Just Wondering if someone can look over, edit and give feedback
Introduction
Recidivism of Sex Offenders, in particular Child Sex Offenders (CSO), is an area of research that has received intense scrutiny in recent years as policy makers demand a 'solution' to the issue of child sexual abuse (Theodis & McMurran, 2006). Smallbone and Ransley (2005) state that while the number of sex offenders in prisons has grown, the incidence of child sexual abuse has declined, although median sentence lengths in prisons have increased by twenty percent. Lievore (2004) makes mention of the fact that the acknowledgement that CSO's are a distinct class of Sex Offenders has also seen research focussed on new areas. Even within themselves, CSO's are a varied group of offenders, with one percent of offenders having admitted to committing child sexual abuse (CSA) on more than one hundred victims, (Smallbone and Ransley, 2005) and with a number of different psychological definitions placed within the Diagnostic and Statistical Manual of Mental Disorders (Looman et al, 2005, Bahroo, 2003). Drake & Ward (2006) identify the tendency to treat all CSO's offence and relapse processes as the same, which has been a problem of many treatment methods. This begs the question, how do policy makers work to prevent recidivism amongst such a varied group of offenders?
This literature review intends to look at treatment methods used on CSO's, both traditionally and in the modern day. It also intends to discuss why CSO's engage in recidivism, the various methods used to predict recidivism and the economic costs and benefits of recidivism and treatment programs.
Treatment of Child Sex Offenders
Historically, a diverse range of methods have been used in the treatment of Child Sex Offenders. McConaghy (1997) states that nearly all methods developed have resulted in the 'mismanagement' of CSO's and Sex Offenders in general. Historical methods of treatment included incarceration, Behavioural therapy and interest modification therapy.
Queensland Currently uses four models of treatment for sex offenders (Lievore 2004). These are the Sex Offender Intervention Program (SOIP), The Preparation for Intervention Program (PIP) the Sexual Offender Treatment Program (SOTP) and the Indigenous Sex Offender Treatment Program (ISOTP). The programs in themselves are flexible, but have not been tailored in the same fashion that programs in other jurisdictions, such as New Zealand (Bakker et al, 1998) and Canada (Wilson et al, 2000). Queensland also utilizes extended detention orders and has recently adopted the use of electronic surveillance bracelets for Child sex offenders (Smallbone & Ransley, 2005, Affix Source here), both of which can be argued to be disintegrative shaming methods. (McAlinden, 2005). Warner, (2005) acknowledges consequences for the victim and questions legislation that includes preventative detention and extended supervision orders on the basis of lower recidivism rates and the principle of double punishment.
Recent treatments for CSO are based on cognitive theories and on the assumption that the offenders' have common disfunctionalities. Therefore they are treated using a manualised group-based approach where offenders' are classified and each resulting group receives the same treatment. This approach has had modest success in reducing recidivism (Drake & Ward 2006). It is argued that this is an insufficient way to treat all offenders' and suggest a formulation based method which takes into account their individual differences. It is suggested the cognitive approaches in the group based method ignore dysfunction in belief systems and social attitudes and that each offender takes a different pathway to offend and re-offend. These pathways show where formulation based treatment could be effective. For example, a case formulation of an offender includes developmental factors that could influence a person to sexually offend (being a victim themselves). It is argued that this treatment is especially useful if the offender has a variety of problems or has a little researched problem, such as female offenders, or where previous treatment has failed. The formulation based method could give therapists more precision in matching treatment to offender. It is recommended that the two treatments, manual and formulated be combined to better serve and understand individual differences in offenders. (Drake and Ward 2006)
Although knowledge of risk factors to sexual recidivism is necessary to guide treatment programs, it is not sufficient. More information regarding the relationship between specific risk factors and sexual reoffending both within and across individuals is needed to help determine the best methods of intervention. One way to accomplish this is by testing for mediators and moderators of change. Mediators are those processes through which change occurs, while moderators are those characteristics that affect the extent to which changes occur (Baron & Kenny, 1986). In order to create more effective treatments, we need to determine the mechanisms of change that produce reductions in sexual recidivism. Understanding why treatment works, can help to maximize treatment effects by ensuring that the critical ingredients of treatment are included, and can help to identify those variables upon which the effectiveness of a treatment might depend (Kazdin & Nock, 2003).
Theoretical Development
Recently, a number of theories have been developed about certain aspects of Child Sex Offending and CSO's in general. Ward (2003) outlines his recent theory as he thinks it represents theory built on the work of Finkelhor, Precondition model (1984), Marshall and Barbaree, Integrated theory (1990), Hall & Hirschman, Quadripartite model (1992) - each of these has its weakness to provide a satisfactory explanation of child sexual abuse.
Toman, (2003) asks whether deniers and resisters can be successfully treated and the problems for CSO when compelled into treatment for release orders (like D'arcy). If it is not a behaviour of choice is it changeable? It is suggested that if we can learn from homosexual experiments that if an individual can't be responsible for their patterns of arousal they can be responsible for their sexual behaviour. Nicholson (1997) asks whether it is medical or criminal and questions the retributive approach that can further harm the victim and offender.
According to Vandier (2006) those who exhibit aggression during their childhood, escalating into delinquency and violence during their adolescence and adulthood, are considered to fall into the life-course persistent offender category. Those, however, who engage in violent acts during their adolescence, are only known as adolescence-limited offenders. Subsequent research has indicated the existence of a group of offenders who begin offending during their adolescence and continue offending into their adulthood (Kempf-Leonard, 1988; Nagin, Farrington, & Moffitt, 1995). Furthermore, research has supported the notion that those who begin offending early (early-starters) are more likely to offend more seriously and persistently (Krohn, Thornberry, Rivera, & Le Blanc, 2001).
Polaschek, (2003) argues that situations such as the offender's emotional state (feeling depressed) or their environmental surroundings (living near a children's playground) can influence whether or not they re-offend. Situations such as these can threaten the individual's sense of self-control and leads them into a lapse. Pithers et al. (1983), defines this lapse as an indication that an offence may occur. This can include fantasising about offensive sexual behaviour with children or purchasing pornography. This point of the lapse is where the short term benefits of sexual offending outweigh the consequences, as the immediate gratification of the offence is too great. This experience of immediate gratification provides the momentum for a lapse to slide into a relapse where the first sexual offence leads to the offender re-offending.
Feelgood et al. (2005), argues that sex offenders use violent sexual acts as a way of coping through experiences such as rejection or loneliness. The sexual act giving them a sense of control and power in their lives and this may be an explanation as to why they continue to offend. Edwards (2004) found that those who were successful therapy candidates were more likely to admit their abusive activities, were lacking in seriously distorted sexual beliefs and had good sexual knowledge (social sexual desirability).
Recidivism
Recidivism can be defined as lapsing into previous patterns of criminal behaviour (Maltz, 1984). By this definition, recidivism does not only refer to behaviour that breaks the law, but can include any conduct that is indicative of previous offence patterns. Falshaw et al. (2003), argues that recidivism includes any offence-related behaviours. A specific example of this may be a convicted child molester loitering outside a primary school. Whilst this behaviour is not against the law, it mirrors previous patterns of offending behaviour displayed by that offender. Recidivism is the broadest definition and it incorporates both reconviction and reoffending.
Serin et al (2001) has identified four types of recidivists: termination or suspension, non-violent, violent and sexual. Termination/suspension recidivism involves returning to prison for violating conditions of release. Non-violent recidivism is defined as a new conviction that was neither violent nor sexual. Violent recidivism involves a conviction for a new nonsexual offence committed against a person and sexual recidivism refers to a new conviction for a sexual offence. This literature review will investigate sexual recidivism in relation to child sex offenders.
A major issue amongst researchers and policy makers is the development of a working definition of recidivism in relation to Child Sex Offenders (Lievore 2005). Police statistics in Australia are typically skewed by the definition of recidivism typically including child sex offenders who have committed a crime unrelated to that conviction being counted as recidivists (Lievore, 2004). In the Relapse prevention model, a variant of which is extremely popular for use amongst child sex offenders in Australian programs and a key component of Queensland's Sex Offender Treatment Program (SOTP), considers thought about the negative action as a form of recidivism (Laws, 2003). Psychiatric and psychological perspectives on the issue tend to focus on both thinking about and commission of the act, and generally focus (McConaghy, 1997). This confusion has aided to uncertainty in the understanding of recidivism amongst child sex offenders, and is an issue in need of clarification.
Smallbone and Wortley, (2005 in Smallbone and Ransley, 2005) found in examining the unofficial extent of offending found that half of the CSO had only offended against one child, 10% said they had offended against more than 10 children and less than 1% (of 180) had offended against more than 100 children. Andrews & Bonta (1998 in Ward, 2003) found there is a subgroup who commits a wide variety of offences.
Hanson and Bussiere (1998, in Smallbone and Ransley, 2005) found that the average sexual recidivism rate was 13% within five years. Non-familial offenders were more than twice as likely to be convicted for further sexual offences (19% against 8% over 5 years). Over a longer term, 20 years at risk, Terence (2003 in Smallbone and Ransley, 2005) estimated the average sexual recidivism rate to be around 35 %. Leclerc & Tremblay, (2007) found that adolescent CSO's, even though they may have limited rationality, still attempt to minimise their risk of apprehension. This is consistent with the psychiatric opinion of Bahroo (2003), who states that outside of a core of obsessive, fixated offenders, there is a large number of non-obsessive, opportunistic offenders who are not deterred, due to the low likelihood of conviction, and indeed, reconviction. If the offender finds a successful method they will use it again. This may give credence to the hypothesis of Rational Choice Theorists, who state that crime involves the attainment of a goal that is instrument in achieving an overall outcome (Leclerc & Tremblay, 2007)
Hanson and Morton (2005) indicate through a meta-analysis of 82 recidivism studies (1,620 findings from 29,450 child sexual offenders) identified deviant sexual preferences and antisocial orientation as the major predictors of sexual recidivism for both adult and adolescent sexual offenders. Antisocial orientation was the major predictor of violent recidivism and general (any) recidivism. The review also identified some dynamic risk factors that have the potential of being useful treatment targets (e.g., sexual preoccupations, general self-regulation problems). Many of the variables commonly addressed in sex offender treatment programs (e.g., psychological distress, denial of sex crime, victim empathy, stated motivation for treatment) had little or no relationship with sexual or violent recidivism.
One of the problems in recidivism research on sexually abusive youth is the lack of a clear and consistently used definition of the term 'recidivism'. This can be firstly illustrated by Kahn and Lafond (1988) who had a sample of over 350 adolescent sexual offenders, aged 12 to 18 years (mean of 14.5). The offenders had attended a state juvenile correction treatment facility in Washington for sentences ranging from between 2 months and 4 years. Follow-up ranged between a few weeks and 6 years after leaving the residential facility and the authors stated the results indicated that approximately 9% had sexually re-offended while 8% had committed new non-sexual offences
Predicting Recidivism
Smallbone and Ransley, (2005) state that risk prediction research indicates that average recidivism rates are lower than assumed, and wide variations in recidivism patterns have been detected (different types of CSO). At present risk prediction methods have high rates of false-positive and false-negative predictions. They believe that actuarial risk prediction scales could give important information to psychological and psychiatric risk assessment as long as their limitations are understood by the courts. The actuarial prediction methods have been shown to carry more weight than clinical prediction that rely on professional judgment and has proven to be only somewhat better than chance
A commonly used device in predicting recidivism amongst child sex offenders in recent years has been the STATIC 99 Test (Friendship & Beech, 2005, Bahroo, 2003). This test combines the Rapid Risk of Sex Offence Recidivism (RRASOR) Scale and the Static, Anchored, Clinical Judgement (SACL) scale, with a sample of 531 released sexual offenders from the English and Welsh prison system to test its validity. There are 10 items: prior sex offences, number of prior sentencing occasions, nonsexual violence, male victims, single, committed non-contact sex offences, unrelated victims, stranger victims and finally, the offender's age. Despite a number of statistical measurement issues, the STATIC-99 test has been used extensively in clinical studies. In Western Australia, the Violent Offender Treatment Program Risk Assessment Scale (VOTPRAS) found that recidivism differed according to cultural background and whether or not the offender is regarded as violent or nonviolent, indicating that treatment methods in operation fail to correctly treat CSO's if they aren't tailored towards their type of offending. (Allan, Dawson & Allan, 2006).
Looman et al (2005) studied how CSO's high on the Psycopathy Checklist-Revised test (PCL-R) responded to treatment and the likelihood of recidivism. They found, outside of a group of High PCL-R offenders who appeared to be highly receptive to treatment and motivated, that those regarded as more psychopathic were more likely to engage in recidivous behaviour than those how scored low on the PCL-R. Serin, Mailloux and Malcolm (2001) found that deviant sexual arousal, when related to CSO's with higher psychopathy scores, was a strong indicator in prediction of recidivism.
It is notable that the 3-Predictor model is entirely based on dynamic risk factors. This is in sharp contrast with the Hanson and Bussiere (1998) finding, when they reviewed 61 sexual offending studies, that long-term recidivism was best predicted by static factors. However, Hanson and Harris (2000b) found that even after controlling for pre-existing differences in static risk factors, dynamic factors continued to be strongly associated with recidivism.
In assessing risk of recidivism in sexual offenders, a number of static and dynamic factors are considered. A static risk factor (such as age, developmental history or offence history) is something that is useful for evaluating long-term risk, but because it is historical in nature it cannot be used to assess changes in levels of risk over time (Craig et al, 2005). However, in addition to static risk, dynamic risk factors have also been demonstrated to contribute to risk prediction for sexual offenders in England and Wales (Friendship et al, 2005). Dynamic risk factors refer to factors that are changeable and open to intervention such as sexual preferences, personality factors, mood and intoxication (Craissati, 2003). Hanson and Harris (1998) distinguish between dynamic factors that are stable and acute.
Stable dynamic risk factors (factors that are expected to persist for months or years) are relatively persistent characteristics of the offender, which are subject to change such as levels of responsibility, cognitive distortions and sexual arousal. Acute dynamic factors (factors that may last for minutes or days) are rapidly changing factors such as substance misuse, isolation and negative emotional states, the presence of which increase risk. Factors that predicted general recidivism were not the same as those associated with sexual recidivism. Previous sex offences, poor social skills, male victims, and two or more victims in index offence were all risk factors associated with sexual recidivism. Early conduct disorder, previous convictions, psychopathy, and the use of death threats or weapons at the index sex offence were predictors for general criminality (Hanson et al, 1998).
Recommendations
Mercado (2005) looks at the development of recent statutory schemes, in both the United States and Australia, with its aim to keep the most dangerous sex offenders detained beyond the expiration of their prison sentence.. More recently, the Australian High Court in Attorney-General (Qld) v. Fardon (2004) similarly upheld the constitutionality of Queensland's Dangerous Prisoners (Sexual Offenders) Act (2003), which allows for the post-sentence preventive detention of sex offenders deemed to be at high risk of serious sexual recidivism
Of concern in Australia is that there has been no published research on the validity of the various sexual risk assessment measures. This is less of a concern for structured professional judgment measures, such as the SVR-20 that do not provide probability data. However, for actuarial measures that provide specific probability estimates, we cannot rely on the probability estimates since they are necessarily defined by the population from which the risk assessment measure was developed or validated. Given the differences across samples of offenders, and other cultural and population differences, we simply cannot determine the extent to which the actual measures would differ in Australian samples.
Moulden et al (2005) argues that positive approaches to sexual offender therapy are rooted in the belief that individuals working toward positive changes will be more successful and maintain those changes longer than those working toward lives based on avoidance.
Thanks
Christian
Introduction
Recidivism of Sex Offenders, in particular Child Sex Offenders (CSO), is an area of research that has received intense scrutiny in recent years as policy makers demand a 'solution' to the issue of child sexual abuse (Theodis & McMurran, 2006). Smallbone and Ransley (2005) state that while the number of sex offenders in prisons has grown, the incidence of child sexual abuse has declined, although median sentence lengths in prisons have increased by twenty percent. Lievore (2004) makes mention of the fact that the acknowledgement that CSO's are a distinct class of Sex Offenders has also seen research focussed on new areas. Even within themselves, CSO's are a varied group of offenders, with one percent of offenders having admitted to committing child sexual abuse (CSA) on more than one hundred victims, (Smallbone and Ransley, 2005) and with a number of different psychological definitions placed within the Diagnostic and Statistical Manual of Mental Disorders (Looman et al, 2005, Bahroo, 2003). Drake & Ward (2006) identify the tendency to treat all CSO's offence and relapse processes as the same, which has been a problem of many treatment methods. This begs the question, how do policy makers work to prevent recidivism amongst such a varied group of offenders?
This literature review intends to look at treatment methods used on CSO's, both traditionally and in the modern day. It also intends to discuss why CSO's engage in recidivism, the various methods used to predict recidivism and the economic costs and benefits of recidivism and treatment programs.
Treatment of Child Sex Offenders
Historically, a diverse range of methods have been used in the treatment of Child Sex Offenders. McConaghy (1997) states that nearly all methods developed have resulted in the 'mismanagement' of CSO's and Sex Offenders in general. Historical methods of treatment included incarceration, Behavioural therapy and interest modification therapy.
Queensland Currently uses four models of treatment for sex offenders (Lievore 2004). These are the Sex Offender Intervention Program (SOIP), The Preparation for Intervention Program (PIP) the Sexual Offender Treatment Program (SOTP) and the Indigenous Sex Offender Treatment Program (ISOTP). The programs in themselves are flexible, but have not been tailored in the same fashion that programs in other jurisdictions, such as New Zealand (Bakker et al, 1998) and Canada (Wilson et al, 2000). Queensland also utilizes extended detention orders and has recently adopted the use of electronic surveillance bracelets for Child sex offenders (Smallbone & Ransley, 2005, Affix Source here), both of which can be argued to be disintegrative shaming methods. (McAlinden, 2005). Warner, (2005) acknowledges consequences for the victim and questions legislation that includes preventative detention and extended supervision orders on the basis of lower recidivism rates and the principle of double punishment.
Recent treatments for CSO are based on cognitive theories and on the assumption that the offenders' have common disfunctionalities. Therefore they are treated using a manualised group-based approach where offenders' are classified and each resulting group receives the same treatment. This approach has had modest success in reducing recidivism (Drake & Ward 2006). It is argued that this is an insufficient way to treat all offenders' and suggest a formulation based method which takes into account their individual differences. It is suggested the cognitive approaches in the group based method ignore dysfunction in belief systems and social attitudes and that each offender takes a different pathway to offend and re-offend. These pathways show where formulation based treatment could be effective. For example, a case formulation of an offender includes developmental factors that could influence a person to sexually offend (being a victim themselves). It is argued that this treatment is especially useful if the offender has a variety of problems or has a little researched problem, such as female offenders, or where previous treatment has failed. The formulation based method could give therapists more precision in matching treatment to offender. It is recommended that the two treatments, manual and formulated be combined to better serve and understand individual differences in offenders. (Drake and Ward 2006)
Although knowledge of risk factors to sexual recidivism is necessary to guide treatment programs, it is not sufficient. More information regarding the relationship between specific risk factors and sexual reoffending both within and across individuals is needed to help determine the best methods of intervention. One way to accomplish this is by testing for mediators and moderators of change. Mediators are those processes through which change occurs, while moderators are those characteristics that affect the extent to which changes occur (Baron & Kenny, 1986). In order to create more effective treatments, we need to determine the mechanisms of change that produce reductions in sexual recidivism. Understanding why treatment works, can help to maximize treatment effects by ensuring that the critical ingredients of treatment are included, and can help to identify those variables upon which the effectiveness of a treatment might depend (Kazdin & Nock, 2003).
Theoretical Development
Recently, a number of theories have been developed about certain aspects of Child Sex Offending and CSO's in general. Ward (2003) outlines his recent theory as he thinks it represents theory built on the work of Finkelhor, Precondition model (1984), Marshall and Barbaree, Integrated theory (1990), Hall & Hirschman, Quadripartite model (1992) - each of these has its weakness to provide a satisfactory explanation of child sexual abuse.
Toman, (2003) asks whether deniers and resisters can be successfully treated and the problems for CSO when compelled into treatment for release orders (like D'arcy). If it is not a behaviour of choice is it changeable? It is suggested that if we can learn from homosexual experiments that if an individual can't be responsible for their patterns of arousal they can be responsible for their sexual behaviour. Nicholson (1997) asks whether it is medical or criminal and questions the retributive approach that can further harm the victim and offender.
According to Vandier (2006) those who exhibit aggression during their childhood, escalating into delinquency and violence during their adolescence and adulthood, are considered to fall into the life-course persistent offender category. Those, however, who engage in violent acts during their adolescence, are only known as adolescence-limited offenders. Subsequent research has indicated the existence of a group of offenders who begin offending during their adolescence and continue offending into their adulthood (Kempf-Leonard, 1988; Nagin, Farrington, & Moffitt, 1995). Furthermore, research has supported the notion that those who begin offending early (early-starters) are more likely to offend more seriously and persistently (Krohn, Thornberry, Rivera, & Le Blanc, 2001).
Polaschek, (2003) argues that situations such as the offender's emotional state (feeling depressed) or their environmental surroundings (living near a children's playground) can influence whether or not they re-offend. Situations such as these can threaten the individual's sense of self-control and leads them into a lapse. Pithers et al. (1983), defines this lapse as an indication that an offence may occur. This can include fantasising about offensive sexual behaviour with children or purchasing pornography. This point of the lapse is where the short term benefits of sexual offending outweigh the consequences, as the immediate gratification of the offence is too great. This experience of immediate gratification provides the momentum for a lapse to slide into a relapse where the first sexual offence leads to the offender re-offending.
Feelgood et al. (2005), argues that sex offenders use violent sexual acts as a way of coping through experiences such as rejection or loneliness. The sexual act giving them a sense of control and power in their lives and this may be an explanation as to why they continue to offend. Edwards (2004) found that those who were successful therapy candidates were more likely to admit their abusive activities, were lacking in seriously distorted sexual beliefs and had good sexual knowledge (social sexual desirability).
Recidivism
Recidivism can be defined as lapsing into previous patterns of criminal behaviour (Maltz, 1984). By this definition, recidivism does not only refer to behaviour that breaks the law, but can include any conduct that is indicative of previous offence patterns. Falshaw et al. (2003), argues that recidivism includes any offence-related behaviours. A specific example of this may be a convicted child molester loitering outside a primary school. Whilst this behaviour is not against the law, it mirrors previous patterns of offending behaviour displayed by that offender. Recidivism is the broadest definition and it incorporates both reconviction and reoffending.
Serin et al (2001) has identified four types of recidivists: termination or suspension, non-violent, violent and sexual. Termination/suspension recidivism involves returning to prison for violating conditions of release. Non-violent recidivism is defined as a new conviction that was neither violent nor sexual. Violent recidivism involves a conviction for a new nonsexual offence committed against a person and sexual recidivism refers to a new conviction for a sexual offence. This literature review will investigate sexual recidivism in relation to child sex offenders.
A major issue amongst researchers and policy makers is the development of a working definition of recidivism in relation to Child Sex Offenders (Lievore 2005). Police statistics in Australia are typically skewed by the definition of recidivism typically including child sex offenders who have committed a crime unrelated to that conviction being counted as recidivists (Lievore, 2004). In the Relapse prevention model, a variant of which is extremely popular for use amongst child sex offenders in Australian programs and a key component of Queensland's Sex Offender Treatment Program (SOTP), considers thought about the negative action as a form of recidivism (Laws, 2003). Psychiatric and psychological perspectives on the issue tend to focus on both thinking about and commission of the act, and generally focus (McConaghy, 1997). This confusion has aided to uncertainty in the understanding of recidivism amongst child sex offenders, and is an issue in need of clarification.
Smallbone and Wortley, (2005 in Smallbone and Ransley, 2005) found in examining the unofficial extent of offending found that half of the CSO had only offended against one child, 10% said they had offended against more than 10 children and less than 1% (of 180) had offended against more than 100 children. Andrews & Bonta (1998 in Ward, 2003) found there is a subgroup who commits a wide variety of offences.
Hanson and Bussiere (1998, in Smallbone and Ransley, 2005) found that the average sexual recidivism rate was 13% within five years. Non-familial offenders were more than twice as likely to be convicted for further sexual offences (19% against 8% over 5 years). Over a longer term, 20 years at risk, Terence (2003 in Smallbone and Ransley, 2005) estimated the average sexual recidivism rate to be around 35 %. Leclerc & Tremblay, (2007) found that adolescent CSO's, even though they may have limited rationality, still attempt to minimise their risk of apprehension. This is consistent with the psychiatric opinion of Bahroo (2003), who states that outside of a core of obsessive, fixated offenders, there is a large number of non-obsessive, opportunistic offenders who are not deterred, due to the low likelihood of conviction, and indeed, reconviction. If the offender finds a successful method they will use it again. This may give credence to the hypothesis of Rational Choice Theorists, who state that crime involves the attainment of a goal that is instrument in achieving an overall outcome (Leclerc & Tremblay, 2007)
Hanson and Morton (2005) indicate through a meta-analysis of 82 recidivism studies (1,620 findings from 29,450 child sexual offenders) identified deviant sexual preferences and antisocial orientation as the major predictors of sexual recidivism for both adult and adolescent sexual offenders. Antisocial orientation was the major predictor of violent recidivism and general (any) recidivism. The review also identified some dynamic risk factors that have the potential of being useful treatment targets (e.g., sexual preoccupations, general self-regulation problems). Many of the variables commonly addressed in sex offender treatment programs (e.g., psychological distress, denial of sex crime, victim empathy, stated motivation for treatment) had little or no relationship with sexual or violent recidivism.
One of the problems in recidivism research on sexually abusive youth is the lack of a clear and consistently used definition of the term 'recidivism'. This can be firstly illustrated by Kahn and Lafond (1988) who had a sample of over 350 adolescent sexual offenders, aged 12 to 18 years (mean of 14.5). The offenders had attended a state juvenile correction treatment facility in Washington for sentences ranging from between 2 months and 4 years. Follow-up ranged between a few weeks and 6 years after leaving the residential facility and the authors stated the results indicated that approximately 9% had sexually re-offended while 8% had committed new non-sexual offences
Predicting Recidivism
Smallbone and Ransley, (2005) state that risk prediction research indicates that average recidivism rates are lower than assumed, and wide variations in recidivism patterns have been detected (different types of CSO). At present risk prediction methods have high rates of false-positive and false-negative predictions. They believe that actuarial risk prediction scales could give important information to psychological and psychiatric risk assessment as long as their limitations are understood by the courts. The actuarial prediction methods have been shown to carry more weight than clinical prediction that rely on professional judgment and has proven to be only somewhat better than chance
A commonly used device in predicting recidivism amongst child sex offenders in recent years has been the STATIC 99 Test (Friendship & Beech, 2005, Bahroo, 2003). This test combines the Rapid Risk of Sex Offence Recidivism (RRASOR) Scale and the Static, Anchored, Clinical Judgement (SACL) scale, with a sample of 531 released sexual offenders from the English and Welsh prison system to test its validity. There are 10 items: prior sex offences, number of prior sentencing occasions, nonsexual violence, male victims, single, committed non-contact sex offences, unrelated victims, stranger victims and finally, the offender's age. Despite a number of statistical measurement issues, the STATIC-99 test has been used extensively in clinical studies. In Western Australia, the Violent Offender Treatment Program Risk Assessment Scale (VOTPRAS) found that recidivism differed according to cultural background and whether or not the offender is regarded as violent or nonviolent, indicating that treatment methods in operation fail to correctly treat CSO's if they aren't tailored towards their type of offending. (Allan, Dawson & Allan, 2006).
Looman et al (2005) studied how CSO's high on the Psycopathy Checklist-Revised test (PCL-R) responded to treatment and the likelihood of recidivism. They found, outside of a group of High PCL-R offenders who appeared to be highly receptive to treatment and motivated, that those regarded as more psychopathic were more likely to engage in recidivous behaviour than those how scored low on the PCL-R. Serin, Mailloux and Malcolm (2001) found that deviant sexual arousal, when related to CSO's with higher psychopathy scores, was a strong indicator in prediction of recidivism.
It is notable that the 3-Predictor model is entirely based on dynamic risk factors. This is in sharp contrast with the Hanson and Bussiere (1998) finding, when they reviewed 61 sexual offending studies, that long-term recidivism was best predicted by static factors. However, Hanson and Harris (2000b) found that even after controlling for pre-existing differences in static risk factors, dynamic factors continued to be strongly associated with recidivism.
In assessing risk of recidivism in sexual offenders, a number of static and dynamic factors are considered. A static risk factor (such as age, developmental history or offence history) is something that is useful for evaluating long-term risk, but because it is historical in nature it cannot be used to assess changes in levels of risk over time (Craig et al, 2005). However, in addition to static risk, dynamic risk factors have also been demonstrated to contribute to risk prediction for sexual offenders in England and Wales (Friendship et al, 2005). Dynamic risk factors refer to factors that are changeable and open to intervention such as sexual preferences, personality factors, mood and intoxication (Craissati, 2003). Hanson and Harris (1998) distinguish between dynamic factors that are stable and acute.
Stable dynamic risk factors (factors that are expected to persist for months or years) are relatively persistent characteristics of the offender, which are subject to change such as levels of responsibility, cognitive distortions and sexual arousal. Acute dynamic factors (factors that may last for minutes or days) are rapidly changing factors such as substance misuse, isolation and negative emotional states, the presence of which increase risk. Factors that predicted general recidivism were not the same as those associated with sexual recidivism. Previous sex offences, poor social skills, male victims, and two or more victims in index offence were all risk factors associated with sexual recidivism. Early conduct disorder, previous convictions, psychopathy, and the use of death threats or weapons at the index sex offence were predictors for general criminality (Hanson et al, 1998).
Recommendations
Mercado (2005) looks at the development of recent statutory schemes, in both the United States and Australia, with its aim to keep the most dangerous sex offenders detained beyond the expiration of their prison sentence.. More recently, the Australian High Court in Attorney-General (Qld) v. Fardon (2004) similarly upheld the constitutionality of Queensland's Dangerous Prisoners (Sexual Offenders) Act (2003), which allows for the post-sentence preventive detention of sex offenders deemed to be at high risk of serious sexual recidivism
Of concern in Australia is that there has been no published research on the validity of the various sexual risk assessment measures. This is less of a concern for structured professional judgment measures, such as the SVR-20 that do not provide probability data. However, for actuarial measures that provide specific probability estimates, we cannot rely on the probability estimates since they are necessarily defined by the population from which the risk assessment measure was developed or validated. Given the differences across samples of offenders, and other cultural and population differences, we simply cannot determine the extent to which the actual measures would differ in Australian samples.
Moulden et al (2005) argues that positive approaches to sexual offender therapy are rooted in the belief that individuals working toward positive changes will be more successful and maintain those changes longer than those working toward lives based on avoidance.
Thanks
Christian
