From “Nothing Works” to “What Works” to “Making it Work” From “Nothing Works” to “What Works” to “Making it Work” James Bonta James Bonta Oporto, Portugal - April, 2008
From “Nothing Works” to “What
Works” to “Making it Work”
From “Nothing Works” to “What
Works” to “Making it Work”
James BontaJames Bonta
Oporto, Portugal - April, 2008
2
Outline of the Presentation
1. The Dark Ages: 1975 to 1990
“Nothing Works”
2. The Renaissance Period: 1990 to 2000
“What Works”
3. The Reality Check: 2000 to ?
“Making it Work”
3
Before the Plunge into Darkness
Review N % Effective
Kirby (1954) 4 75
Bailey (1966) 22 60
Logan (1972)* 18 50
N = number of studies reviewed* Logan reviewed 100 studies, but only 18 were studies
of counseling/therapy with a comparison group
“ l i t t l e r e a s o n t o h o p e t h a t
w e h a v e… f o u n d a s u r e
w a y o f r e d u c in g r e c id i v i sm
t h r o u g h r e h a b i l i t a t i o n ”
The Darkness Descends
5
“Nothing Works”
Review N % Effective
Lipton et al. (1975) 231 40-60
Findings not new but the consequences were…
6
Consequences of Nothing Works “Getting Tough”
• Utilitarian Models of Crime• Since crime pays, up the costs• Get tough• U.S. has ¼ of world prison
population (7 million under correctional supervision)
• Just Deserts• Punishment as a deserved social
value; It is the “fair” thing to do• Has this worked?
7
“Getting Tough” and Recidivism
Sanction φφφφ N
Prison vs Community Adults (71) -.07* 76,287
Juveniles (24) -.09* 4,118
Longer Sentence Adults (228) -.03* 68,303
Juveniles (5) .00 38,862
Intermediate Sanctions Adults (104) .02 44,870
Juveniles (59) .00 11,141
Smith et al., 2002
8
Checklist for Effective Punishment
⇒ Inhibits behavior - does not teach new behavior
⇒ Vary punishments (few universal punishers)
⇒ Immediate
⇒ Appropriate intensity
⇒ Type of person works best with:
* non-impulsive, future oriented
* average to above average IQ
* minimal punishment history
* cautious, avoids/minimizes excitement
9
Bonta fiddles while rehabilitation burns
10
Psychopathological Models
Assessment Treatment
Anxiety � Relaxation, medication
Intellectual/ � Educational/RemedialCognitive Deficits
Self-esteem � CounselingDepressionAlienation
Schizophrenia � Hospitalization,Manic-depression medicationHallucinationsDelusions
Emotional Discomfort and Crime
Risk Factor k n r
Emotional D iscomfort 66 19,933 .05
Gendreau et al., 1996
Theory Assessment Treatment
Limited Social Position EmploymentOpportunity/ (Race, SES, EducationAnomie/Strain Ethnicity) Anti-poverty
Labeling Social Position Non-intervention
Conflict Social Position Social JusticeRedistribute wealth
Subcultural Alienation Group cohesionSelf-esteem Leadership training
Sociological Criminology
Social Class and Crime
Risk Factor k n r
Lower Class 23 13,080 .05
Gendreau et al., 1996
14
Correctional Quackery
� “treatment interventions that are based neither on existing knowledge of the causes of crime or programs that have shown to change offender behavior”
� Dismissive of evidence
15
Correctional Quackery: Examples
� Music/Drama/Art/Horticultural Therapies
� Acupuncture
� TM; “Refining the Breath”
� Pet Therapy
� Men required to dress as women
� Vision Training
16
This is a Good Day to Die
17
The Renaissance
“What Works”
18
D. A. Andrews (aka “Moses”)
19
Risk and Treatment (% Recidivism)
Study
Risk
Low
Intensive
Bonta et al. (2000) Low
High
15
51
32
32
O’Donnell et al (1971) Low
High
16
78
22
56
Baird et al (1979) Low
High
3
37
10
18
Andrews & Kiessling (1980) Low
High
12
58
17
31
Treatment
20
Treatment Principle 1
Risk Principle
Match treatment services to risk level
21
Treatment Dosage
Lipsey’s (1995) meta-analysis
⇒ young offenders⇒ 100+ contact hours
Carleton University: Dosage and effect size
⇒ r = .05 (all studies; k = 374); but # hours r = .41⇒ r = .11 (youth; k = 193)⇒ r = .28* (female; k = 45)⇒ r = .22* (minority; k = 105)
Bourgon & Armstrong (2005) ⇒ matching of risk level with treatment⇒ precise measure of risk and treatment length⇒ at least 200 hours
22
Needs and Dynamic Risk
Criminogenic Noncriminogenic
Procriminal attitudes Self-esteem
Criminal associates Vague feelings
Family affect/discipline Physical training
Antisocial personality Group cohesion
(self-control, anger)
Work training with job Increase ambition
23
Treatment Principle 2
Need Principle
Target criminogenic needs
24
Treatment Principle 3
Responsivity Principle
Match treatment style to offender’s learning style
General: Cognitive-Behavioural
Specific: Personal factors
25
1. We need good assessments of risk and criminogenic needs
2. Not all treatments are the same
Implications from the
Principles
26
What Factors are Important in Offender Risk Assessment?
27
History of antisocial behaviour
Antisocial personality pattern
Antisocial cognition
The “Big Four”
Risk Factors
Antisocial associates
Family and/or marital
School and/or work
Leisure and/or recreation
The “Central Eight” Risk Factors
Substance abuse
28
Not all Treatments are the Same
1990 Meta-analysis: The Resurgence of Offender Rehabilitation
Implications from the
Principles: Part 2
29
Service or Sanctions?
Service:
2006: r = +.12
(n = 273)
Sanction:
2006: r = -.03
(n = 101)
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
Service
Sanction
30
Treatment by Risk
High Risk Cases:
r = .10
(n = 278)
Low Risk Cases:
r = .03
(n = 96)
0
1
2
3
4
5
6
7
8
9
10
High Low
31
Targeting Criminogenic Needs
Criminogenic
r = .19
(n = 169)
Noncriminogenic
r = -.01
(n = 205)
-2
0
2
4
6
8
10
12
14
16
18
20
Crim Noncr
32
Intermediate Targets
Criminogenic needs targeted exclusivelyr = .22 (CI = 18 - 25; k = 111)
Some criminogenic needs targeted, some non-criminogenic needs targeted
r = .13 (CI = 10 - 18; k = 81)
Unspecified targetsr = .00 (CI = -04 - .04; k = 83)
Non-criminogenic needs targeted exclusivelyr = -.03 (CI = -.06 - .01; k = 99)
33
Behavioural Treatment
Behavioural
r = .23
(n =77)
Nonbehavioural
r = .04
(n = 297)
0
5
10
15
20
25
Beh
Nonbeh
34
Traditional
Punishments
(30 studies)
Effectiveness of RNR Adherence:
% Difference (Expt-Control group)
ISPs
(47 studies)
Unspecified
Treatment
(54 studies)
Inappropriate
Treatment
(32 studies)
Appropriate
Treatment
(38 studies)
-7% -7% - 6%
13%
30%
-10
0
10
20
30
Adherence to Principles by Setting
-15
-10
-5
0
5
10
15
20
25
30
35
40
0 1 2 3
# of Treatment Conditions
Decrease
Increase
Recidivism
�Community
� Residence
36
Violent Recidivism: Mean ES by Level of RNR Adherence
-0.05
0
0.05
0.1
0.15
0.2
0 None 1 Low 2 Mod 3 Full
37
.02 .03
.17
.36
-.02
.02
.18
.24
-0.1
0
0.1
0.2
0.3
0.4
None Low Mod High
Women
Men
Adherence to RNR by Gender
38
RNR Adherence and Program Type
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0 None 1 Low 2 Mod 3 Full
Family
Academic
Drug/Alcohol
Work
39
Sex Offender Treatment: Adherence to Risk, Need and Responsivity
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
0.12
average phi
None 1 principle 2 principles All three
4 studies 6 studies 12 studies 1 study
40
Comparative effects sizes for selected interventions
Intervention Target Effect size
Aspirin Heart attack 0.03
Chemotherapy Breast cancer 0.11
Bypass surgery Heart disease 0.15
Offender Treatment Recidivism 0.12 (ns)
0.29 (approp)
41
42
Reality Check
Making it Work
43
Percent of Studies Adhering to the Principles of Effective Treatment
P r in c ip le %
R isk 7 4
N e ed 4 5
R e sp o n s iv ity 2 1
A ll 1 6
44
Treatment Principles
� Risk Principle -Match treatment services to risk level
� Need Principle- Target criminogenic needs
� Responsivity Principle -Match treatment style to offender’s learning style
� Program Integrity – Delivering the service as intended
45
The “Real World” vs. the “Small Demonstration” Project
• Mark Lipsey noted how the effects of treatment are much higher in demonstration projects than in the “real world” of routine programming
• We too find this effect in our meta-analyses
46
Level of RNR Adherence
Program Type 0 1 2 3
Demonstration .01 .07 .31 .34
Real World -.02 .04 .09 .15
RNR Adherence
47
Community Programs in Ohio
Latessa & Lowenkamp (2006)
• 13,221 offenders: 37 Halfway Houses; 15 Community Based Facilities
• Two-year follow-up conducted on all offenders
• Recidivism measures included new arrests & incarceration in a state penal institution
48
Supervision Intensity & Treatment (Risk Principle)
7
1
0
2
4
6
8
10
High Risk/More
Service
Low Risk/More
Service
Change in Recidivism
49
Criminogenic Programming and Treatment (Need Principle)
5
-16-20
-15
-10
-5
0
5
10
More than 25%
Criminogenic
Less than 25%
Criminogenic
% Change in Recidivism
50
Behavioural Treatment Model & Recidivism (Responsivity Principle)
8
00
2
4
6
8
10
Cognitive Behavioral Other
% Change in Recidivism
51
What is it about Demonstration Projects?
• Staff are carefully selected, trained and supervised
• Training and supervision according to relationship and structuring skills
• Specific model of treatment with manuals
• Adequate dosage
• Monitoring of process/intermediate change
• Small numbers
• Involved evaluator
52
Indicators of Program Integrity
Present
Indicator no yes
Selected for relationship skills .08 .48
Staff trained .02 .15
Staff clinically supervised .07 .18
Monitor intermediate targets .07 .11
Specific model .02 .14
Evaluator involved .04 .24
Small program/sample .05 .14
Appropriate dosage .07 .12
53
Making it Work
Adhering to “RNR Principles”
⇒ Risk-Needs assessment
⇒ Training staff to target criminogenic needs⇒ Responsivity: Relationship & Structuring
Monitoring of Service Integrity
⇒ Constantly evaluate what you are doing
54
Tony Soprano
• Tony’s psychotherapist has given up her pursuit of enhancing Tony’s emotional health
• If only she had known the RNR model
• Of course, crime prevention was never really her goal
• Traditional mental health concern with enhanced personal well-being
• Healthy more confident criminal