiveness of cognitive behavioural programmes (Lipsey et al, 2007) Review of 58 studies published between 1980 2004. Majority quasi experimental designs: only 33% randomized Average follow up interval of 12 months. Mean Odds Ratio = 1.53 This corresponds to a 25% reduction in recidivism. Significant heterogeneity: (Q) = 214.02. No significant differences found between randomized and nonrandomized designs. Most prominent moderators = risk level of the participants, and the quality of implementation of programme. Mean Odds Ratio for studies with best practice features = 2.86. This corresponds to a 52% decrease in recidivism. 17

Cost benefit analysis of programmes The graph shows the ratio of benefits (money saved) relative to costs (money invested) for a number of intervention programmes. Aos, S., Phipps, P., Barnoski, R. & Lieb, R. (2001). The Comparative Costs and Benefits of Programs to Reduce Crime. Olympia, WA: Washington State Institute for Public Policy. 18

The evolution of programmes: Local and national projects in the UK Experimental Day Training Centres (1970s) as alternatives to prison Other staff led initiatives in local areas Elements of programmes gradually introduced (1970s 1980s) 1980s) Increasing emphasis on offence focused work: 1985 The Mid Glamorgan Straight Thinking on Probation (STOP) experiment (1991) Prison based sex offender programme (SOTP) introduced (1992) Partial dissemination of positive results through conferences ( e.g. What Works ) (1992, 1994, 1996, 1998) Prison based General Accreditation Panel established in 1996 Government commissioned reviews of research: 1997, 1998 Home Office Crime Reduction Programme: 2000 onwards Establishment of fcorrectional lservices Accreditation i Panel 19

UK Crime Reduction Programme o A major policy initiative launched in 1998, that became operational in 2000 o Considerable resources provided o Investment in dissemination of manualised programmes o Effort to preserve integrity of delivery Staff training Accreditation and monitoring systems o Research and evaluation component 20

Programme accreditation criteria used by the Correctional Services Accreditation Panel A clear model of change Selection of offenders Targeting a range of dynamic risk factors Effective methods Skills orientated Sequencing, intensity and duration Engagement and motivation Continuity of Programmes and Services Process Evaluation and Maintaining Integrity Ongoing Evaluation Each item is scored as follows: 2 = fully met 1 = partially met 0 = not met A programme must score between 18 20 points to be awarded accredited status 21

Programmes currently accredited for use in the community (Probation) o o o o o o o o o o o Enhanced Thinking Skills Think First Reasoning and Rehabilitation Priestley One to One Cognitive Skills Booster Thinking Skills Programme Women s Acquisitive Crime Programme ASRO (Addressing Substance Related Offending) PRISM (Programme for Reducing Individual Substance Misuse) Drink Impaired Drivers Offender Substance Abuse Programme o o COVAID (Control of Violence for Angry Impulsive Drinkers) Lower Intensity Alcohol Programme o Sex Offender Group Programmes (4 versions) o o o o o o Internet Sexual Offending Treatment Programme Aggression Replacement Training Controlling Anger and Learning how to manage it (CALM) Cognitive self change programme Integrated Domestic Abuse Programme Community Domestic Violence Programme 22

Programme documentation: Each programme requires five manuals Theory manual Programme manual Assessment and Evaluation manual Management or Operating manual Staff Training manual 23

Cognitive behavioural interventions A continuum of theory and methods Behaviourally-oriented Behaviour modification Behaviour therapy Social skills training Self-instructional training Problem-solving training Cognitive therapy Cognitively-orientedoriented Schema-focused therapy

Theoretical structure of the Think First Programme Programme elements and their conceptual inter linkages Social problemsolving training Self management & self control Social interaction i skills Attitudes & values

Relationship of programme components to criminogenic risk factors How programme elements target selected risk factors Cognitive limitations, e.g. Avoiding problems, poor skills in generating possible solutions, anticipating consequences, formulating steps; rigid thinking style Self management problems, e.g. Poor anger or impulse control, substance dependence, mood swings, low motivation, Social cognition, e.g. Failure to consider others, attributional biases, cognitive distortions, self efficacy expectations Attitudes and values, eg e.g. Antisocial beliefs, use of neutralizations, biased beliefs (gender, race); hostility; exploitativeness Social interaction, e.g. Social isolation, poor interpersonal skills, dominance/submissiveness, status concerns, susceptibility to peer pressure 26

Social problem solving training The problem solving sequence Recognise problems Put problem into words Make clear and specific Describe as outcome Generate alternative solutions Formulate means-ends steps Anticipate consequences Make decision Implement and monitor

The Pathfinder Evaluation Project Leicester and Liverpool Universities Process evaluation Retrospective evaluation Prospective evaluation Cost benefit evaluation (with CRG Associates, Cardiff) Study supported by the Home Office Research, Development and Statistics Directorate 28

Programmes and sample sizes Reasoning and Rehabilitation Rhbili i Retrospective sample: 2,141 (R&R) 36/37 two hour group sessions Prospective sample: 2,409 Enhanced Thinking Skills (ETS) 20 two and a half hour group sessions Comparison samples: 2,630/3,305 Think First 4 individual sessions 22 two hour group sessions 6 individual post group sessions Other programmes: Addressing Substance Related Offending (ASRO) One to One Aggression Replacement Training (ART) Drink Impaired Drivers (DIDs) Data collected from 26 probation areas Cross checked with Offenders Index and Probation Index Analysed using descriptive, nonparametric and multivariate statistics 29

Intention to treat to treat (ITT) analysis Experimental group 69.9% reconvicted Comparison group 57.9% reconvicted Significant difference in favour of the comparison group! 30

Percent reconviction rates from a prospective evaluation of general offending behaviour programmes (Hollin et al, 2007) Experimental group = 2,409; Comparison group = 3,305 SUB-GROUP Reasoning and Rehabilitation ti (R&R) Enhanced Thinking Skills (ETS) Think First Completers 34.96 42.64 37.41 Non-starters t 80.8686 80.0404 74.22 Non-completers 81.48 75.81 71.63 Comparison 63.66 63.66 63.66 Regression model χ 2 38.37 p < 0.001 73.36 p < 0.001 80.55 p < 0.001 31

Survival curves for probation samples (n=929) 32

Are completers those who would do well anyway? Appropriateness of assessed risk level (OGRS 2) Effect size are higher for those in specified risk band than for those in too high and too low bands Probation officers ratings of offender motivation No detectable association between assessments at referral stage and subsequent outcomes (completion Referral numbers Programme referrals for Think First 2001-2004 16000 14000 12000 10000 8000 6000 4000 2000 0 versus non completion) 6000 Number of motivated offenders appears to rise on an annual basis Statistical model predicting completion using propensity scores does not predict re offending 1 2 3 Year Completers Commenced but did not complete Referred but did not commence 33

Propensity score stratification and reconviction outcomes (McGuire et al, 2008) Quintile Completer Comparison Noncompleter Nonstarter Chisquare p 3 (n=133) Not reconvicted Reconvicted 29 19 20 16 7 8 17 17 % reconvicted 39.6 44.4 53.3 50.0 1.33.720 4 (n=634) Not reconvicted 76 51 37 54 38.66 Reconvicted % reconvicted 71 48.3 68 57.1 106 74.1 171 76.0 <.001 5 (n=159) Not reconvicted Reconvicted % reconvicted 11 8 42.1 7 31 81.6 4 19 82.6 16 63 79.77 13.81.003 34

Percentage of starters completing programmes 35

Crime Reduction Programme overall outcomes: Comparing predicted and actual recidivism (Hollis 2007; total sample = 25,255) Programme category Sample size Percent reduction (overall) General offending behavior Percent reduction (completers) 12,924 11.0 26.4 Substance misuse 5,081 6.7 19.9 Drink impaired 4,101 12.5 28.3 drivers Anger management 1,148 14.2 25.5 Domestic violence 1,148 3.2 na Sexual offences 791 37.7 61.0 36

Re offending of adults (2006 cohort) Ministry of Justice Statistics Bulletin (September 2008) Comparing the reoffending results for the 2000 and 2006 adult cohorts shows: The frequency rate fell 22.9% (189.4 146.1 offences per 100 offenders) The number of offences classified as most serious fell 11.1% (0.78 0.69 offences per 100 offenders) The proportion of offenders reoffending (actual yes/no rate) decreased by 10.7% (4.7 percentage points = 43.7% 39.0%) The proportion p of offenders who reoffended fell by 10.6% when controlling for changes in offender characteristics 37

Frequency of re offending 2000 2008 (per 100 offenders) (Ministry of Justice, 2010) 38

Novaco s model of anger arousal EXTERNAL EVENTS frustration inequity assault COGNITIVE APPRAISAL BEHAVIOURAL RESPONSE verbal hostility passive withdrawal aggression AROUSAL + labelling 39

Aggression Replacement Training (ART): Outcome evaluation with adults (Sugg, 2000; McGuire & Clark 2004) Combines anger control + social skills training + values education One year follow up study of 297 offenders 230 traceable through Offenders Index 77 still on probation Of 153 evaluated cases, 113 completed (74%) Evaluation in terms of: comparison with predicted rates using OGRS 2 reconviction relative to matched samples on other community orders 40

ART results at 1 year follow up PREDICTED rates of reconviction (2 year) ART 34.6% ART revoked 54.0% Comparison 37.9% Comparison revoked 58.0% ACTUAL rates of reconviction (1 year) ART 20.4% ART revoked 65.0% Comparison 34.5% Comparison revoked 60.0% 41

Controlling Aggression Group (CAG) The Devon Probation Programme (Priestley, Gates, Phillips & McGuire, 2006) 14 sessions x 2 hours Problem solving, Anger management Two year follow up: 24 participants, 27 prisoners Non significant trend for general offending 42% versus 67% Significant ifi reduction in violent re offending 17% versus 44%

CAG outcomes at 2 year follow up The figure compares predicted and actual rates of 80 70 60 Predicted reconviction Actual reconviction Vi l reconviction 30 Violence re offending for 50 programme 40 participants (CAG) and a prison control sample, for both 20 general and 10 violent reoffending 0 CAG Prison

Some conclusions (so far) We can be more confident than ever regarding the evidence: on reducing recidivism outcomes on economic outcomes (benefit cost analysis) There is a strong case for retaining programmes as a core component of an effective criminal justice strategy However, improvements in effect sizes might not be a matter of developing better programmes, but of improving delivery More research is needed d on contextual t and other factors that influence programme implementation, individual engagement, g and quality of services There is evidence that criminal justice could be made more cost effective through a significant transfer of resources from prison to probation 44