You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
Once an offender has been convicted, the criminal justice system faces a practical question that forensic psychology is well placed to answer: what should be done to reduce the likelihood that they offend again? Custodial sentencing, examined in the previous lesson, is one response, but it has a mixed record and high costs. This lesson examines three psychologically grounded ways of dealing with offending behaviour that can operate either inside custody or in the community: behaviour modification in custody (token economy programmes), anger management programmes, and restorative justice. Each rests on a different theoretical tradition — operant conditioning, cognitive-behavioural therapy (CBT), and a restorative, victim-centred social philosophy respectively — and each is therefore vulnerable to a different set of criticisms. A recurring evaluative theme is the contrast between interventions that change surface behaviour (behaviour modification), those that change the cognitions driving behaviour (anger management), and those that change the offender's relationship to the harm they have caused (restorative justice). The discussion is conducted in a measured, academic register, treating offenders and victims with appropriate seriousness and focusing on the evidence.
Key Definition: Behaviour modification is the systematic use of operant-conditioning principles — reinforcement and, sometimes, punishment — to change observable behaviour. In forensic settings it is most commonly delivered as a token economy within an institution.
This lesson addresses the following point from the AQA A-Level Psychology (7182) specification, Paper 3, Section D — Forensic Psychology:
It develops the theoretical basis, procedure and evidence for each of the three named interventions and prepares you to describe (AO1) and evaluate (AO3) them. Exam questions on this material are normally split AO1/AO3 only, with no AO2 unless a scenario stem is provided (for example, a description of a particular offender or institution to which a named intervention must be applied) — this lesson flags that distinction explicitly. The content links closely to the neighbouring custodial-sentencing lesson (custody's rehabilitative aim is delivered partly through these very techniques) and to the learning and cognitive approaches studied elsewhere on the course.
Token economy programmes apply Skinner's operant conditioning — the principle that behaviour is shaped by its consequences, so that behaviour followed by reinforcement becomes more frequent and behaviour followed by punishment or the removal of reward becomes less frequent. The token economy translates this laboratory principle into an institutional management system. Its logic is to make desirable conduct pay in a literal, immediate and consistent way, so that prisoners learn, through reinforcement, to behave in ways the institution values.
A token economy has four defining features:
A token bridges the gap between the moment of good behaviour and the later reward, which is why tokens are necessary at all: they allow reinforcement to be delivered instantly, when it is most effective, even though the privilege it buys comes later. The approach is closely related to the Incentives and Earned Privileges (IEP) schemes used in the modern English and Welsh prison estate, in which prisoners move between basic, standard and enhanced regimes according to behaviour.
graph LR
A[Target Behaviour Performed] --> B[Token Awarded Immediately]
B --> C[Tokens Accumulated]
C --> D[Exchanged for Primary Reinforcers / Privileges]
D --> E[Behaviour Positively Reinforced]
E --> A
F[Undesirable Behaviour] --> G[Tokens Withheld or Removed: Response Cost]
Hobbs and Holt (1976) introduced a token economy across three cottages of a youth detention facility, retaining a fourth cottage as a no-token control. Target social behaviours (such as following rules and cooperating) were rewarded with tokens exchangeable for privileges. Their finding was that desirable (prosocial) behaviour rose substantially in the token cottages compared with the control cottage, in which no comparable change occurred. The conclusion is that a properly run token economy can reliably increase targeted prosocial behaviour within an institution. Supportive evidence also comes from adult prison settings: studies of token systems have likewise reported improved compliance and reduced disruptive behaviour while the contingencies are in force.
A point often missed at A-Level is that the success of a token economy depends heavily on how faithfully it is implemented — its fidelity. For the conditioning to work, tokens must be awarded immediately, consistently and contingently (only for the target behaviour), and staff across all shifts must apply the rules in the same way. In practice, busy or untrained staff may award tokens erratically, delay them, or give them out of sympathy, which weakens or destroys the contingency. This means that an apparent "failure of behaviour modification" is sometimes really a failure of delivery, an important nuance for evaluation.
| Strengths | Limitations |
|---|---|
| Demonstrably effective in institutions — Hobbs and Holt (1976) and adult-prison studies show increases in targeted prosocial behaviour while the system is running | Poor generalisation — behaviour learned for tokens may not transfer to the community, where no one dispenses tokens; the change can be situation-specific rather than internalised |
| Easy and cheap to administer — target behaviours are clearly defined and measurable, and the system can be run by ordinary staff without specialist therapists | Does not address causes — it changes surface behaviour but leaves the cognitive, emotional and social drivers of offending untouched, so its effects may be shallow |
| Grounded in a well-supported theory — operant conditioning has extensive empirical support across species and settings | Ethical and rights concerns — using privileges to control prisoners can be seen as manipulative, and there are limits on what can legitimately be made contingent (basic decency and welfare must not be withheld) |
| Produces immediate behavioural improvement, which can make institutions safer and more orderly | Requires high fidelity — inconsistent or delayed token delivery undermines the contingency; effects often fade once the programme stops, suggesting weak durability |
Anger management programmes apply cognitive-behavioural therapy (CBT), drawing specifically on Novaco's (1975) analysis of anger and on the broader logic of stress inoculation training. The central assumption is cognitive: anger is not caused directly by an event but by the way the individual appraises (interprets) that event. A perceived slight interpreted as a deliberate, threatening insult produces far more anger than the same event interpreted as accidental. Where anger reliably triggers aggression and offending, helping the offender to recognise and restructure these maladaptive appraisals — and to deploy coping skills before anger escalates — should reduce violent behaviour. Anger management therefore treats anger as a learned, cognitively mediated response that can be relearned.
Novaco's approach is typically delivered in three phases:
graph TD
A[Stage 1: Cognitive Preparation] --> B[Stage 2: Skill Acquisition]
B --> C[Stage 3: Application Practice]
A --> A1[Identify triggers]
A --> A2[Recognise physiological cues]
A --> A3[Understand role of appraisal]
B --> B1[Relaxation techniques]
B --> B2[Cognitive restructuring]
B --> B3[Self-talk and assertiveness]
C --> C1[Role-play scenarios]
C --> C2[Graded exposure]
C --> C3[Therapist feedback and reinforcement]
Stage 1 — Cognitive preparation. The offender works with a therapist to become aware of their own anger: identifying the triggers (situations, people, events) that typically provoke them, recognising the physiological signs of rising anger (raised heart rate, muscle tension, clenched fists), and grasping the central insight that it is their interpretation ("They're disrespecting me") rather than the event itself that drives the anger. This stage reframes anger as something with identifiable antecedents that can be anticipated and managed.
Stage 2 — Skill acquisition. The offender is taught concrete coping skills: relaxation techniques (controlled breathing, progressive muscle relaxation) to dampen the physiological arousal; cognitive restructuring, replacing hostile, irrational appraisals with calmer alternatives (e.g., reinterpreting "He's trying to wind me up" as "He probably didn't mean anything by it"); self-talk, using rehearsed internal statements ("Stay calm, I can handle this") to manage the emotional response; and assertiveness and communication skills, so that needs and grievances can be expressed without aggression.
Stage 3 — Application practice. The new skills are rehearsed in controlled, simulated situations through role-play of anger-provoking scenarios, using graded exposure that begins with mildly irritating situations and works up to more provocative ones. The therapist provides feedback and reinforces successful self-regulation, so that the skills become more automatic before the offender has to use them for real.
Key Definition: Cognitive restructuring is the CBT technique of identifying irrational, maladaptive appraisals and systematically replacing them with more rational, adaptive interpretations — here, of the events that trigger anger.
Keen et al. (2000) evaluated the National Anger Management Package with young offenders aged 17–21 in custody, delivered over several sessions. Although early sessions were marked by some non-cooperation (forgetting materials, reluctance to engage), offenders who completed the course reported increased awareness of their anger difficulties and greater capacity to control their anger on self-report measures. Ireland (2000) compared 50 young offenders who completed an anger-management programme with a control group: assessed by self-report, prison-officer ratings and behavioural checklists, the treated group showed clear improvement on at least one measure in a substantial majority of cases, whereas the control group did not improve. The conclusion across these studies is that anger management can produce measurable short-term reductions in anger and aggression in custodial young offenders, particularly on cognitive and self-report measures.
Restorative justice (RJ) reframes crime not primarily as an offence against the state but as harm done to a victim and a community, and it focuses on repairing that harm. Rather than (or alongside) punishment, RJ typically brings the offender and victim — sometimes with community members and trained facilitators — into structured, voluntary contact, often a face-to-face conference. Within this meeting:
The theoretical underpinning draws on Braithwaite's (1989) influential idea of reintegrative shaming: the aim is to make the offender feel the wrongfulness of the act (shame) in a way that reintegrates them into the moral community, in contrast to stigmatising shaming, which labels and excludes the offender and thereby tends to push them further into a criminal identity. Done well, RJ is intended to promote empathy and genuine responsibility in the offender while restoring the victim's sense of security and closure.
Sherman and Strang (2007) reviewed a large body of restorative-justice research, including randomised studies. Their findings were that, across many evaluations, RJ reduced reoffending — with the effect strongest for violent crime — and consistently produced higher victim satisfaction than conventional criminal-justice processing. Victims who took part reported reduced fear of the offender, a reduced desire for revenge and a greater sense of closure, and there was evidence of reduced post-traumatic stress symptoms in some victims. The conclusion is that, for suitable offences and willing participants, RJ can benefit victims and modestly reduce reoffending. The review also noted that RJ is less suited to some offences and that effects vary with how programmes are run.
| Strengths | Limitations |
|---|---|
| Centres the victim — conventional justice often marginalises victims; RJ gives them a voice and is associated with high satisfaction and reduced fear (Sherman & Strang, 2007) | Not suitable for all crimes or victims — inappropriate where there is a serious power imbalance (e.g., domestic abuse, sexual offences), or where the victim does not wish to take part |
| Can reduce reoffending, with the largest effects for violent crime, supporting Braithwaite's reintegrative-shaming logic | Depends on genuine offender engagement — some offenders may participate insincerely (e.g., to secure a lighter sentence), undermining the process |
| May build empathy and responsibility, addressing a motivational root of offending that punishment alone does not reach | Risk of secondary harm — a poorly facilitated meeting can re-traumatise the victim, so skilled facilitation and careful selection are essential |
| Cost-effective relative to custody, and flexible (can run pre-sentence, post-sentence or instead of prosecution) | Perceived as "soft" — political and public resistance, and a concern that it may seem to under-punish serious offences |
| Feature | Token Economy | Anger Management | Restorative Justice |
|---|---|---|---|
| Theoretical basis | Operant conditioning (Skinner) | CBT / Novaco (1975) — appraisal and stress inoculation | Reintegrative shaming (Braithwaite, 1989); victim-centred justice |
| What it targets | Observable behaviour | Cognitive appraisals and behavioural anger responses | The offender–victim relationship and the harm caused |
| Where it operates | Within the institution | Within the institution (or community) | Community, institution, or pre-/post-sentence |
| Key evidence | Hobbs & Holt (1976) | Keen et al. (2000); Ireland (2000) | Sherman & Strang (2007) |
| Addresses root causes? | No — surface behaviour only | Partly — addresses the cognitions behind anger | Partly — addresses empathy, responsibility and victim need |
| Main limitation | Poor generalisation beyond the institution | Only helps offenders whose offending is anger-driven | Unsuitable for some offences; needs genuine engagement |
Exam Tip: The strongest evaluative move is to argue that these approaches are complementary, not competing. No single intervention addresses every cause of offending: token economies manage behaviour in the short term, anger management targets the cognitions behind violent offending, and restorative justice addresses empathy and victim need. A combined, individually matched approach is likely to be most effective — and saying so explicitly is a top-band discriminator.
Subscribe to continue reading
Get full access to this lesson and all 10 lessons in this course.