Offender Treatment Programmes


Which forms of treatment are likely to be the most effective?


Fagan (1990) has argued that the actual length of time that a young violent offender spends in custody is not necessarily a crucial factor in determining reoffending rates. What appears more important is that such high-risk youths are prepared for their release back into the com­munity through the provision of well run reintegration programmes which are rooted in sound theory and advanced practices. In addition, Fagan argues that on release there should be intensive supervision with an emphasis on a gradual re-entry. Reviewing a number of such pro­grammes, he notes that: ‘The well-implemented programs resulted in significant reductions in the number and severity of arrests ... as well as in significantly greater time until rearrest’ (1990: 258).

Andrews et al. (1990) have identified a number of characteristics which appear to differentiate successful rehabilitation programmes from the less successful ones. They observe importantly that ‘The effectiveness of correctional treatment is dependent upon what is delivered to whom in particular settings’ (1990: 372).

Andrews et al. argue that successful programmes are those which have the following characteristics:

(1)  They are highly structured and focused and address a distinct problem in a distinct way.

(2)  They employ staff who are firm but fair and who reinforce anti-criminal values. In addition, the staff are committed and enthusiastic and are supported by good management.

(3)  The programmes target in particular those attitudes and values which support offending behaviour.

(4)  The programmes employ problem-solving procedures based on cog­nitive and social learning perspectives.

(5)  Programmes are matched to offender characteristics.

(6)  Programmes are monitored and evaluated in relation to both recidi­vism rates and personal growth.

(7)  Programmes target medium and high risk offenders.

(8)  Programmes attempt to generalise beyond the institutional setting.


Cognitive skills programmes

Cognitive skills programmes are regarded by some as an effective means of reducing the recidivism by those taking part (Blud, 1999). The extent of this reduction is put in the order of 10-30%. Arguably, these treatment programmes are effective because they work to correct a number of cognitive deficits exhibited by offenders. These cognitive deficits are identified as follows:

. Self-control - to correct impulsiveness and action.

. Cognitive style - lack of empathy with abstract social concepts. Such matters as social harmony and justice may be involved here. Such offenders may be rigid and inflexible in their thinking. They may be poor at tolerating ambiguity, which results in what may be described as rather simplistic, dogmatic thought processes.

. Interpersonal problem solving - offenders are often socially handicapped without recognising the fact. They are sometimes unaware of the other courses of action available to them. They rarely examine the likely consequences of their actions.

. Social perspective taking - prisoners are often egocentric. They fail to understand why they should consider other people. They may lack the skill involved in seeing matters from the perspective of other people. They tend to interpret the actions of others in their own terms and int~rpret the actions of others wrongly as a consequence.

. Values - their moral reasoning skills are poor. They do not recognise the incongruity between their actions and their beliefs.

. Critical reasoning - their thinking is often irrational and illogical. Self-analysis is avoided. They justify what they do by blaming others and, consequently, do not see themselves as to blame.


Anger reduction schemes


Some of our current understanding of anger stems from the work of Novaco (1975). Novaco stressed the role which cognition plays in the emotional arousal which generally precedes aggressive and violent acts. Novaco argued that anger is often quick to surface because the individ­ual is already upset about other things in their life. The individual tends to displace angry feelings onto an available target, even if this is entirely inappropriate. In this case, expressing anger may be one way in which the individual feels that he/she can take control of a situation. This is particularly likely in circumstances in which the person feels insecure or threatened. In behaviourist terms, becoming angry is reinforced by the individual’s feeling of control in the situation.

Novaco suggests that any intervention should not try to stop the individ­ual from experiencing anger — rather it should allow him/her to monitor, control and manage angry reactions. Thus, the individual will need to be taught new strategies for regaining self-control, and be encouraged to develop techniques which allow the resolution of conflict situations with­out resort to violence. Most anger management programmes involve individuals working in groups and comprise three distinct stages.


1. Cognitive preparation


In this phase, individuals are encouraged to think about and to analyse their own patterns of anger. In particular, group members would be encouraged to try to identify the specific situations which have triggered an angry response in the past. By doing this it is hoped that individuals can learn to recognise the types of situation which are potentially dan­gerous (in terms of invoking an angry response). Individuals would also be encouraged to try to recall their pattern of thought during the inter­action. By doing this, the therapist/counsellor may be able to challenge some of the irrational thoughts and assumptions which the individual makes (for example, about what another’s actions signify). By tracing the individual’s pattern of behaviour, it is hoped that it will be possible to break what may have become an almost automatic response. For exam­ple, the individual might recognise that teasing by others invariably leads to physiological arousal, which quickly leads to a violent outburst.

Individuals would also be encouraged to think about the conse­quences of their anger. Specifically they would be taught that although their anger may have served some purpose at the time, it probably had long-term negative consequences. For example, the individual may be left with feelings of guilt, may be banned from a favourite pub, lose a girl­friend or be punished by the courts.



2. Skill acquisition


As the name implies, this stage allows individuals to learn a range of new skills which can allow them to deal with potentially anger-provoking situ­ations in a more effective way. The skills to be learned are both cognitive and behavioural. Cognitive skills would be predominantly concerned with challenging the automatic thought processes which are perhaps irrational yet well rehearsed. Thus, the individual might be instructed simply to say ‘stop’ to themselves when the normal thought patterns emerge. Alternatively, they might practise saying to themselves some­thing along the lines of ‘I can deal with this situation in an effective way’ rather than relying on previous maladaptive responses.

There are a number of different behavioural skills which might be taught including relaxation, assertiveness, social and communication skills and problem-solving strategies. Relaxation training may be particu­larly helpful as it serves two purposes: it teaches the individual to relax, thus making him/her less tense generally and, secondly, it teaches the individual that it is possible to take control of emotions, rather than be ruled by them. Assertiveness training can also be very useful for some individuals, as the techniques learned can allow the person to achieve an objective without resorting to violence. Many people confuse assertive­ness with aggressiveness, but the two are very different. In the case of assertiveness training, the individual learns to make his/her point in a calm and effective way. The individual is also encouraged to acknowl­edge others’ feelings and wishes, but to stick to their legitimate request.


3. Application practice


In the final phase, individuals would be allowed the opportunity to prac­tise their skills in controlled conditions. Each person might, through role-play, be exposed to a range of situations which have caused prob­lems in the past. The person would typically begin with situations which were only slightly problematic and would then gradually work up to other more difficult scenarios. Throughout these exercises the individ­ual would be encouraged to monitor their progress, but would also receive encouragement (or reinforcement) from others. The individual would then need to transfer the skills from the controlled situation within the group to everyday occurrences.

There is no doubt that anger management programmes can be very useful as a way of helping individuals to change their established patterns of behaviour. However, programmes can only work if sufficient resources are devoted to the initiative, the appropriate individuals are assigned to the programme and careful monitoring takes place. While anger man­agement can be useful as a general method of reducing aggression in an offender population, in some cases a more individual and targeted pro­gramme is necessary such as with men who have been convicted of violence against their female partner.

Text Box: Commentary: Controlling aggression reward or punishment
•	Punishment of aggressive behaviour. Generally punishment of aggressive behaviour has some effect, but it can lead to displaced aggression elsewhere. Sherman and Berk (1984) found that arresting the violent partner in domestic violence cases was most effective by the Minneapolis police department. There were only 10% cases of re-offending compared to 24% for ‘separating partners for cooling down period’ or 19% for ‘counselling both parties’.
•	Punishment of aggressive models — Bandura’s work on observational learning and modelling has shown that seeing an aggressive model being punished reduces overt aggression (Bandura, Ross and Ross, 1963). But it does not produce long-term change as the aggressive behaviour will appear if rewards are offered later for such behaviour.
•	Rewarding alternative behaviour patterns — it is better to reward non-aggressive behaviour patterns. Brown and Elliot (1965) asked nursery school teachers to ignore aggressive behaviour among the children and give attention only when desired behaviour was shown. The average number of aggressive acts fell from 64 before the experiment to 26 two weeks later.

The National Anger Management Package

Keen (2000) provides an example of the preparation and delivery of an anger management course within a young offender institution, conducted with young male offenders aged between 17 and 21 years. The programme, first devised in 1992 and updated in 1995, was developed by the Prison Service in England and Wales and is known as the National Anger Management Package (Prison Service, 1995).

The aims of the course are as follows:


    to increase course members’ awareness of the process by which they become angry


    to raise course members’ awareness of the need to monitor their behaviour


    to educate course members in the benefits of controlling their anger


    to improve techniques of anger management


    to allow course members to practise anger management during role plays.


The course involves eight two-hour sessions, the first seven over a 2—3 week period, with the last session a month afterwards. The details of the course are as follows (for more details see Towl & Crighton, 1996).

Keen’s experiences of using the course are generally positive, though there are various predictable problems with young incarcerated males, such as failure to bring the anger diary to sessions, grins and sniggers at the mention of ‘bodily arousal’ and a certain degree of egocen­trism which can make relationships and progress in the group a little difficult at times. Overall, though, the feedback from the individuals completing the course indicates that they have increased their awareness of their anger management difficulties and have increased their capacity to exercise self-control.


Ireland (2000) using two groups (experimental and control) and three measures (Interview, Behavioural checklist completed by prison officers and a self-report questionnaire) found that 92% of the experimental group showed an improvement on at least one measure, 48% showed an improvement on the checklist and self-report.  No differences were found for the controls.


Evaluation of anger management programmes


The effectiveness of anger management programmes to reduce recidivism in violent offenders is called into question by several researchers. Although it may seem common sense to propose that anger is a serious problem among those convicted of violent crimes such as assault, researchers disagree about whether there is a link between anger and violent crime. For example, while Zamble & Quinsey (1997) argue in favour of such a relationship and propose that uncontrolled anger is a risk factor in predicting violence and recidivism, others dispute it (e.g. Muirhead, 1997).


Prominent amongst those who do not believe that anger is a primary cause of violent criminal acts are Loza & Loza-Fanous (1999). They argue that many of the research findings linking anger with violence and rape were based on laboratory studies using students (e.g. Zillman, 1993) or offenders’ own explanations of their violent acts (Zamble & Quinsey, 1997). In fact they maintain that


‘Unfortunately, most of the opinions supporting a link between anger and violent behaviour, rape, or recidivism are based on speculations with very little empirical support such as results of valid and reliable psychometrics.’ (page 492)


They studied 271 Canadian male offenders, comparing a group of violent offenders with non-violent ones and a group of rapists with non-rapists. Violent offenders were those who had committed crimes such as murder, assault and robbery with violence. The non-violent offenders had a history of moderate or minor offences such as fraud or property offences. Using several psychometric measures of anger, they found no difference between violent offenders and non-violent offenders and between rapists and non-rapists on anger measures. Loza & Loza-Fanous argue that not only are anger treatment programmes ineffective with violent offenders, they have the potential to be harmful by encouraging the offender to attribute his violent actions to anger for which he cannot be blamed, rather than taking full responsibility for his actions.


Anger management programmes have, as we have seen, proved to be useful in controlling aggressive antisocial behaviour in prisons, but there needs to more research into the relationship between anger and crime before an assessment of its use and effectiveness in reducing rates of recidivism amongst violent criminals can be made.


Tackling male violence against a partner


Kennerley (2001) studied forty-five female perpetrators (FPG) were compared to 42 male perpetrators (MPG) and 46 female college students (FCG) in an attempt to describe female perpetrators of domestic violence. Participation was restricted to those who had been in or were currently in a married or cohabiting heterosexual relationship of at least six months duration. Criticisms of previous examinations of female perpetrators have centered on the failure to consider the incidence and type of violent behaviors they employ, the context in which their violent behaviors occur, and whether they initiate the use of violence in the relationship or in the incident. Participants were administered a Personal Violence History, Motivation Checklist and the Conflict Tactics Scale-2. To further expand the extant literature on female perpetrators, a measure of adult attachment style (Experiences in Close Relationships), propensity for abuse (Propensity for Abuse Scale), and social desirability (Balanced Inventory of Desirable Responses) were also administered. The FPG admitted using violence against a previous partner more often than the MPG. Continuums of domestic violence behaviors ranging from victimization to perpetration with the majority of participants reporting mutually combative relationships were found in both groups. Patterns of responding suggest that females with victimization scores were younger and had younger partners, reported physical assault and psychological aggression by their partners, and endorsed characteristics consistent with relationship avoidance. Social desirable responding tended to be associated with low reports of victimization. Females with perpetration scores tended to be older, had older partners, and reported physically assaulting, inflicting injury upon, and using negotiation skills with their partners. Clinical implications and suggestions for future research are presented.


According to Dobash and Dobash, men who use violence against a partner are seen to be acting within a culturally defined role in which the man has greater power and authority and uses violence as a way of con­trolling and punishing his partner. In this situation, violence is regarded as a learned behaviour and is often used as a way of resolving conflicts. The violence is seen as an intentional act designed to intimidate and control. Typically, men who behave in this way are egocentric, lack empa­thy, minimise the harm done, deny responsibility and try to deflect the blame onto others (usually the victim). Intervention thus tries to instil different attitudes towards women and the learning of new forms of behaviour in conflict situations. It also aims to analyse and help the offender to understand his own behaviour and to persuade the offender to accept responsibility for his actions.

In their recent review, Vennard and Hedderman suggest that offenders who attend cognitive-behavioural programmes achieve a 10—15 per cent lower reconviction rate than those who do not attend such programmes. Further, the authors report that programmes which stick closely to effec­tiveness principles achieve a recidivism rate twenty per cent lower than control groups. In addition, those programmes which include social skills training show the most positive results. It is, however, discouraging to learn that although many probation services in England and Wales recognise the benefits of a cognitive-behavioural approach, many pro­grammes have not achieved good results. Hedderman and Sugg (1987) suggest that the main reasons are threefold: a lack of programme integrity and inadequate staff training, a failure to adopt specialised pro­grammes and a failure to examine whether specialist programmes are effective.

Dobash et al. provide an interesting evaluation of two court-mandated programmes first set up in Scotland (CHANGE and The Lothian Domestic Violence Probation Project (LDVPP)). Their purpose was to establish whether the two programmes studied were more successful than more orthodox sanctions in reducing violence and changing atti­tudes over a one-year period. They note that in this field it can be particularly difficult to provide good objective evaluations. Although there have been more than 30 evaluation studies of interventions in the USA and Canada (most of which claim a success rate of between 50 and 80 per cent), many such evaluations suffer from methodological inade­quacies. The main deficiencies concern small sample sizes, selection bias, lack of control groups, insufficient follow-up period and a reliance on offender’s self-reports or official arrest records. In addition, many evaluation studies have failed to control for the possible termination of contact between the two parties concerned. Thus, in some cases, a man s conviction for domestic assault may result in separation or divorce. In these cases there may be no subsequent contact between the perpetrator and victim and thus no arena for violence to re-emerge. Interestingly, Dobash et al. report that post treatment evaluations rarely include follow-up interviews with women who have been victims — such a measure may well provide a better indicator of an offender’s subsequent behaviour and thus the programme s success.

Dobash et al.’s evaluation study was designed to eliminate most of the inadequacies of previous studies, although for a number of reasons it was not possible to use a randomised design. It was, therefore, necessary to show that people assigned to the two programmes under scrutiny were broadly similar to those given other sentences/treatment. Dobash et al suggest that this was indeed the case. Their evaluation used a number of different measures and it is interesting that different types of evaluation gave very different impressions. For example, an examination of court records showed that only seven per cent of men on the programmes and ten per cent of men dealt with by alternative means came before the courts for a repeat offence in the following twelve months. Such figures taken alone might suggest that most of those convicted of domestic assault do not commit further offences and that treatment programmes are only marginally better than other forms of disposal.

However, an examination of other data reveals a rather different pic­ture. For example, three months after the initial interview, 30 per cent of women whose partners were in the programme group reported having suffered further violence. This compares with 62 per cent of women whose partners were given other sentences. This difference was statisti­cally significant. Data gathered after twelve months showed a similar pattern — 34 per cent of men who had been on the treatment pro­gramme committed further violence over the year, compared with 75 per cent who had received other disposals. Such figures do not, however, give the whole picture, as they take no account of the frequency of further vio­lence. In fact Dobash et al. found that after one year, only seven per cent of women whose partners had been on the programmes reported fre­quent violence. This compares with 37 per cent of the women in the other group, a difference which was again statistically significant.

Dobash et al. conclude that:


in contrast to other criminal justice sanctions, programmes for violent men can have significant effects on the prevalence and frequency of violence over a twelve month period following the imposition of a sanction.

(1999: 229)


However, the authors note that such a conclusion can only be reached as a result of the data obtained from victims over the twelve-month follow-up period — evaluations based on court records alone would have given a very different and incorrect impression. There are in fact a number of reasons why simplistic measures such as reconviction rates do not offer a reliable measure of treatment effectiveness. Lloyd et al (1994) suggest that the main reasons are:


(1)  Reconviction rates take no account of changes in the severity of offences nor of changes in the frequency of offending.

(2)  As Dobash et al. discovered, reconviction rates underestimate the real level of reoffending, as many types of crime have low rates of reporting and detection.

(3)  Different areas and different police forces have differing recording and clear-up rates and so the chances of being arrested and con­victed can differ from one area to another.

(4)  There is inconsistency in the follow-up period used in many recon­viction studies. Some studies which follow up for a year may claim a low reconviction rate, whereas those which use a longer period may show a higher rate. An additional problem is that, as we saw earlier, there is invariably a significant time lag between an offender’s crime and their appearance in court. Offenders may actually commit fur­ther offences during the follow-up period but these would not be counted, as the offender’s court appearance would be up to a year after the actual offence.

(5)  The follow-up period for community sentences begins from the date of sentence, whereas that for custodial sentences begins with the date of release. It is, therefore, not always possible to make valid comparisons between the two types of sentence.


Dobash et al. suggest that their results are in line with a number of evalu­ation studies carried out in the USA in recent years. Successful programmes appear to be mainly educational and have a core of central characteristics, including high programme integrity, good management and the delivery of structured interventions focusing on the offender and his offending behaviour. Successful programmes also appear to have a cognitive-behavioural focus. As such they attempt to change the values, attitudes, beliefs and actions of offenders. Dobash et al. suggest that suc­cessful programmes are those which improve a person’s internal control, develop appropriate social skills, increase critical reasoning about the offence, establish good problem-solving strategies and increase empathy. However, the authors also note that in order to be successful, pro­grammes should not be voluntary, but rather should be part of an overall process or package of sanctions. The process of being arrested, charged and prosecuted provides a lever or incentive for the individual to partici­pate in the programme and, hopefully, to avoid more severe punishment in the future.


Psychotherapy or Psychodynamic therapy

Psychodynamic therapy is based on the work of Sig­mund Freud. The aim is to discover the childhood and/or unconscious cause of the behaviour. This tends to have limited use because it is time-consum­ing and can be superficial, if not handled correctly.

Persons (1967) performed one of the few out­come studies on its use with boys at a reform school. The 82 boys were matched on demographic and criminal variables, and half were randomly chosen to receive 80 hours of therapy over 20 weeks. The therapy was individual psychoanalysis or ‘eclectic’ (combination of approaches) group therapy. Imme­diately afterwards, 30 of the treatment group showed less psychological problems compared to 12 in the control group, while one year later, 13 of the treated and 25 of the untreated boys had been re-institutionalised. Generally the treated group did better, but the improvement could either be due to the therapy or, alternatively, due to the attention given to the boys in the treatment group.

Psychotherapy has not been used as much in the treatment of adult offenders, but a number of in-patient psychiatric units have been set up in Britain, the best known of which is Grendon Underwood, established in the late 1960’s as a full-scale psychiatric prison. Prisoners are transferred to it from other prisons and are not admitted directly from the courts. It has as many staff as inmates (about 150 of each) and is organized as a “therapeutic community,” the emphasis being on staff—prisoner and prisoner—prisoner relationships. All inmates meet daily in groups, each of which has a therapist. The prisoners’ rooms are well heated and adequately furnished, with views over the countryside (Parker 1970).

After a two-year follow up, Gunn, Robertson, Dell, and Way (1978) found a reconviction rate of 70 per cent among a sample of about 100 men who had been treated in Grendon, compared with 62 per cent of a random sample of British prisoners. However, the authors reject this comparison group as inappropriate, because the Grendon inmates had been selected for treatment, and wished to receive it. They assert that the objective of Grendon is treatment for psychological problems, not the reduction of recidivism, and that interview and MMPI data attest to the success of the former. They agree that if the aim is to assess the efficacy of Grendon in reducing reconviction then random allocation to Grendon and a control regime is the method of choice, but that this is impossible on both practical and ethical grounds. There is general agreement (e.g., Smith 1984) that both Grendon and another therapeutically oriented British unit, Barlinnie, have much lower rates of violence than similar prisons with more custodial emphases, but this is also a separate issue from that of reconviction rates.


Intervention:  non-residential settings

Once again, the results of psychotherapy are no better than those of control conditions. Extensive casework with 970 inner city convicted young offenders as part of the Chicago Youth Development Project, compared with 570 who received no treatment, resulted in no dif­ferences in either reconviction rates, or in unemployment (Gold and Mattick 1974). Adams and Vetter (1981) used a commendably long follow-up of 10 years and found reconviction rates to be higher in a group of young offenders treated by psychodrama (in effect “acting-out” antisocial impulses) than in an untreated comparison group.


Methods: the behaviour therapies



The psychology of learning has been the major source of the behaviour therapy approach (also termed behaviour modification) to the treatment of psychological problems, but other areas of experimental psychology, particularly cognitive and social psychology, are making increasing contributions. The aim is to modify the current, behavioural problems of the individual, couple or group. A conservative view of the relatively short history of the behaviour therapies (the major developments started only in the 1950s) is that their therapeutic efficacy is encouraging, particularly for phobias, anxiety problems, obsessions and sexual difficulties.



Scared straight.” This is a very dramatic and highly publicized approach which, though not formally behavioural, is appropriately placed here because it can be considered as an intended instance of negative modelling.

In September 1976, the Lifers Group at New Jersey Rahway State Prison had a rap session with a group of juveniles which evolved into a program of “group intimidation more akin to shock therapy” (Homant and Osowski 1982). In April 1978, a Los Angeles TV station filmed the documentary “Scared Straight,” which was shown throughout the United States and won Emmy and Oscar awards for best documentary film. According to Homant and Osowski (1982), by that year there had been at least 28 imitation programs in the United States and Canada, with some claiming up to 90 per cent rates of success, but that careful studies were rare. In their own research in Michigan, they randomly assigned to “treatment” and control groups boys who had been referred by the juvenile courts and had individually agreed to participate. At six months follow-up, those in the JOLT (Juvenile Offenders Learn the Truth) group exposed to a confrontation session (“if you don’t change you’ll be in this hole permanently”) were compared for police arrests with a control group: 31 per cent of the JOLT group and 29 per cent of the controls had been arrested. An even worse result was reported by Finkenaur (1982), who found that juveniles “scared straight” were more likely than controls to commit subsequent and more serious offenses.


Achievement Place. The next set of preventive approaches is less dramatic but much more carefully based. The Achievement Place program of Phillips (1968) and his associates has resulted in copies of their method in many parts of the USA. Essentially, boys between 12 and 14 who have been convicted for minor offences and are regarded by the county court as being at risk of carrying out more serious ones, live in small groups, “as in a family” with a pair of experienced “house parents.” The boys undergo a complex program based on token economy principles (Ayllon and Azrin 1968). This is a system of reinforcers, typically poker chips, applicable to nearly all behaviours, which can be distributed or removed readily in response to desired or undesired behaviours, and has been used in a very wide range of settings. There is also partial self-government, and much attention is given to skills and achievements relevant to the outside world.

Achievement Place has substantial benefits in terms of educational and social skills (Braukmann, Fixsen, Phillips, and Wolf 1975, Weinrott, Jones, and Howard 1982), but what of legally related indices? An early report was optimistic. Two years after release, the reconviction rate of a group of Achievement Place boys was only 19 per cent as against over 50 per cent of those placed on probation or committed to a conven­tional institution (Fixsen et al. 1972). But, as the authors point out, the boys were not randomly assigned to treatments, so that the result may have been due either to a “population effect” or to a “treatment effect” Three subsequent, large scale, studies agree in their conclusion: what is now termed the TFM (teaching family model) is not superior to comparison group home programs for recorded offences during follow-up periods of up to three years (Jones 1978. Kirigin, Braukman, Attwater, and Wolf 1982, Weinrott et al. 1982). The last named report also considered the TFM model to be expensive relative to alternative programs, a point of great importance to those who fund facilities for offenders.


Family-based programs. Generally, family therapy as a treatment has become more popular in an attempt to reduce delinquency. There is Parent Management Training (PMT), which focuses on training the parental skills to deal with prob­lem behaviour, while Functional Family Therapy (FFT) concentrates on the interactive process within the fam­ily system. Alexander and Parsons (1973) found lower recidivism in an eighteenth month follow-up, when comparing a FFT group with a control group (that is, no family therapy). The pitfall is with family therapy is that the family often gets the blame as the cause of crime.

 According to Patterson (1986) the problems of social behaviour that leads eventually to delinquency first arise in the ways parents and children interact with each other. The clear implication is a need to train many parents in effective child-management skills before their children reach the age at which they are at risk of becoming offenders. Essentially this is an example of the triadic approach advocated by Tharp and Wetzel (1969) for work in the natural environment. (The behaviour therapist trains and monitors professionals or parents who then work directly with the offender or pre-offender.)

O’Dell (1974) set out the advantages of behaviour modification pro­grams with families as follows:

1.     Programs can be carried out by previously unskilled non-professionals.

2.     Many people can be trained at any one time, and relatively short training periods suffice, although Alexander, Barton, Schiavo, and Parsons (1976) tend to contradict this.

3.     A minimum of professional staff is required (hence costs are relatively low).

4.     Many parents like the behavioural treatment model because it does not assume “sick” behaviour.

5.     Many childhood problems consist of well-defined acts appropriate to a behavioural approach.

However, in a very influential book on behavioural work in the natural environment, Tharp and Wetzel (1969) suggest caution, pointing out that many parents raise objections. These include a belief that rewarding children for desired behaviour is bribery, as they should be intrinsically “good.” Some parents prefer punishment to positive rewards on the basis of “spare the rod, spoil the child.” Others want to treat all family members equally, and see differential reinforcement as discriminatory. Still others become so angry with their children that they are unable to maintain the self-control necessary for behaviour modification techniques. Finally, carrying out a systematic program requires a harmonious relationship between the partners, exactly as in other areas of behaviour modification in which family relationships are the focus of treatment (e.g., McGovern, Stewart, and LoPiccolo 1975).

Both sets of arguments are well illustrated by reports of work in this area, the two most active and influential research groups being those led by Patterson and Alexander. The Patterson group has focused on parent training programs in general; Alexander and colleagues have in addition emphasized the relationship between therapists and client families — the “non-specific” factors which have often been neglected in behaviour therapy work.

The main findings of these two groups are as follows:

1.  The patterns of interactions in families with delinquent children differ from comparison families.

2.  Short-term behavioural interventions with the former families result in significant changes in family interaction patterns when compared with untreated controls.

3. These changes are related to reduced rates of recidivism for status offences, for follow-up periods of up to 18 months. Results for criminal offences are more variable, but Marlowe, Reid, Patterson, and Weinrott (1986) reported that, even if the Patterson method was not associated with a better outcome at follow-up than a conventional community treatment, it was significantly less expensive (presumably because there was less professional involvement). Moreover, improved family interactions were associated with a lowered frequency of offences, giving support to the basic thesis.

4. In behaviourally treated families, but not in controls, siblings of delinquent adolescents have significantly lower delinquency rates at three years follow-up. This suggests that the parents concerned had applied their newly acquired skills to the development of appropriate behaviours in those sibs, thereby pre-empting antisocial behaviours by them.

5. The better the relationship skills of the therapists the lower the dropout rate for families in behavioural treatment and the lower the recidivism rate for status offences.

6.  Parent-management training (PMT) makes major demands on parents — they must master educational materials, make systematic observations and respond appropriately to their children over long periods of time. At least one highly co-operative parent is needed, but may not be available. Indeed, according to McCauley (1982), PMT works best in modifying the behaviour of children in stable middle-class families. In contrast, Patterson, Cobb, and Ray (1973) comment that mothers raising families alone and in extreme poverty have great difficulty in learning and applying PMT.

7.  Berg, Hullin, and McGuire (1979) found that, when parents of truanting children were required to attend court for any subsequent truancy, both truancy and criminal offences were reduced, compared with a group supervised by probation officers and social workers, with no extra effort being required of the parents. This suggests that, even without PMT, if a child’s misdemeanours result in some penalty for the parents they will put more effort into child management. A study which looked at the effect on offending of a combination of PMT and various forms of penalty imposed on parents for child misdemeanours would be of interest.


Intervention:   institutions

Behaviour modification

These techniques have been adapted into the ‘Token Economy System’ (TES). The individual gains points or tokens for ‘good behaviour’, and loses them for ‘bad behaviour’. After a while, the tokens gained can be exchanged for desirable activities (like watching TV) or objects (like cigarettes).

The TES has been implemented in whole prisons in the USA (for example, ‘Cell Block Token Econ­omy’ Milan et al, 1974). In the UK, there has been limited use in ‘Young Offenders Institutions’ (for example, Cullen and Seddon, 1981).

At Camp Butner, North Carolina, aluminium tokens were used with the inmates. For example, fighting lead to ‘seclusion’, a charge of fifteen tokens, with inmates sent to their rooms to be quiet for 30 minutes. If they fulfilled this criterion, they received five tokens.

By 1980, some variant of the token economy was in operation in nearly all states in the USA (Burchard and Lane 1982). Initial enthusiasm was great because criminal behaviours, like behaviours in general, seemed to be acquired in the offenders past and present environment. Hence, with an appropriate use of learning principles, socially approved behaviours could be enhanced and criminal behaviours diminished. The institutional setting, because it is physically self-contained, was thought particularly appropriate for the learning approach. Contingencies could be precisely manipulated and, given careful management of the transition, there seemed good reason to expect considerable generalization of the results of treatment from the institution to the environment to which the prisoner would return.

The reality was very different from initial expectations. Several of the token economies were applied in a particularly unsatisfactory manner, the most notorious example being the START program at Springfield, Missouri, which relied heavily on negative reinforcement to shape inmate compliance. Amid much legal and professional criticism this was closed after two years (see Nietzel, 1979, for a full description). In all, Nietzel reviewed six prison-based token economies and concluded that there were no meaningful outcome data associated with any of them.

What went wrong? Burchard and Lane (1982) make some key points, including resistance to change by existing prison staff, inconsistent application of the behavioural techniques used, and failure to plan properly for generalization to the natural environment. It should be noted that well-conducted token economies (reviewed by Ayllon and Milan, 1979) do improve behaviours within institutions, such as general rule-keeping and interpersonal aggression. These are not minor gains, but the crucial target must remain criminal behaviour after release, on which Ayllon and Milan (1979) present no evidence.

Other problems include the creation of a mercenary attitude, or that some inmates who are unable to control their behaviour, receive no tokens and experience learned helplessness. This is the feeling that the individ­ual has no control over events, and they tend to be pas­sive to life’s challenges. These feelings can easily lead the offender into the old cycle of behaviour and recidivism. Other techniques based on behaviourist principles have been tried. For example, reparation programmes require the offender to make a personal apology and/or financial compensation to the victim. This is uncomfortable, and acts as punishment to stop future offending.


Intervention:  community-based programs

The severe difficulties of working in institutions, and the poor post-discharge results obtained have led to a major shift in emphasis in behavioural work to community-based programs, some residential, some not. In the latter, the young offender or potential offender continues to live in his/her own home. Some programs involve professionals, but increasingly professionals work through key figures in the natural environment, usually parents.


A residential program. Ostapiuk (1982) has described in detail the SHAPE program based in an inner-city area of Birmingham, in the UK Midlands. SHAPE is a multi-stage voluntary behavioural program for offenders aged 18-24.  In the first stage the participants live in a group and identify each other’s social and work-related needs.  In the second stage the participants live in pairs in a flat that they have decorated themselves.  They practice their new skills and work closely with employers and police.  In the third stage the participants live alone in their own apartment and continue to practice appropriate work-related behaviour as well as survival skills.  93 young men were referred between 1977 and 1979.  Staff accepted 80% of those referred and of these, half refused the conditions imposed upon them.  36 participants completed the course over six-months.  A follow-up at 6-18 months found that 22% had been reconvicted (which is encouraging).


A temporary change of environment.  By 1980 there had been more than 100 “adaptive wilderness” programmes in North America.  For one such programme no difference in recidivism was found after six months.


School-based programmes.  These are based on the assumption that educational progress can be expected to reduce initial offending or reoffending.  Preparation through responsive educational programmes (PREP) involved over 500 pupils in the Maryland school system.  Students were recommended to the programme because of falling school grades, truancy and police contacts.  Teachers were trained in methods of teaching social and educational skills and then the teachers taught the pupils.  The treated groups showed moderate short-term academic gains but this disappeared after a four-year follow-up.  No improvement was found in non-academic areas such as delinquency.


Probation programmes. Jesness (1975) reports a large-scale study.  Ninety probation officers received 40 hours of classroom training in behavioural principles.  The emphasis was placed on contingency contracting.  Each offender received 22 hours of consultation time.  400 male offenders, average age 16, were seen.  No significant difference was found between the treatment and control group with regard to offence rates.  However those probation is with a positive regard for their probation officer were less likely to be recidivists.



Employment related programmes.

1.      Shore and Massimo (1979). Twenty male school drop outs aged 15 to 17 with a long history of anti-social behaviour took part voluntarily and were randomly assigned to the programme or to an un-treated control group.  Job placement was based on the boys’ own interests and goals support and training concerning interpersonal problems, remedial education and job skills were given which continued during the employment.  After 15 years 90% of the controls but only 30% of the treated group were arrested at least once and 60% of the controls but only 10% of the treated group served at least one prison term.40% of the controls of 70% of the treated group were in successful and prolonged employment.  80% of both groups were married but whereas 70% of the marriage controls subsequently divorced this was true of only 25% of the treated group. However, a problem with this study is that the sample was rather small.

2.      The Kentfields Rehabilitation Project (Davidson and Robinson 1975).  One hundred and 17 male chronic delinquents took part.  95 were in the experimental group and 22 were controls.  On average the delinquents were aged 16, had been on probation for two and half years and had committed three known offences per year whilst on probation.  The programme lasted nine weeks and involved 6 hours a day on work projects and education.  After 18 months 17% of the experimental group were in penal institutions compared with 53% for the controls.  35% of the experimental group were employed and 25% were in some form of education. However a problem with this study is that we cannot be sure that the control group did not have a worse pre-trial criminal record.

3.      Mills and Walter (1979).  This study involved 53 young men and women (30 in the experimental group and 23 in the control group) with an average of 3.85 previous convictions.  Local employers were given information on behavioural principles and contingency contracts were drawn up between the youths and the researchers.  The treatment group were trained in interview skills and regular work routines prior to being placed in employment.  Support for the experimental group from the employers was gradually faded out.  One year later only 10% of the experimental group had been rearrested compared with 70% of the control group.  52% of the control group were in institutions.  However the groups were not matched for gender or prior work experience nor were participants randomly assigned.


Individual and group therapy

In the American Cambridge-Somerville Youth Study, started in 1939 young males in high delin­quency areas were assigned counsellors, given aca­demic teaching, medical and psychiatric care and other community assistance. The programme lasted up to eight years. Research finds that again this proj­ect does not reduce future offending. However, many problems were created by the counsellors imposing social values which produced conflict for the participants and their families (McCord, 1978).


Non-family community programmes.  O’Donnell, Chambers, and Ling (1973) set up an athletics program for which any boy from an economically poor community was eligible. The opportunity to play football and other games was offered as a reward for academic work and progress. During the four years of the project not one of the 200 participants had an official arrest record. Although no comparison group was set up without access to sports, the complete absence of arrests is much better than could have been expected in almost any random sample of disadvantaged youth. How­ever, it is possible that the fact that the parents of the boys had to apply for their son’s participation may have resulted in a group from concerned homes providing a good degree of social control which supplemented that provided by the project. And there are no data on the pre-program arrest record of the boys.

Such programs are replicable on a potentially very large scale indeed. Since the 1970’s the British juvenile courts have able been to “sentence to treatment” young offenders. That is, they are required to spend a period of time in a community based “intermediate treatment” (IT) program, typically an amalgam of education and leisure activities but with little planning or monitoring from entry to discharge. Preston (1982) described a study of one IT centre in which, having gained the cooperation of existing staff, she re-designed on behavioural lines both pre- and post-assessment procedures and the intervention program itself. Contracts were drawn up between staff members and program participants (boys and girls, aged 15—16) who continued to live in their own homes and attended the centre daily. Nearly 90 per cent of participants had both criminal histories and a lengthy record of truancy. The contracts specified short and long term-goals in the areas of education, work and social skills and self-management (e.g., time-keeping and personal hygiene). Points were earned for achieving targets and were exchanged daily for special activities such as ice-skating, swimming and snooker.  The overall program lasted eight months.

Preston and Carnegie (1989) compared the behaviourally managed group (n = 24) with a group which had attended the centre in the year prior to the introduction of the behavioural program (n = 26), in both cases after a follow-up of one year. The former was twice as likely to be employed as the latter (50 per cent versus 23 per cent), and only about half as likely to have been reconvicted (33 per cent versus 58 per cent). Pre-program offence records were the same for the two groups.


Social skills training.

One approach to help the offenders is by giving them social skills training (SST). It is hoped that this will give them an alternative to violence or crime. It can also be a way of improving the individual’s employment skills. Clients are first taught the skills by a combination of modelling, instruction, role play and rehearsal. They then attempt to re-enact these skills themselves in various arranged situations and receive feedback on their performance, the emphasis being on the use of social reinforcement such as praise. They are also given homework assignments which aim to help them practise and consolidate the skills they have learned in various situations, including real-life ones for those who are not incarcerated. Some programmes start by teaching certain non-verbal skills, known as micro-skills, such as eye contact, gesture and posture and then move on to all-round skills such as how to maintain a good conversation, how to interact with members of the opposite sex or how to negotiate (e.g. Hollin et al, 1986). The type of situation that might be practised is how to enter a room full of strangers, how to return faulty goods to a shop, how to politely decline getting involved in a drinking binge when you’ve already had enough, and how to say what you want to say without being embar­rassed and ‘tongue-tied’.


Results of SST

SST does change the social behaviour of partici­pants, but it may not necessarily prevent crime or re-offending behaviour. Spence and Marziller (1979) worked with five male young offenders who had few close friends, and were aggressive and rude. Videos of their behaviour showed they avoided eye contact, made inappropriate head movements, had excessive hand fiddlings, gave few verbal acknowledgements and a lack of feedback. Each participant received seven to ten training sessions using modelling, role-playing, videotaped feedback and social reinforce­ment. The results were increased eye contact and reduced hand fiddling, but no change otherwise. This is because of problems with maintaining and generalizing changes in skills, the poor motivation of some individuals to change, and even institutional resistance in some prisons (Howells, 1986).

Feedback from social skills training programmes is mixed. Goldstein (1986) reviewed 30 studies of SST used with aggressive or delinquent teenagers and found that various skills such as the use of appropriate eye contact and how to negotiate with a probation officer had been learned. However, Goldstein et al. (1989) found that only 15—20 per cent of trainees could use the skills they had learned during training in a more real-life situation. These researchers did, however, manage to increase this to 50 per cent by providing additional teaching.

Some other programmes report improved self-esteem and a feeling of greater control over life (a shift to a more internal locus of control) (Spence & Marziller, 1981). However, in this programme, individuals in a control group who received no training but spent an equal amount of time discussing their problems showed an equivalent improvement in self-esteem, so perhaps attention alone is the key to improvement. With respect to locus of control, Hollin et al. (1986) found no change in individuals to whom they gave SST.

Rates of recidivism (reconviction) are a more long-term measure of effec­tiveness of SST When the offending records of individuals in the Spence & Marziller programme were examined six months later, the SST group did have a lower level of conviction, but when asked about offending this group reported having committed more offences. There is evidence that both the appearance of a suspect and their general demeanour (bearing and expression), including the amount of respect shown, can affect whether an adolescent is detained by the police for questioning (Piliavin & Briar, 1964). In this case, it may be that the lads trained in SST were better at talking to the police and therefore less likely to be arrested.

There have been very few studies that have investigated the effects of SST on recidivism. Hollin (1990) has made two important points in this respect. Firstly, there simply is no research to show whether or not lack of social skills is associated with offending and secondly, in some SST programmes, there has been no evidence that the participants who received it were actually lacking in the skills in the first place.

However, there are two specific areas where social skills training might contribute to a diminution in the crime figures. The first concerns crime avoidance: Aiken, Stumphauzer, and Veloz (1977) showed that knowing how to resist peer pressures helped adolescents in the Los Angeles inner-city avoid criminal involvements. Second, there is evidence that both the appearance of a suspect and his general demeanour contribute to the decision to arrest or to caution a juvenile (Piliavin and Briar 1964). Werner et al. (1975) found that juveniles can be trained to display behaviours acceptable to the police. A report by Gross, Brigham, Hopper, and Bologna (1980) claims fewer arrests after a one year follow-up for boys trained in how to talk to policemen than for an untrained control group.

It seems reasonable to conclude that short-term changes in social behaviour can be achieved with SST, but it has yet to be shown that they are either long-lasting or that they can be generalised to real-life situa­tions. As Blackburn (1993) suggests, it may be a useful programme for offenders who have very serious social difficulties, but ‘the available evidence does not warrant the routine use of SST in correctional settings’ (page 357). Hollin (1990) also comments that SST alone is unlikely to be a cure for crime but it can be a powerful means of personal change.


Behavioural treatments for offence-related behaviours

Emory and Marholin (1977) identify three possible strategies for the direct modification of criminal behaviours; each of which has severe problems. The first is to decrease them by negative reinforcement. There are three difficulties here: criminal behaviours are of low fre­quency, are difficult to observe, and occur in conditions over which the behaviour modifier has very little control. The second option is, to change completely the environment of the offender. Since transportation ended doing so has involved only a temporary placement in an institution. While this removes the offender from potential targets of crime for the duration of his confine­ment, he returns eventually to the environment in which offences may occur, possibly even more likely to carry them out. The third possible strategy, and the most frequently used, is to strengthen behaviours which are incompatible with delinquency. The problem here is the term “incompatible.” Most of the behaviours intended to fall under this head, such as educational, job or social skills, are in fact quite compatible with criminal behaviour. Truly incompatible behaviours, such as refusing to take part in a crime or paying for goods in a shop rather than stealing them, have rarely been the targets of behavioural programs. The few published reports, almost all to do with stealing, usually from shops, are reviewed below, as are behaviour modification programs with sex offenders, an important exception to the general rule of a lack of behavioral work directly concerning the problem which led to arrest and conviction.

There are two single-case reports of adult shoplifting, both of which involved covert sensitization (Cautela 1967). Essentially, the client visualizes himself in the store, stealing, being caught and then sent to prison. Gauthier and Pellegrin (1982) used a careful single-case design with a woman with 11 previous convictions for shoplifting, which she engaged in several times per week. Having been trained in the method she practiced it on her own 10 times per day for five weeks and then whenever she felt tempted subsequently. At the final follow-up, 14 months later, she had reported stealing only once, when briefly out of the country. There were no officially recorded charges or convictions during the follow-up period. Another single-case report (Clover 1985) used a similar method and had a similar result.

Henderson (1983) reports using a range of behavioral techniques to alter clients’ responses to stimuli which were previously “signals for stealing.” A two-year follow-up of 27 children referred to child psycho­logical services for stealing (no information is provided on offense records) resulted in a recidivism rate of 20 per cent for the behaviorally treated group compared with 60 per cent for those who received “other treatments” and 70 per cent for an untreated group. As Henderson himself points out, there was no random allocation to behavioral or comparison groups and he carried out all treatments and did most of the follow-ups. Nevertheless, other than sex offenses, this seems the only example in the literature of direct behavioral intervention in the offense related behaviors of a sizeable group.




The treatment of sex offenders is controversial. Many would argue that this group of offenders, perhaps more than any other, deserves punishment and that the offer of treatment whilst they are incarcerated is pointless since they clearly either do not wish or are unable to change their behaviour. Moreover, they may opt for treatment quite cynically in order to facilitate earlier release or less intrusive supervision in the community. Perkins (1990), however, strongly argues the case for treating sex offenders on the following grounds.


     Treatment may enable sex offenders to modify their behaviour, whilst no treatment clearly will effect no change.

     Treatment may counteract the negative influences of imprisonment, e.g. confined contact with other sex offenders which might lead to a strength­ening of sexual deviation and long periods of incarceration which may further impair offenders’ social competence.

     Treatment may enable researchers to build up a body of knowledge which might help to reduce the possibility of sexual offences.

     Treatment may be entirely appropriate for those individuals whose offend­ing could be related to their own sexual victimisation as children.

     Where sexual abuse has occurred within a family setting and there is a possibility of reintegration, treatment will be essential.


Perkins also points out that criticisms of treatment programmes for sex offenders based on less than impressive recidivism rates often fail to take into account the number of further offences which have been prevented since these are inevitably ‘invisible’. We only usually know when offences have been committed, not when potential offences are successfully resisted. The provision of treatment programmes for sex offenders is limited currently 25 prisons run such programmes (attended by about 600 prisoners) and the Probation Service has established 90 programmes throughout the country, treating almost 2000 offenders each year.



What are sex offences?

Since ‘normal’ sexual activity is a social construction which varies according to time and cultural mores, it is perhaps not surprising that defining ‘abnormal’ sexual activity is problematic. For instance, acts of sadism between consent­ing adults are illegal in Britain.  Defining current legal categories of sexual offences is relatively simple attempted rape, rape, indecent assault, gross indecency, soliciting, unlawful intercourse, incest, procuration and indecent exposure but not all forms of sexual deviation are against the law.   Prins (1991) suggests that a useful classification of sexual deviation would include the following.


     Sexual activity which does not require a human partner, e.g. the use of animals (zoophilia) or objects (fetishism).

     Sexual activity which does not require a willing partner, e.g. rape, necrophilia, voyeurism or indecent exposure.

     Sexual activity which occurs under unusual conditions, e.g. with the very young or the very old, within the family or with excessive violence.

     Activity which is sexually motivated but presents in ‘masked form’, e.g. stealing female underwear, excessive interest in bodily secretions or pyro­mania (fire setting).


For many people the most obvious treatment for paedophiles would be castration and some would argue that recidivism rates following surgical castration are very low under 4% (Bradford, 1985), though in the studies cited most of the men were volunteers. Similarly, chemical castration, using synthetic steroids to reduce testosterone, has produced mixed results (Hucker et al., 1988). Whilst castration might have emotive appeal, its effects are by no means certain. Sexual satisfaction can be obtained without an erection and since many sex offenders do not wish to change their behaviour, they are unlikely to volun­teer for castration. Any move to impose surgical or chemical castration on convicted offenders would be strongly resisted by civil liberties groups.


The most common approach used in these programmes is cognitive-behavioural therapy which focuses on identifying sexual assault cycles, correcting distorted thinking patterns (for example denial and minimisation), controlling deviant fantasies, increasing empathy for victims, increasing social competencies, and developing relapse prevention skills.


Cognitive behavioural therapy (CBT)

The STEP programme (Beech et aI., 1998).

Fundamental to a treatment programme is a set of decisions about how, when, where, to and by whom therapy is to be delivered. The basic decisions for the prison Sex Offender Treatment Programme included the following:

. Where: in a limited number of prison establishments which could be resourced appropriately to deliver the treatment.

. Who: priority is given to offenders at the greatest risk of reoffending according to a formal risk-assessment procedure. Some characteristics such as mental illness, lack of English, very low IQ, suicide risk and severe personality disorder effectively debar the individual from the programme.

. When: at an appropriate stage in a sentence of two or more years provided that time is available to complete the programme.

. Whom: the approach is multi-disciplinary. That is, all sorts of staff besides psychologists may be involved such as prison officers, teachers and chaplains. Of course, they are given appropriate training in terms of knowledge of cognitive behavioural treatments and relevant skills in working with others in groups.

The therapeutic situation is based on structured group work. A group consists of eight offenders and two tutors/therapists. The treatment manual contains a structured series of cognitive behavioural activities and exercises that are explained and described in detail. It concentrates on the thought processes involved in offending as well as attempting to place limits on the behaviour on the offender. The core programme is designed to work on the motivation of the offenders to avoid reoffending and to develop personal skills that enable this. These latter skills are collectively known as relapse prevention. According to Beech et al. (1999), the programme consists of 20 blocks (treatment sessions) that cover a number of areas. These for the purposes of this description may be classified as cognitive modification and relapse prevention. The methods employed in group treatment of this sort are a mixture of methods familiar to those who have engaged in any type of group work, and matters much more specific to sex offending. Some of the techniques involved include the following:

. Brainstorming and group discussion: topics discussed by the group are often written down as a list on a board or flip chart. An individual often responds to this collection of ideas in terms of his experiences.

. Smaller or buzz groups: some activities are carried out by a pair of offenders, perhaps three or four. This sort of activity helps the offender develop communicative skills with others, assertiveness and a degree of empathy. The experience with the small group can then be combined with that of other groups in a 'plenary' seSSiOn.

. Role-playing: members of the group (and this may include the tutors or facilitators) may play out a situation. The rest of the group observe and respond to the role-play. The actual participants may also analyse their experiences.

. Focus on the individual: the work of one individual is subject to scrutiny and evaluation by the rest of the group.

. Videos: film is available which deals with various aspects of sex offending. In particular, there may be video available about the experiences of victims. Viewing this is then followed by individual response and group discussion.

. Homework: activities such as keeping a diary are carried out by members outside the group meeting itself. This is almost always written work.

The above are the main treatment methods employed. Substantial areas are covered using these methods including the following:

. Describing the offence: it is known that sex offenders tend to describe their offences in ways that are self-exculpatory. Often the offender will present himself almost as if he were the victim. Vagueness and being non-committal is characteristic of their responses. The following is a short extract from an interview (Howitt, 1995a, p. 95):

Interviewer: Did you kiss her on the breast?

Bennie: Maybe I did maybe I didn't. . . [when you are arrested] they try to use psychology on you, they make you say you did. . . so I am going to say I did.

Interviewer: . . . that's no use to me. . . . I don't want to know what they say, I want to know what it is. .

Bennie: . . . maybe I probably did. . .

At the end of the exchange, the reader may feel that they still do not know whether or not Bennie accepts that he kissed the girl's breasts sexually despite a clear challenge from the interviewer. Nevertheless, despite this, Bennie uses the phrase 'maybe I probably did. . .' which leaves him free psychologically to maintain his position that he said things because he, in fact, was the victim of pressure from other people. Therapists may describe this as a 'passive account' since it does not truly describe what the offender did. An active account would be much more direct. For example, in the above example, the offender might have said 'I encouraged the girl to roll about in front of the television with me. We were pretending to play at being animals. I pulled up her clothes and played at biting her stomach. Then I took it further and sucked her breast for a couple of minutes.' In order to encourage the active account, it is necessary for the offender to provide information about the following:

- How the offender actually planned the offence - it did not just happen.

- The offender's sexual or emotional preoccupation with the victim.

- That the offender was responsible for initiating all of the aspects of the abuse.

- That he took measures to try to prevent the victim from disclosing the abuse to others.

In terms of the content of treatment programmes, Prentky (1995) provides a useful summary of the target areas for treatment together with strategies for dealing with each of these areas:


   Correcting distorted thinking patterns

Most sex offenders justify their behaviour by the use of cognitive distortions, or irrational ideas, which minimise the degree of force used during the offence, the impact on their victims, their role in the offence, or the inappropriateness of their behaviour (see Hatch-Maillette et al., 2001). Since these cognitive distortions tend to be learned attitudes, supported by societal and cultural systems which perpetuate misogyny (hatred of women) and the exploitation of children, they have to be confronted and challenged in a systematic way known as cognitive restructuring. This involves providing accurate information about sexual abuse and its consequences, and helping to reveal the functional role of distorted attitudes, how they allow the offender to avoid responsibility and facing up to his actions. It also includes creating discomfort in the offender by emphasising the victim’s distressed response to abuse their shame, fear, pain and humiliation in an attempt to help the offender own’ his behaviour and its dreadful consequences rather than hiding behind convenient cognitive distortions.


   Controlling deviant fantasies

There is a clear link between deviant sexual fantasy and deviant sexual arousal which may lead to inappropriate and unacceptable behaviour. Sexual arousal is usually measured using an instrument called a plethysmograph which measures penile volume changes in response to a range of auditory and visual stimuli. This provides a baseline against which to assess changes in arousal which follow treatment. Once a pattern of deviant sexual arousal has been identified, for example to photographs of children, then behavioural tech­niques can be employed either to decrease deviant arousal or to increase more appropriate arousal. Those techniques designed to reduce deviant arousal include covert sensitisation, aversion and masturbatory satiation, whilst those which aim to shape the individual’s arousal pattern towards more appropriate targets include systematic desensitisation, fantasy modifica­tion and orgasmic reconditioning.


All these techniques rely on basic classical and operant conditioning principles and the most commonly used involve some element of aversion therapy, whereby an unpleasant stimulus is paired with the visual stimulus which pro­duces deviant sexual arousal. Early examples of unpleasant stimuli included brief electric shocks or nausea-inducing chemicals, though it is now more likely that noxious smells will be used. When covert sensitisation is used, the offender is told to imagine his deviant activity resulting in unpleasant conse­quences such as being arrested at work and humiliated or assaulted. ‘Assisted’ covert sensitisation involves pairing arousal with an imagined aversive event such as vomiting or having a tooth filled without anaesthetic. The technique of satiation involves the offender being told to masturbate to orgasm using a non-deviant fantasy (e.g. consensual sex with an adult) and then to continue masturbating after ejaculation whilst verbalising deviant fantasies. The idea is that the unpleasant experience will become associated with deviant fantasies and a consequent reduction’ of sexual arousal to those fantasies. This activity is undertaken in private but the offender is obliged to make audiotapes of his experiences which will be checked for authenticity by the therapists.


The most popular technique for increasing non-deviant arousal is orgasmic reconditioning, whereby the offender is instructed to masturbate to his pre­ferred deviant fantasy but must switch to a non-deviant fantasy at the point of orgasm, thus taking away the satisfaction of the deviant fantasy. The non-deviant fantasy is then progressively moved back in time towards the start of masturbation.  Fantasy modification involves gradually reducing the inappro­priate elements of a deviant fantasy and moving the offender toward a more acceptable fantasy which can be associated with sexual satisfaction.


   Increasing victim empathy

A lack of empathic concern for victims is a significant feature of all interpersonal violence. Thus, providing offenders with the opportunity to develop empathy may enable them to recognise the pain their victims will suffer and then desist from hurting them. This can be accomplished by exposing offenders to videos of victims describing their feelings and experiences following abuse, enabling them to visualise themselves in a similar situation of betrayal and violation, role play and role reversal taking the role of their victim, and writing letters of apology to their victim (which will not be sent). Many offenders will themselves have experienced victimization as children but may need help to develop more affective appreciation, in order to generalize these feelings towards their own victims. Pithers believes that developing empathy is crucial in treatment programmes for sex offenders:


If empathy can be established, significant effects may be observed in sexual arousal, cognitive distortions, intimacy within interpersonal relationships, realistic self-esteem, and motivation to change and maintain change. (1993, p.190)


Establishing empathy with victims is a vital part of early treatment because it significantly shifts perception so that denial of pain becomes impossible. The offender can then be assisted to resist the urge towards denial.


   Improving social competence

It is recognised that many paedophiles lack competence in their social and interpersonal skills, feeling awkward with adults and much more comfortable with children. Addressing this via social skills training, sex education, self-esteem enhancement, relaxation training, and assertiveness training may help such individuals improve their interpersonal relationships with adults, though they may need substantial rehearsal opportunities and positive feedback.


. Relapse prevention: relapse prevention (e.g. Marshall et al., 1992; Pithers et al., 1988) prepares the offender to deal with the feelings and experiences that he will have on release which are known to be progenitors of offending behaviour. These may be regarded as warning signs that must not go unheeded. It is known,for example, that sex offending patterns in some offenders are preceded by a negative mood state. Thus depression and anxiety might serve as danger signals for imminent offending. Similarly, the return of deviant fantasies may serve a similar function. There are other aspects of offending that a relapse prevention strategy would signal as dangerous: for example, moving into a job involving children, moving to a neighbourhood where there is a school nearby or just offering to babysit for a neighbour. For most people, these may be innocuous life-events; for the paedophile or child molester, they may be a precursor to the offending process.

 Antisocial personality

There would appear to be a link between the characteristic behaviours associat­ed with antisocial personality, in particular impulsivity, and sexual assault. This is particularly important when considering the possibility of relapse or recidi­vism following treatment and release. Attempts to address impulsivity include identifying the ‘assault cycle’, i.e. the chain of events and emotions which lead up to offending behaviour, so that the offender can develop strategies to avoid high-risk situations. Families can also be involved in this aspect of treatment so that they can identify predictors which might suggest that further offending behaviour is likely to occur and inform the appropriate authorities. 


Evaluation how do we know when treatment programmes are successful?

Some therapists will argue that any treatment of sex offenders is better than none since no form of treatment makes offenders more deviant (Laws, 1985), but evaluating the success of treatment programmes is particularly difficult because of under-reporting of offences (by victims and offenders) and the timescale necessary to be sure of effective rehabilitation. Dwyer and Myers (1990) report a ten-year follow-up of a particular group of sex offend­ers who had volunteered to undergo a particularly comprehensive treatment programme. Recidivism was under 4%, although over two-thirds of the offenders reported experiencing urges to reoffend. Abel et al. (1988) carried out a similar study and found a recidivism rate of 12% after one year. When comparing untreated sex offenders with treated sex offenders, Marshall et al. (1991) claim that 20—60% of untreated offenders reoffend in the five years following release from prison, whereas typically only 15% of treated offenders repeat their offences. They thus offer strong support for the continuation of treatment programmes for sex offenders, particularly those with a cognitive behavioural perspective.

In the UK, evaluation of sex offender programmes has taken place in relation to community programmes and prison programmes. Beckett, Beech, Fisher and Fordham (1994) used what has become known as the STEP method to evaluate community programmes (discussed above), using pre- and post-psychological measures involving offence-specific and personality assessment techniques, and found that 54% of the offenders in their sample had profiles after treatment which fell within a non-offending range. The same team concluded from their study of prison programmes that two-thirds of the offenders in their sample were successfully treated with regard to offending attitudes, and that longer programmes were more effective than shorter ones (Beech, Fisher & Beckett, 1999).


Overall, what treatment works?

It seems that most outcome studies find little suc­cess in reducing recidivism, whatever the technique used. But many studies have methodological weak­nesses or are too general in the approaches to recidi­vism.

Based on 231 studies, Martinson (1974) is best known for concluding that ‘nothing works’. But Thornton (1987) re-analysed the same data using a clear definition of recidivism and controls in the allocation of groups for the studies. On this basis, only 34 of the studies were acceptable in comparing psychological therapies to an untreated control group. Of the 34, 16 showed a significant benefit from treatment, 17 showed no difference and only one study found a significant disadvantage follow­ing treatment. This has led to the current position that ‘clinical intervention works with some offend­ers, some of the time’ (Hollin, 1989). This is known as ‘differential treatment’.

Recent research has involved meta-analysis (a sophisticated statistical technique) to compare other studies. For example, Losel (1995) summarized the meta-analysis of 500+ studies on the effect of cor­rectional treatment since the mid-1980s. The mean effect size was 0.10 (which means a 10% drop in recidivism due to treatment).

Overall, for treatments to reduce recidivism the programme needs to follow certain criteria.

·                     Application of a specific structure to a specific problem (for example, Behaviour Therapy for aggression).

·                     The staff need to model and reinforce anti-crimi­nal values.

·                     The staff need to be committed to the programme.

·                     Attitudes that encourage offending behaviour must be challenged.

·                     The programme must be matched to the offender.

·                     These should be an attempt to generalize the pro­gramme outside of the institution.


Hollin (1992) draws some conclusions about ther­apeutic programmes, as listed below.

·         Multi-modal structured treatments are effective. For example, Aggression Replacement Training (ART) uses three main approaches: structured learning training (including social skills training and social problem-solving), anger control train­ing, and moral education. In a New York study, re-arrest rates for those on ART were 15% com­pared to 43% for the control group (Goldstein et al, 1989). Lipsey’s (1992) meta-analysis of 400 studies of juvenile delinquency found a reduction in recidivism of around 20% for skill-oriented, CBT, and multi-modal programmes.

·         Behavioural-based approaches are much better than psychotherapy.

·         Community programmes and those involving the family are also effective.

On the other hand, there are threats to such pro­grammes, such as: ‘programme drift’ (a move from the original aims of the programme), ‘programme reversal’ (staff modelling inappropriate behaviour), and ‘programme non-compliance’ (the programme changing half-way through) (Hollin, 1995).


What happens when sex offenders are released?

Doubts about the effectiveness of treatment programmes for sex offenders have led to calls to publish the names of known offenders so that their new neigh­bours are aware of the potential dangers. Whilst the Sex Offenders Act 1997 has established a register of all convicted paedophiles in England and Wales and has required all on the register to inform the police of their addresses, access to this register is restricted and some newspapers have taken the step of ‘outing’ convicted sex offenders living in the community The results have been forced evictions, arson and severe beatings, sometimes involving innocent men.


The move towards releasing information on sex offenders stems from legisla­tion in America following a particularly horrifying crime. In 1994 Jesse Timmendequas sexually assaulted and murdered Megan Kanka, aged seven, in New Jersey He had previous convictions for sex offences but when he moved into Megan’s neighbourhood no one knew about his past. After Megan’s death her mother campaigned vigorously for the right to know if paedophiles move into a community and her efforts resulted in ‘Megan’s Law’ whereby this infor­mation has to be made public. Unfortunately, this legislation has not reduced the rate of child sex offences and critics of ‘Megan’s Law’ point out that not only does the majority of child sexual abuse occur in the home, but if paedophiles are hounded out of communities they may have to go underground and conceal their identities, making it more difficult for them to continue treatment and maybe even increasing the possibility of relapse.


A more extreme response to the potential for sex offenders to reoffend lies in the decision made by the American Supreme Court in June 1997 that repeat sex offenders can be detained indefinitely if a doctor reports a likelihood of recidivism. Whilst civil liberties groups have protested, this legislation has been seen as one which will be popular with the general public, though understandably less popular with the medical profession upon whose judge­ment the public will rely for the protection of children.


A programme for prisoners with a sentence of a year or less

In Britain between 1999 and 2002 the three main types of rehabilitation scheme have been- psychological 'offending behaviour programmes', drug treatment and basic skills education.  In 2001, 6,127 inmates completed offending behaviour programmes, more than 11 times as many as in 1994. That figure will have risen to 9,000 in 2002. Another 16,000 are being taught numeracy and literacy - the basic skills of more than two-thirds of prisoners are so poor they are automatically excluded from 94 per cent of jobs. Research shows that nothing succeeds in preventing recidivism more effectively than employment.   The Enhanced Thinking Skills course is currently being employed in British prisons, being aimed at prisoners on sentences of one year or less.  A typical example of such a course is described by David Rose in his Observer article (05-05-02).  The technique involves small groups of prisoners talking about their crimes and the thinking patterns they engaged in at the time; they are also asked to think about how they now feel towards the victim(s).  Through talking and role play the prisoners begin to understand the thought patterns that led to their crimes and discover ways of dealing with situations that potentially lead to criminal behaviour.  ETS in 2002 was used in 79 jails and a similar programme adopted from Canada at another 24. Peer-reviewed research by Caroline Friendship, a Prison Service psychologist, compares 670 inmates who went through these courses with 1,801 offenders matched by offence and social categories who did not attend a programme. All types of offender who had the treatment were significantly less likely to be reconvicted within two years. Among those judged 'medium-low risk', for example, only 18 per cent were reconvicted, against 32 per cent in the comparison group. The research concludes that prisoners who take the courses in 2002 can be expected to commit 21,000 fewer crimes.   One criticism of this research is that the follow-up period of just two-years is too soon as many of the inmates could drift back to crime after two years, but still the tentative results look promising.





Peter B. Ainsworth, 2000, Psychology and Crime, Longman, 0-582-41424-5

Kevin Brewer, 2000, Psychology and Crime, Heinemann, 0-435-80653-X

Diana Dwyer, 2001, Angles on Criminal Psychology, nelson thornes, 0-7487-5977-8

Philip Feldman, 1993, The Psychology of Crime, Cambridge, 0-521-33732-1

Julie Harrower, 1998, Applying Psychology to Crime, Hodder & Stoughton, 0-340-70556-6

Julie Harrower, 2001, Psychology in Practice – Crime, Hodder & Stoughton, 0-340-84497-3

Dennis Howitt (2006) Forensic and Criminal Psychology, Pearson Prentice Hall, ISBN 0-13-129758-9