1. Introduction

Sexual violence in Australia is more widespread than criminal justice statistics indicate. This is primarily because sexual violence often goes unreported (Australian Bureau of Statistics [ABS], 1996, 2002a; Mouzos & Makkai, 2004), due to the majority of sex offenders being known to their victims. Common misconceptions locate sex offenders as 'deviant predators', when in fact most men who perpetrate sexual violence are connected to their victims as family members, partners or friends (ABS, 1996, 2004a), and thus operate under a façade of 'normal', 'caring' or 'mutual' relationships. This misconception is underpinned by societal structures and attitudes that presume all sexual violence is committed by 'deviant' men unknown to their victims. These factors limit the focus, conceptualisation and approach to social policies and programs aimed at stopping sexual violence and the treatment of sex offenders. This paper outlines the underpinning theoretical perspectives and prevalence of sexual violence, and what is known about sex offenders, responses to them and their treatment. The authors take the position that sexual violence is a gendered phenomenon, requiring a comprehensive application of contemporary theory, across relevant sectors, that ethically accounts for the experience of victims who are criminally violated. The focus of this paper is male sexual violence against women and children. In instances where this is not the case, it will be specified.

The paper provides an overview of literature concerning explanations, definitions and prevalence of sexual violence in Australia. These are important in order to understand how community, legal and treatment responses have been shaped. Historical and contemporary approaches to preventing sexual violence and a review of evaluated sex offender treatment programs are discussed. A number of aspects about sex offender treatment programs and sexual violence prevention initiatives are then reviewed. Finally, suggestions for future practice and research in the area of sex offender treatment programs are identified.

2. The social problem of sexual violence: Theoretical explanations, definitions and prevalence

2.1 Theoretical explanations of sexual violence

Overshadowing all theoretical explanations of sexual violence is the ongoing influence of the morality and punishment paradigm - the earliest explanations of sexual violence were based on questions of morality and punishment. Positions of morality and punishment discriminated against certain acts of sexual violence, but not others. This was highlighted in the early work of Gebhard, Gagnon, Pomeroy, and Christenson (1965), where they stated that the '...inevitable fusion of secular law and religious belief has given rise to some curious inconsistencies' (p. 3) in what is deemed a sexual offence. These processes of morality and punishment are underpinned by factors such as:

• the legacy of institutional patriarchal processes - Women and children were property. In this way, these explanations focused on the external threat from men outside intimate or family relationships. Thereby, rape is seen as a heinous crime warranting severe penalties when committed by men who are not in legitimate positions of control over female victims.
• the morality surrounding the sexual act - This has influenced the law, regardless of the status of consent. This is especially evident in the case of homosexuality. In this way, sexual violence and what is categorised as sexual deviance have been collapsed into a range of sexual offences.

The historical legacy of these positions remains in community attitudes and the legal system. Until relatively recently, sex between men was a criminal offence, but rape within marriage was not considered a criminal act, and there remains a tenuous recognition of the occurrence and pursuit of charges relating to incest, especially where the offender is the father. The social norms surrounding what is considered acceptable sexuality and sexual relations continue to influence understandings and attitudes towards sexual violence (Taylor, 2004).

Explanations of sexual violence can be conceptualised in terms of differing theoretical perspectives, from individual psychopathological perspectives to structural/feminist theory through to post-structural and postmodern explanations. Each category has its internal variance. In some cases, explanations focus on the individual variation from 'normal sexual practice' or the abhorrence of the act within societal morality. It is important to view each of the explanations with these influences in mind.

Much of the research in sexual offending stems from an aetiological premise to identify individual characteristics that cause sexual violence. These explanations are steeped in the tradition of scientific enquiry in the fields of biology, psychology and psychiatry, where the focus has been on psychopathological and cognitive behavioural causes of sexual offending. This has strongly influenced the historical (for example, see Gebhard et al., 1965) and contemporary (for example, see Doren, 2002; Schwartz & Cellini, 1997; Ward & Beech, 2006) perspectives on sexual violence shaping the overall debate, along with legal and treatment responses. Pioneering research by Gebhard et al. set out to identify how sex offenders differed from those who had not committed sexual offences. There continues to be a growing body of literature attempting to capture characteristics and circumstances to differentiate sex offenders from other 'normal' men.

Explanations within the psychological category include: biological functioning, neuroscience, cognitive deficits or abnormalities, and social learning theory. Ward and Beech (2006) utilised much of this literature on sexual offending to develop an integrated theory of sexual offending. Their theory placed particular importance on the integration of how biological functioning of the brain and social learning interact with neurological processes pertaining to motivation, emotion, perception, memory and decision-making. While they acknowledged the social and cultural environment, their emphasis was on individual aetiological factors that impacted on understanding and treating sex offenders. Ward's and Beech's theory provided an overview of individualistic explanations of sexual offending. The use of theories such as that proposed by Ward and Beech has resulted in most treatment approaches focusing on psychopathology, cognitive distortions and behavioural deficits.

Social and structural explanations of sexual violence have attempted to attribute sexual violence to particular social structures and processes. Social psychology and sociology have utilised social learning and socialisation theory to explain how sexual violence is initiated and continued. For example, pornography has been linked to rape by social psychologists because of its portrayal of women as willing participants (Brehm & Kassin, 1989). Sociologists have also used class and power models of analysis to theorise about the causes of sexual offending.

This tradition of social enquiry has influenced the development of feminist perspectives on sexual violence. Central to the feminist explanation is the structure of gender relations that create an imbalance of power between men and women. Early feminist liberation work theorised that sexual violence was 'a conscious process of intimidation by which all men keep all women in a state of fear' (Brownmiller, 1975, p. 87). Overarching social, structural and feminist explanations of sexual violence have disputed that sexual offending is separate and deviant; rather it is seen as a consequence of current social and human conditions. This premise is at odds with individualistic explanations from within psychology and psychiatry.

Contemporary feminist theory has developed a sophisticated theoretical argument by not differentiating between men who commit sexual violence and those who do not (Cossins, 2000). This development has been influenced by postmodern/post-structural theory that recognises multiple explanations (Lancaster & Lumb, 1999). The focus of this social enquiry does not exclude individual or structural factors, but rather examines both the embodiment and institutions of sexual practice in gender relations. This makes it possible to examine acts of sexual coercion that remain hidden or taken for granted as 'normal' social practices within the confines of heterosexual dominance (Chung, 2005).

Postmodern and constructivist perspectives on sexual offending have evolved from a critique of sexual deviancy and power relations. A range of perspectives have emerged from postmodern theorists. Some writers in cultural studies have used postmodern theory to suggest that sexual offences have been constructed because they differ from 'normal' sexual practice. This literature critiques 'contemporary concern' for sexual violence - for example, in the case of child sexual abuse - suggesting that these concerns rest on essentialist and uncritical understandings of intergenerational sexual relations (Plummer, 1991).

Conversely, some writers examine the position of ethics, citing ways in which particular masculinities hold a position of privilege that overrides women's and children's right to consent. The use of a multilevel analysis of sexual offending in postmodern perspectives has utilised studies in masculinity, especially Connell's (1995, 2000) concept of hegemonic masculinity. Connell suggested that a diversity of masculine identities exists, with differing amounts of social acceptability and power. However, there are dominant and dynamic constructions of masculinity, and this hegemonic version of masculinity is at the foundation of why sexual violence occurs. These theoretical perspectives have given rise to therapeutic approaches, using narrative therapy (White, 2000), which shift the focus from the causes of sexual violence to positively constructing men's preferred identities in ways that exclude their use of violence. Jenkins (1990) also developed an approach to therapy that focuses on what 'restrains' men from engaging in respectful relationships.

Table 1 illustrates the diverse range of explanations of sexual violence. The paradigm of morality and punishment continues to strongly influence explanations and responses to sexual violence. Individualistic explanations focusing on psychopathology and cognitive/behavioural deficits have had the most influence on contemporary sex offender treatment. This has strongly shaped how sexual offending is constructed as a clinical deviance requiring treatment. These influences are elucidated and critiqued in more detailed in section 4.

Table 1 Explanations of sexual violence

In summary, the majority of explanations of sexual offending have largely focused on individual psychopathology, and this has had a significant impact on sex offender treatment orthodoxy. More recent developments in post-structural and contemporary feminist theories offer an opportunity to better conceptualise sexual offending as occurrences within a complex set of social relations and institutions. This enables a more sophisticated understanding of sexual violence that allows for an environment of disproportionate power relations between men, women and children, which creates the social conditions for an independent conscious act of sexual violation, as opposed to a deviant unconscious act determined solely by intra-psychic processes. Incorporation of these contemporary understandings will allow for a more comprehensive response to sexual violence across society.

2.2 Towards a gendered analysis and response to sexual violence

The influence of feminist theory on explanations of sexual violence has been relatively recent and has not yet significantly impacted on sex offender treatment approaches. Rape and incest have long been acknowledged as crimes in Australia; however, it was not until the 1970s that the concept of sexual violence emerged (Cook, David, & Grant, 2001). The 1970s saw challenges to the prevailing notion that sexual assault was primarily 'stranger rape' (Cook et al., 2001); the feminist movement contested the notion that rape is a crime committed in public, by strangers, against women who 'asked for it', wore 'suggestive' clothing or lied to protect their reputation.

Until this time, rape and incest were understood solely as the individual pathology of the men who committed such acts. These sex offenders were considered 'deviant' from the rest of the population and were characterised as having poor impulse control and being poorly socialised in childhood. As a result, they were considered unable to have appropriate intimate sexual relations, which led to their 'deviant' sexual behaviour. Such an understanding of sexual assault implied that sexual gratification of a male biological sexual drive cannot be controlled. This individually focused 'hormonal'/'biological' understanding could then be used to explain why men are the perpetrators in the vast majority of sexual violence, while also reducing their level of responsibility for their actions by pointing to their biological drive. Consequently, it followed that reforming and treating sex offenders would require that such individuals be detected by the state and incarcerated and/or treated through a focus on their individual 'deficiencies' and with medical interventions that aimed to reduce their sexual drives.

Feminist analyses of rape and sexual assault that emerged in the 1970s did not view these actions as a matter of sex, but rather as the gendered abuse of power - an act of humiliation and control (Brownmiller, 1975). Feminists argued that sexual violence is a crime perpetrated by men against women in both the public and private spheres, and in the context of intimate partner and family relationships (Cook et al., 2001). The introduction by feminists of the term 'sexual violence' expanded the definition to include a range of sexual behaviours that could be used by men to degrade, humiliate and maintain power over women.

Such understandings of sexual assault place a very different emphasis on who is considered to be a perpetrator of sexual violence, and the policy and program approaches to stopping sexual violence. Perpetrators of sexual violence within this explanation are broadened to include male partners and acquaintances who harass, coerce, pressure and/or force women to have sex against their will or engage in any sexual behaviour with women to which they do not or are unable to consent. A feminist analysis therefore assumes that the elimination of sexual violence is intrinsically linked to gender equality, as it is men's social power as a group and personal power in their intimate relationships which enables the continuation and condoning of sexual violence.

More recently, the study of masculinity, with its history in feminist understandings of gender, has provided insights in understanding men's use of gendered violence. Connell's concept of hegemonic masculinity (1995, 2000), which referred to dominant prescribed ways of being a man that emphasise heterosexuality, power, success, non-victimisation and control, has been influential. The consideration of masculinity has furthered our understanding of the ways in which sexual violence is a means of men having personal and social power over women; in particular, how acts of sexual violence are presented as 'normal' heterosexual masculinity that is all powerful and privileges men's 'needs'. The performance of successful heterosexual masculinity by men in groups is often predicated on the objectification and sexualisation of women. Recently, such behaviour has been considered within the context of men and sport (Benedict, 1998; Parissa, 2002). These social practices of hegemonic masculinity have also often been associated with victim-blaming in cases of sexual violence, where women are considered to be teasing and consenting to sexual activity by merely being present in such social settings. There has been an increase in knowledge about what sexual violence is and who commits such acts. However, the recent media responses to professional sportsmen's alleged sexual violence highlights how social attitudes still tend towards victim-blaming of women and normalising or trivialising of men's abuse of power. For example, sportsmen have been seen as victims of female fans' adoration, making accusations of sexual violence appear farcical.

2.3 Definitions and implications

Using a gendered analysis, policy approaches to sexual violence suggest the importance of responses being intersectoral, including the use of: law enforcement, justice and correctional systems to respond to individual perpetrators; education to promote gender equality and respectful intimate relationships; employment and community services to enable women to access resources for economic and personal equality; and changes in social attitudes to promote gender equality. These policy and program responses are much more complex and wide ranging than the detection of individuals committing sexual offences. One of the tensions that emerges in such approaches is that, while there is a social and political analysis of sexual violence, offences and treatment tends to remain individually focused for the safety of their victims and the community.

As knowledge and understandings about sexual violence have increased over the last 35 years, definitions have expanded to include a wider range of behaviours and circumstances. The variations in definitions of those who sexually abuse children are linked to both understandings of sexual violence and the discipline or field from which it has originated, as is shown below.

Psychopathological explanations have shaped the definitions outlined in the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM IV) (American Psychiatric Association [APA], 1994). The DSM IV classifies sexual violence towards adults and children within the broad category of 'paraphilias'. The essential diagnostic features of paraphiliacs described in the DSM IV are:

...recurrent, intense sexually arousing fantasies, sexual urges, or behaviours generally involving 1) nonhuman objects, 2) the suffering or humiliation of oneself or one's partner, or 3) children or other non-consenting persons, that occur over a period of at least 6 months (p. 522).

Within the broad category of paraphilias, those who are sexually attracted to and/or sexually assault children are referred to as 'paedophiles' and the definition of 'sexual sadism' best describes, amongst other behaviours, the sexual assault of adult victims.

The DSM IV definition of paedophiles has been criticised because it implies that such people are psychiatrically abnormal or are suffering with a mental disease or disorder predisposing them to perpetrate sexual offences, and that is very rarely the case (Glaser, 1997). Regarding paedophiles as psychiatrically abnormal can facilitate attitudes implying that the actions are not the perpetrators' fault, thus reducing their responsibility for their sexual offending. There are implications for treatment when the clinical diagnosis is regarded as an explanation for the offending behaviour. Treatment may focus on 'curing' or 'reducing' the diagnosis rather than addressing the offending behaviour and holding perpetrators accountable for their actions.

The DSM IV definition identifies paedophiles as those who have a sexual attraction and/or sexual fantasies about children, but do not necessarily act upon these attractions; whereas legal definitions of sexual violence, including paedophilia,¹ focus on the perpetrator's behaviours (James, 1996). Thus, a person can be clinically diagnosed as being a paedophile without sexually abusing a child.

The DSM IV description of the sexual sadist, which includes descriptions of non-consenting sexual acts being carried out, identifies the chronic nature of the problem (APA, 1994):

Sexual Sadism is usually chronic. When Sexual Sadism is practiced with non-consenting adults, the activity is likely to be repeated until the person with Sexual Sadism is apprehended (p. 530).

The category of sexual sadist does not refer exclusively to sexual assault; however, as is the case with the paedophilia definition, DSM IV pathologises the perpetrator's controlling and humiliating behaviour of another person.

The legal system relies on narrow definitions of child and adult sexual assault, based on the behaviours of perpetrators. The definitions have expanded from times when sexual assault was considered to only be heterosexual, vaginal intercourse for which there was no consent (Heath, 2005). Current definitions of adult sexual assault (Heath, 2005) and child sexual abuse (Bromfield & Higgins, 2005) vary between jurisdictions and states and territories in Australia. There are three key components to legal definitions of sexual assault, which include:

• behaviour defining sexual assault, such as genital sexual contact, oral sexual contact and the continuation of sexual activity after consent is withdrawn
• non-consent and the inability to consent to sexual acts²
• the mental state of the accused (Heath, 2005).

An example of a legal definition of sexual assault is the Western Australian Criminal Code Act 1913 (section 319), which names the sexual offence as sexual penetration without consent. This definition defines penetration as:

1. to penetrate the vagina (which term includes the labia majora), the anus, or the urethra of any person with -
1. any part of the body of another person; or
2. an object manipulated by another person, except where the penetration is carried out for proper medical purposes;
2. to manipulate any part of the body of another person so as to cause penetration of the vagina (which term includes the labia majora), the anus, or the urethra of the offender by part of the other person's body;
3. to introduce any part of the penis of a person into the mouth of another person;
4. to engage in cunnilingus of fellatio; or
5. to continue sexual penetration as defined in paragraph (a), (b), (c) or (d). ...'consent' means a consent freely and voluntarily given and, without in any way affecting the meaning attributable to those words, a consent is not freely and voluntarily given if it is obtained by force, threat, intimidation, deceit, or any fraudulent means...

Legal definitions also influence the categorising of sex offenders based on: the types of sexual offences they perpetrate; victims' characteristics; and the relationship of the offender to the victim (Greenberg, Da Silva, & Loh, 2002). For example, offenders who perpetrate sexual violence against children are often referred to as 'child molesters' or 'paedophiles', and offenders who perpetrate sexual violence against adults are commonly known as 'rapists'.

Such definitions and categorising of perpetrators rely on legal interpretations of sexual assault and arguably do not adequately define, nor account for, all forms of sexual violence because they usually exclude non-physical and non-assaultive offences, such as sexual coercion, exposure and voyeurism (Lievore, 2004).

As a result of the narrow legal definitions of sexual assault and the complex legal processes associated with sexual assault cases, only a relatively small percentage of such cases lead to a conviction.³ Therefore, a large number of people who are reported for sexual assault are not later convicted. This can be misinterpreted by the general public as indicating that there was no sexual assault and that women must be lying about being raped. Consequently, this perpetuates myths about sexual assault, such as: women commonly make false allegations, and women say 'no' when they mean 'yes'. Such myths lead to victim-blaming, and they play an important role in preventing women from naming the experience as coercion or assault, and in enabling men to see such behaviour as acceptable in a heterosexual relationship (Byers, 1996; Daws, Brannock, Brooker, Patton, Smeal, & Warren, 1995; Patton & Mannison, 1995; Russo, 2000).

3. Sexual violence and sex offenders in Australia

The notion that sex crimes are perpetrated by a small number of habitual, sexually deviant offenders, usually strangers, conflicts with the fact that most victims are sexually assaulted by someone they know (Lievore, 2005, p. 293).

An accurate assessment of the prevalence of sexual violence in any community is always likely to be underestimated because the vast majority of sexual violence remains unreported (ABS, 1996, 2002a; Mouzos & Makkai, 2004). Rates of prevalence will also differ according to a number of methodological issues (ABS, 2004b; Neame & Heenan, 2003; Schweitzer & Dwyer, 2003). These include:

• the definition of sexual violence used in a study - a narrow definition will produce smaller numbers than that of a broader definition, which may include experiences such as sexual pressure and coercion, and unwanted touching
• the study sample - prevalence may be estimated from what is commonly known as a 'clinical sample', which refers to those people known to have committed sexual offences or those seeking help for the effects of sexual violence. The sample may also be drawn from the general population, or a sub-population (for example, students) where the prevalence of sexual violence is unknown.

There are two main ways in which estimates of sexual violence have been gathered: surveys of the population to identify rates of sexual violence and records on the numbers of people arrested and charged with committing sex-related crimes.

Surveys generally reveal a higher prevalence of sexual assaults when compared to police records (ABS, 2004b), as it is recognised that not all sexual offences are reported to the police (ABS, 1996, 2004b; Schweitzer & Dwyer, 2003). Criminal justice records tell us about the number of people and incidents reported and convicted for sex crimes; therefore all forms of sexual violence which do not come to the attention of the authorities or are outside legislative definitions are not recorded. These figures are important; however they underestimate the prevalence of sexual violence in a community. To estimate sexual violence prevalence, surveys of victims are more frequently used, as they can include both reported and unreported incidents. Both survey and criminal justice record data are presented in the following sections to offer a more complete overview.

3.1 Sex offenders

Sex offenders are a heterogeneous group of people who are not easily identifiable. They come from all types of social, income, racial, ethnic and religious groups (James, 1996). Offenders can be married or not, employed or unemployed, have children and partners or not. Probably the only generalisations that can be safely made about sex offenders are that they are almost always male and they are almost always known to their victims (ABS, 1996, 2004a).

A recent development that extends insights into sex offenders has drawn on the principles of situational crime prevention (Wortley & Smallbone, 2006). Wortley and Smallbone have argued that this approach moves beyond sole reliance on the dominant intra-psychic understandings of offenders and tertiary treatment to considering how the environment can be modified and monitored to prevent child sexual abuse and its re-occurrence. It focuses on the person in their environment so that the likelihood of sex crimes being committed is based on both the individual actor and how his immediate setting facilitates or prevents further sex crimes. Rather than just profiling child sex offenders to differentiate them from the rest of the population, the situational approach acknowledges their individual traits as well as examining how the environment can be modified. For example, Wortley and Smallbone suggested an examination of the settings where the crime is committed - domestic, institutional and public spaces - and how potential offenders can be deterred in such environments. They proposed other modifications to the environment, such as the monitoring of offenders in the community, and primary prevention through increasing children's protective strategies. However, Wortley and Smallbone were careful to warn that they are not suggesting the child is in any way responsible for the sexual abuse. Within the situational approach, the heterogeneity of sex offenders is acknowledged and there is a strong rationale for moving beyond individual treatment approaches.

In a survey of the general population, Finkelhor and Lewis (1988) found that between 4 and 17 per cent of men acknowledged having sexually abused a child. In a survey of male university students in Australia and the United Kingdom, approximately 15 per cent of participants self-reported some likelihood of sexually abusing a prepubescent child if they could do so without risk of being caught (Malamuth, 1989).

As Lievore (2004) noted, the official statistics on apprehension, conviction and incarceration of sex offenders 'represent only the tip of the iceberg of all sex crimes perpetrated' (p. 24) as the majority of sexual offences remain unreported. However, research and data on incarcerated offenders is still important to examine because it provides a profile of at least one group of sex offenders and knowledge about the impacts of treatment programs and other responses. Such data sets can also provide information about trends in the state's responses to sexual violence and identify potential areas for reform.

In Australia, between 1988 and 1998, the population of prisoners incarcerated for perpetrating a sexual offence increased from 10 per cent to around 14 per cent (Smallbone & Ransley, 2005). According to the ABS (2002b), between 1996 and 2001, the proportion of Australian prisoners sentenced for sexual assault decreased from around 13 per cent in 1996 to 10 per cent in 2001. On 30 June 2001, the total Australian prison population comprised 22,458 prisoners, of whom 11 per cent of male prisoners (1873 men) were incarcerated for committing sexual assault. However, this figure does not represent prisoners who were convicted of sexual assault on ancillary charges only.

Smallbone and Ransley (2005) revealed that the number of sex offenders entering North American prisons has increased in recent years. Between 1980 and 1994 the number of prisoners entering North American prisons increased by 200 per cent and the number of prisoners incarcerated for sexual offences increased by 330 per cent (Greenfeld, 1997). In 1998, 94 000 prisoners were incarcerated for perpetrating sexual offences (Beck & Mumola, 1999). The proportion of incarcerated sex offenders increased by 26 per cent by 2002 (Smallbone & Ransley, 2005), with 118 500 prisoners incarcerated for perpetrating sexual offences (Harrison & Beck, 2003).

Police apprehension statistics provide some information about sex offenders who come to the attention of the criminal justice system (Lievore, 2004). Data on police apprehensions are available for South Australia, Victoria and Queensland from police services, and from the Crime Research Centre in Western Australia. Lievore cautioned that apprehension statistics do not necessarily equate to the number of individuals apprehended or to the number of offences committed in a counting period as jurisdictions vary in counting methods.

In South Australia, between 1 January and 31 December 2001, South Australian Police made 740 apprehensions for sexual offences (Office of Crime Statistics and Research, 2002). Fifty-nine per cent of the apprehensions were for rape and indecent assault. Of all those apprehended for sexual offences, 65 per cent were aged between 25 and 59 years and one quarter were aged 24 years or younger. Nine per cent of all persons apprehended for alleged rape were aged 17 years or younger.

Victorian Police apprehended 3439 alleged sex offenders in the financial year ending 30 June 2001 (Victorian Police, 2002). Of these apprehensions, 528 were for rape and 2911 were for other sexual offences, while ten per cent of all apprehensions were persons under the age of 17 years.

During the 2000-2001 financial year, Queensland Police made 5816 apprehensions for sexual offences (Queensland Police Service, 2001). Of the total apprehensions, 1076 apprehensions were for rape and attempted rape and 4740 were for other sexual offences.

Fernandez and Loh (2001) found that, in 2001, 516 individual persons were arrested in Western Australia for sexual offences. Males under 17 years comprised 13 per cent (66 young men) of those arrested, with 3 per cent of these being under the age of 14 years.

While individual states' apprehension counting methods may differ, Lievore (2004) highlighted that some variables remain constant across jurisdictions. Lievore showed that across the jurisdictions:

• between 97 and 99 per cent of alleged sex offenders were male
• juvenile sex offenders comprised approximately 10 per cent of all alleged sex offenders
• alleged sex offenders categorised as being of Indigenous appearance comprised between two and nine per cent of all alleged sex offenders
• Caucasians comprised between 76 to 85 per cent of all alleged sex offenders (where such information was recorded) (p. 18).

3.2 Offender-victim relationships

Most sex offenders are known to their victims. Women are less likely to report sexual violence to the police if the perpetrator is known to them, especially if the perpetrator is a current intimate partner (ABS, 1996, 2004a). The Australian Bureau of Statistics' (2004a) Recorded Crime Statistics indicated that 58 per cent of female victims of sexual assault knew the offender and, of these offenders, approximately 50 per cent were family members. These results are consistent with the finding of the Women's Safety Australia survey (ABS, 1996), which showed that, of the women aged 18 years and over who reported that they had been subjected to one or more incident of sexual assault since the age of 15 years, only 11 per cent of the women were sexually assaulted by strangers. Of the women who reported having been subjected to sexual violence since the age of 15 years, the following data about the perpetrators' relationship to the victim were identified:

• 3.6 per cent were current partners
• 21.5 per cent were previous partners
• 34.1 per cent were boyfriends
• 8.3 per cent were family members
• 26.5 per cent were friends
• 7.4 per cent were bosses or co-workers
• 5 per cent were other men known to the women (ABS, 1996, p. 23).

Thus, the data overwhelmingly show that most women are subjected to sexual violence by men who are known to them. However, recorded statistics on the prevalence of sex offenders in the community only represent 'visible' perpetrators. As the majority of sexual violence incidents remain unreported, the majority of sex offenders remain undetected.⁴

3.3 Juvenile sex offenders

Literature specifically focusing on young male sex offenders was rare until the 1980s (Becker & Johnson, 2001; Davis & Leitenberg, 1987; Smith, Wampler, Jones & Reifman, 2005). The lack of attention on this group contributed to the belief that young men rarely perpetrate sexual violence with 'serious' or physical consequences (Davis & Leitenberg, 1987). However, 20 to 30 per cent of rapes and 30 to 50 per cent of child sexual assaults are perpetrated by adolescents (Becker, Cunningham-Rathner & Kaplan, 1986; Davis & Leitenberg, 1987).

Common cultural attitudes, such as the saying that 'boys will be boys', also downplay the incidence of young men perpetrating sexual violence (Smith et al., 2005). This attitude also views acts of sexual violence employed by young men against their girlfriends as 'experimentation', 'play', 'curiosity' or 'just part of growing up' (Becker & Johnson, 2001; Grant, 2000; Smith et al., 2005). This attitude has also been responsible for young sexual offenders not being held accountable and responsible for their offending behaviours (Grant, 2000).

Recent research on juvenile sex offenders has highlighted that common cultural attitudes towards young sex offenders and not holding juvenile perpetrators accountable for their actions is highly problematic (Connolly & Wolf, 1995; Grant, 2000). Aljazireh (1993) argued that patterns of sexual offending often begin in adolescence and that many adolescent sex offenders go on to perpetrate more serious forms of sexual violence in adulthood. Thus, it is imperative that there are adequate and effective interventions with juvenile sexual offenders that include mechanisms of accountability insofar as treatment options are concerned.

3.4 Women's reporting of experiences of sexual violence

There is relative agreement in the literature that various forms of sexual violence, such as unwanted advances, sexual harassment, and sexual pressure and coercion are common experiences for many women (ABS, 1996, 2004b; Bateman, 1991; Byers, 1996; Hird, 2000; McIntosh & Griffin, 2001; Patton & Mannison, 1995; Russo, 2000; Western Australia Crime Research Centre & Donovan Research [WACRC&DR], 2001). The rates of sexual violence in Australia are based on women's self-reports from various surveys:

• the National Crime and Safety Survey (ABS, 2002a)
• the Women's Safety Australia survey (ABS, 1996)
• Recorded Crime: Victims, Australia (ABS, 2004a)
• the Australian component of the International Violence Against Women Survey (IVAWS) (Mouzos & Makkai, 2004).

The results of these studies are described and their findings discussed in the following section.

The National Crime and Safety Survey (ABS, 2002a) collected data on sexual assaults experienced by women, aged 18 years and over, via posted questionnaires to private residences. The Women's Safety Australia survey (ABS, 1996) utilised face-to-face interviews with a random sample of 6300 women. The Australian Bureau of Statistics report (2004a), Recorded Crime: Victims, Australia, is a compilation of all the crimes reported to the police in each Australian territory and state. The Australian component of the International Violence Against Women Survey (Mouzos & Makkai, 2004), reported on interviews with 6677 women, aged between 18 and 69 years, about their experiences of physical and sexual violence.

In the twelve months to April 2002, 28 300 women were subjected to 62 700 incidents of sexual assault, of whom 66 per cent reported being subjected to a single incident of sexual assault and the remaining 34 per cent experiencing two or more incidents (ABS, 2002a). From the Women's Safety survey findings, it was estimated that in the twelve months prior to the study, 100 000 women experienced an incident of sexual assault in Australia (ABS, 1996). Of the crimes reported to the police between 1 January 2003 and 31 December 2003, there were 18 237 incidents of sexual assault, which represents an incident rate of 91.7 per 100 000 people (ABS, 2004a).

The National Crime and Safety Survey found that women were subjected to significantly more sexual assaults than men. The recorded crime figures for 2003 reveal that women were the victims in 82 per cent of reported cases (ABS, 2004a). The majority of offenders were male in all the Australian surveys. Ninety nine per cent of the sexual assaults were perpetrated by male offenders in the Women's Safety survey (ABS, 1996).

Young women experienced a higher prevalence of sexual assaults than other age groups. In the Women's Safety survey, women aged 18-24 years experienced the highest prevalence rate of sexual assaults of all women (19 per cent) (ABS, 1996). The Recorded Crime study found young women aged 16-19 years recorded the highest victimisation rate - 497 victims per 100 000 people (ABS, 2004a).

The percentage of women who reported incidents to the police is small. The Women's Safety survey found 15 per cent of women reported incidents of sexual violence to police (ABS, 1996), and 20 per cent of incidents were reported in the National Crime and Safety Survey (ABS, 2002a). The IVAWS found similar rates of reporting incidents of sexual violence to the police - only 14 per cent of women experiencing sexual violence from an intimate partner and 16 per cent of women experiencing sexual violence from a person other than an intimate partner (Mouzos & Makkai, 2004).

Consistent with previous research, contemporary Australian studies show the majority of women are sexually assaulted by men known to them. The National Crime and Safety Survey found that perpetrators were known to 58 per cent of the women (ABS, 2002a). The IVAWS showed only 1 per cent of women were raped by a stranger and that some women disclosed violence from both intimate partners and other people (Mouzos & Makkai, 2004).

From these findings, it is apparent that sexual violence is a common experience for many Australian women and takes a variety forms, not all of which fit within the narrow legal definitions of sexual violence. Although the prevalence of sexual violence is difficult to accurately assess, the findings indicate the following:

• women are more likely than men to be the victims of sexual violence
• women are more likely to experience sexual violence from a person they know, rather than a stranger
• perpetrators of sexual violence are most likely to be male.

Furthermore, this review of prevalence studies confirms previous evidence that, in the majority of the cases, sexual violence still rarely comes to the attention of the criminal justice system.

4. Response to sexual violence

In light of the restrictive focus on explanations, definitions and prevalence discussed above, in this section of the paper we will consider responses to stop sexual violence, and the treatment of sex offenders; recall historical approaches to responding to sexual violence and the development of sex offender programs; and examine measures aimed at the primary and secondary prevention of sexual violence.

4.1 Early responses to sexual violence outside the criminal justice system

Traditionally, the prevention of sexual violence emphasised women taking responsibility for their own victimisation; women were advised to modify their own behaviour to avoid sexual assault. Women were encouraged to avoid wearing 'suggestive' clothing, not to go into the community alone and to avoid 'risky situations' (Neame, 2003). Preventing sexual violence was the responsibility of the individual woman and not of the men who were perpetrating the violence. This implied an attitude towards sexual violence where men were not considered responsible for their behaviour; instead, women had the responsibility to not be the catalysts for men's behaviour. Women who were partnered or married in the home were not considered to be at risk of sexual violence as it was only considered to be perpetrated by strangers.

Such approaches to preventing sexual violence have been criticised by feminists for reinforcing social attitudes and myths about sexual violence that place blame with the woman for the sexual assault (Carmody & Carrington, 2000). Such approaches are also regarded as more concerned with women adopting techniques of risk management than with preventing sexual violence. As such, they do not challenge the underlying reasons or community attitudes that are fundamentally linked to sexual violence.

Neame (2003) argued that feminist campaigning in the 1970s raised awareness of sexual violence and was instrumental in revealing the extent of sexual violence women experienced and its context (that is, that predominately perpetrators were known to the victim, as opposed to the 'stranger rape' myth). Men's capacity to perpetrate sexual violence against some women was arguably reduced by this campaigning as the issues of sexual violence became recognised by mainstream society and women became more aware of their rights (Neame). As public awareness of sexual violence increased, women's liberation groups sought to work with relevant government organisations, including partnerships with legal and medical professionals to advocate reforms to policies and practices that would improve the treatment of and services offered to rape victims.

Women's resistance strategies to stop sexual violence were a key focus for some researchers in the 1980s. Neame (2003) noted that various studies examined whether resistance - verbal and/or physical - was an effective rape avoidance strategy (see Bart, 1981; Bart & O'Brien, 1984; Kleck & Sayles, 1990). Although this strategy received much attention and many studies were undertaken, the notion of women's resistance as a rape avoidance strategy was highly problematic because it perpetuated the false notion that the majority of rapes occur in public, and did not address the rape of women by men well known to them. The strategy was also criticised for holding women responsible for their own victimisation (ABS, 1996; Neame; Victorian Law Reform Commission, 2003).

In contrast to this, there were no moves towards making children responsible for avoiding being sexually abused. This quite rightly was seen as unacceptable as children could not be held responsible for their own abuse. At the same time that awareness was raised about childhood sexual abuse, general programs were being introduced, primarily in schools, which encouraged education around what could constitute protective behaviours for children to use against potential offenders. In an audit of child protection prevention programs, Tomison and Poole (2000) found that two-thirds of primary prevention school programs were based on the Protective Behaviours model, which was used both in its initial form and in forms which, over time, were tailored to suit specific target groups. The other main support in recognition of childhood sexual abuse was to encourage adults affected by such abuse to seek support for its negative effects through counselling and self-help groups.

4.2 Historical responses to sex offenders: The emergence of the medical-legal nexus

To respond to the social deviance of sex offenders, laws have been enacted related to rape. These laws have been premised on an individual psychopathology view of sex offenders. In the 1930s in North America, the so-called 'sexual psychopath' statutes were passed and Laws (2003) noted that convicted sex offenders used these statutes to their advantage. On convincing a review panel that their sex offending was the result of a mental disorder, the sex offender was then sentenced to an institution, usually a secure ward in a psychiatric hospital, and not a prison.

The justice system expanded to have three main strategies: punishment, practical help, and treatment. Punishment was seen as an individual and community deterrent (McGuire & Priestly, 1985), although this rationale was applied generally to offending behaviour and was not peculiar to sexual offending. Practical help was focused on trying to reduce recidivism. It was thought that if an offender's 'everyday life' problems, such as housing, money management, employment and family welfare, were addressed, the offender would be less likely to re-offend. To address these issues, offenders would receive counselling from social workers and/or probation officers either in prison or post-incarceration in the community.

Treatment was focused on the personality of the offender and viewed the sexual offences 'as no more than a superficial symptom of some underlying disease or disorder' (McGuire & Priestly, 1985, p. 22). The early treatment of convicted sex offenders was thus largely based on psychodynamic and psychoanalytic treatments, which were the dominant psychological interventions of the time (Laws, 2003). These treatments and associated notions of sexual psychopathy have now been disproved (Laws). Many studies since the mid-1960s (for example, see Gebhard et al., 1965) have repeatedly shown that the majority of sex offenders are not suffering from a mental disorder so severe that it predisposes them to perpetrate sexual violence (Glaser, 1997; Laws). While it is not uncommon for some sex offenders to be diagnosed with a personality disorder, this diagnosis would by no means exempt the perpetrators from taking full responsibility for their offending behaviours (Laws).

The medical/disease model, sometimes referred to as the individual pathology model, was the foundation for early responses to rape and sexual assault, in which sexual violence was seen narrowly as the result of some individual adult men who preyed upon women and children for sexual gratification. The model emphasises the need for the medical and legal systems to operate collaboratively in order for sex offenders to be identified, punished and/or treated as necessary. Such treatment approaches, regulated by medicine and the law, tended to include pharmaceutical and talking therapies. Where treatment was not successful and re-offending for sexual violence occurred, it was deemed appropriate for the perpetrator to be (re-)incarcerated. This social perspective has not changed considerably over time. However, they overlooked gender and other social power structures as having any considerable impact on sexual offending. The sexual assault of intimate partners was, until recently, considered the entitlement and biological need of the adult, heterosexual man rather than sexual violence, and continues to be a major area of under-reporting.

The evolution of treatment approaches is discussed in the following section.

4.3 Sex offender treatment approaches

Although many clinicians often fail to see the relevance of theories, the fact is all practitioners have a perspective on how sexual offending develops and what maintains it; it is simply that most clinicians rarely make their theories explicit although these theories influence what they do in assessment and treatment (Marshall & Laws, 2003, p. 102).

In relation to sex offender treatment approaches, the majority of published work in the area has been from the discipline of psychology, but treatment programs for sex offenders have evolved within criminal justice settings, as the majority of programs have been developed for convicted offenders. In this respect, the development of programs has tended to be influenced by trends in offender programs more generally. The development of treatment approaches began with aversion therapy, commencing in the 1960s. Cognitive techniques were introduced in the 1970s, which aimed at modifying the behaviour and thinking of offenders. From then on, cognitive behavioural approaches became identified with good practice, and, in the last few years, the development of a strengths-based or 'good lives' approach is emerging. Apart from the biological and aversion therapy approaches, all other approaches tend to be run as group programs.

4.3.1 Biological interventions

Biological interventions draw from individualistic medical model understandings of sexual violence. Becker and Johnson (2001) argued that biological interventions should be considered in the treatment of sex offenders. Recently, there has been a focus on hormonal therapy and its effects on sexual behaviour, and hormonal agents, including antiandrogens, are used to reduce sexual offending (Becker & Johnson, 2001; James, 1996). Antiandrogens are able to reduce offending behaviour and fantasies as they suppress libido (Becker & Johnson, 2001). However, antiandrogens and other hormonal agents are not recommended for use in sex offenders under the age of 16 years (Bradford, 1993), with the exception of 'difficult cases' and where other treatments have been unsuccessful (Becker & Johnson, 2001, p. 280). Other organic treatments, such as surgical castration and stereotaxic neurosurgery, are also employed in the treatment of sex offenders, although these treatments are not utilised in Australia (Bradford, 1990; James, 1996). Becker and Johnson argued against surgical castration for juvenile offenders, advocating instead for less invasive treatments. Biological interventions have not been considered a major policy option in the treatment of sex offenders in Australia.

4.3.2 Aversion therapy

Aversion therapies were early behavioural modification treatments used for a range of 'unusual' sexual behaviours, including sex offences, child sex offences, homosexuality, transvestism and fetishism. Laws and Marshall (2003) noted that aversion therapies typically used a variety of noxious stimuli 'paired with either images of the target behavior (Pavlovian conditioning) or the enactment of the deviant behavior (operant forms of punishment)' (p. 83). These therapies included electric aversion, where an electric shock was associated with the deviant sexual behaviour, and injecting a nausea-inducing agent, such as apomorphine, where the nausea was paired with the deviant behaviour (Laws & Marshall, 2003). Laws and Marshall noted that other aversion stimuli were also investigated, including foul odours (see Colson, 1972), covert aversion images (see Cautela, 1967) and the use of embarrassment and/or shame (see Serber, 1970). Recent research has shown aversion therapies to be ineffective in producing permanent changes in sexual behaviour (see Quinsey & Earls, 1990; Quinsey & Marshall, 1983).

4.3.3 Systemic approach

Systemic approaches to sex offender treatment are based on the early systemic family therapy model. When used with a family, the model directs the focus towards how the individual family member's 'presenting problem' is a 'symptom' of what is wrong with the family as a system. Therefore, in order to address the problems of the individual family member, there must be intervention with the whole family. Vivian-Byrne (2004) describes systemic approaches as based on a notion which places 'emphasis outside the individual [offender] and on to the relational patterns of feedback around an individual's behaviour' (p. 188). This systems approach to treating sex offenders considers environmental and personal triggers and the capacity to modify individual responses to them (see Hoffman, 1993; Jones, 1993).

Once power issues are considered by the practitioner and the perpetrator, a systemic approach is thought to be useful in creating opportunities for change for sex offenders (Vivian-Byrne, 2002). Systemic approaches expand the offence cycle to include a range of influences, such as an emphasis on previous relationship experiences which are specific and important to the offender; 'for example, the perpetrator's experience of being abused him/herself, his or her relationships with parents, or his or her ability to form and maintain intimate relationships with adults' (Vivian-Byrne, 2004, p. 189). If a connection can be forged between these past experiences and the offending patterns of behaviour, then these experiences, rather than the sexual offences, can become the focus of intervention (Vivian-Byrne, 2004).

This approach, however, is highly criticised, especially by feminist theorists (Vivian-Byrne, 2002, 2004), as it is based on the assumption that the offender's behaviour is a symptom of other problems in his system of relationships. This then can apportion blame and responsibility for the offending with both victims and offenders (Vivian-Byrne, 2004). It does not look at power differentials between victim and offender and how the offender abuses his power to sexually assault the victim. Feminists argue strongly that sole responsibility lies with the perpetrator and that the victim should not be considered in any way responsible for the sexual offending. Vivien-Byrne (2002) argued that working with sex offenders in ways which examine their patterns of behaviours and relationships can be useful, provided power is taken into account, and the victim not held responsible.

4.3.4 Cognitive behavioural therapy

In the 1970s, behavioural modification interventions were expanded to include cognitive processes and social skills training (Marshall & Laws, 2003). The interventions were essentially two-fold: identifying and changing cognitive distortions, and offering alternative, pro-social behaviours which, in combination, aim to stop offending (Marshall & Barbaree, 1990). In essence, cognitive treatment was intended to 'modify deviant cognitions which precede deviant behaviour' (Mandeville-Norden & Beech, 2004, p. 200). These deviant cognitions are the ideas and beliefs of the offender, often considered to excuse or justify the offending (for example, 'the child wanted a sexual relationship', or 'the woman was game-playing and really wanted to have sex'). These cognitive distortions then become the focus of change in treatment. When such approaches began, it was important that the therapist took a confrontational approach with the sex offender to alter his cognitions and to support him taking responsibility for his actions.

Developing offenders' pro-social relating and 'dating skills' was the other focus, with the understanding that perpetrators would cease sexually offending if they could derive sexual gratification from legitimate, adult sexual relationships (Mann, 2004; Marshall & Laws, 2003). It was considered particularly important for child sex offenders to learn how to have appropriate adult relationships in order to reduce their risk of offending against children. The offenders were then offered alternative ways of behaving that were pro-social, in conjunction with distorted cognitions being challenged.

The focal points of intervention in cognitive behavioural approaches with sex offenders have been: maladaptive sexual arousal, lack of victim empathy, poor social skills and low self-esteem (Vivian-Byrne, 2004). It is argued that change in these areas is needed for there to be long-term cessation of the sexual offending behaviour (Marshall & Barbaree, 1990; Vivian-Byrne, 2004).

Cognitive behavioural treatment for sex offenders has been the most commonly adopted approach internationally. It has received considerable attention in the literature and there has been comprehensive documentation and evaluation of the approach (Lievore, 2004). The evaluation research of these programs will be reported later in this paper.

William L. Marshall has been a major contributor to knowledge about sex offenders and cognitive behavioural treatment programs. Marshall's ideas and concepts in this area continue to evolve and his influential research about sex offenders has been utilised in program development. Marshall (1993) identified sex offenders as having insecure attachment bonds with parents in early childhood, which led to the conclusion that sex offenders have therefore not learned trust and intimacy. Consequently, sex offenders do not have socially acceptable means to have relationships, are socially lonely and have 'intimacy deficits'. This lack of experience of appropriate intimacy can result in a pre-occupation with sex as the only source of intimacy, which leads to promiscuity and the potential for sexual deviance (Mann, 2004). Treatment approaches must therefore address these cognitive distortions and skill deficits of sex offenders.

One of the next developments in the treatment of sex offenders was the introduction of relapse prevention, aimed at stopping recidivism after treatment and court orders had ceased. This was part of a broader trend in relapse prevention that was being used in other areas of corrections, such as alcohol and substance abuse.

4.3.5 Relapse prevention

Marshall (1999) noted that, while cognitive behavioural treatment programs vary in content, most programs developed an emphasis on relapse prevention. Marshall and Laws (2003) argued that the most crucial development in the 1980s was the adaptation of the relapse prevention model to the field of sex offender treatment. This model was widely accepted by practitioners who were 'eager to provide their clients with a way of combating the risks and temptations they faced once direct treatment was over and they were placed back in the essentially free-operant world' (Marshall & Laws, 2003, p. 99).

Sex offenders, both juvenile and adult, make decisions that put them in high-risk situations in which they perpetrate sexual violence, and an offender's chances of sexually re-offending, or relapsing, increase when he places himself into a high-risk situation (Becker & Johnson, 2001). The role of sexual fantasies is also important to a relapse prevention framework, and Becker and Johnson noted that if the sexual offender 'fantasizes and masturbates to inappropriate sexual thoughts, he or she increases the likelihood that the abusive or deviant sexual fantasies will increase in frequency and intensity' (pp. 276-277). Cognitive distortions will eventually develop, which offenders will then use to justify their sexually offending behaviour.

Relapse prevention assumes that the probability of the relapse of sex offenders is based on a range of factors and that offenders should be aware of patterns of behaviours and decisions that can lead to sexually re-offending (Lievore, 2004). According to Marshall (1999), 'relapse prevention describes a set of procedures designed to assist the offender in maintaining the gains he has made in therapy' (p. 233). This model is underpinned by the notion that offence precursors can be identified and addressed (Becker & Johnson, 2001). Marshall noted that there are many elements to a relapse prevention framework; however, he identified the main elements necessary for an effective relapse prevention approach:

• identification of the offence cycle
• specification of the factors (such as anger, intoxication, low self-worth, stress, depressed mood) and situations (such as being alone with a child/children, driving aimlessly) that increase risk
• identifying coping skills that can reduce the risk of relapsing
• creating plans to avoid risk and high-risk situations, and unexpected risks.

4.3.6 Schema-focused therapy

Recent developments in cognitive behavioural approaches have been schema-focused therapy and the strengths model that has been entitled the good lives approach. Mann (2004) described these developments as being a shift from psycho-educational to psycho-therapeutic approaches. Cognitive behavioural approaches assume that the cognitive distortion is the offender's inappropriate or incorrect belief, understanding or interpretation about the offending, the victim and/or the effects of offending. Such cognitive distortions have been confronted and alternative cognitions offered. Mann described the change in this theory as one of examining the underlying cognitive processes in which it is considered a problem of 'faulty information processing' (p. 143). This has led researchers and practitioners to examine the underlying schema.

Young (1990) defined schema as 'extremely stable and enduring themes that develop during childhood [that] are elaborated throughout an individual's lifetime, and are dysfunctional to a significant degree' (p. 9). The underpinning schema hypothesis common among sex offenders has been called the 'suspiciousness schema hypothesis'. Mann (2004) explained that 'sexually aggressive men seem to believe that women are game-playing, deceptive people, who use aggression as a form of seduction, and who are deceitful when they behave seductively' (p. 143).

The hypothesis appears to be an accumulation of what sex offenders have said about women victims. Rather than a 'suspiciousness hypothesis' which is common to sex offenders, it can appear to the feminist eye as a description of misogyny, to which many men ascribe in order not to take responsibility for their behaviour or in order not to accept sexual rejection.

Within this approach, these schema, which are the cognitive structures dominating self-defeating and offending behaviours, need to be altered if the sexual offending behaviour is to change (Mann, 2004; Vivian-Byrne, 2004). In a reframing of previous understandings of cognitive behavioural approaches, post-hoc rationalisations and excuses by the perpetrator, which were once viewed as cognitive distortions in need of confronting, become positive indicators that the offender has insight into violating standards of acceptable behaviour and that there is the capacity for change (Mann).

Vivian-Byrne (2004) noted that in schema-focused therapy, perpetrators are regarded as being 'in need of education, and their thinking in need of restructuring or altering, by the application of techniques [which address the distorted information processing]' (p. 186). Schema-focused therapy is recognised as being a labour-intensive process which places high demands on both the practitioner and the perpetrator. The approach emphasises distorted information processing, identified primarily by the practitioner, and altering these distortions becomes the focus of intervention.

4.3.7 Strengths-based/good lives approach

In recent years, prominent authors William L. Marshall, Tony Ward, Ruth Mann and their colleagues have been advocating a shift away from the confrontational and challenging approaches of the cognitive behavioural therapist to strategies that acknowledge the strengths of the offender and his capacity to lead a fulfilling life. This essentially is a shift in approach from risk management of the offender to a supportive approach. Marshall, Ward, Mann, Moulden, Fernandez, Serran and Marshall (2005) argued that, to date, treatment has been negatively framed both in the language used to describe those who commit sex offences and in the focus of treatment. For example, they describe a concern with the negative emphasis of relapse prevention as a list of 'don'ts' and things to avoid rather than supportive goal-setting that is collaborative and supportive.

The assumption in many programs is that for offenders, avoiding relapse is the most important goal. In fact, although reducing re-offending rates is always the primary goal for the treatment provider, offenders themselves might have other priorities...Traditionally RP programs can fail to engage offenders because they impose a primary goal on them rather than negotiating and agreeing on the goal of intervention (Marshall et al., 2005, p. 1106).

The proposed good lives model works on the underlying assumption that enhancing the social functioning of the offender will decrease the offending behaviour. It is a goal-oriented approach whereby meaningful goals are set for individuals in the program, the means to achieve the goals are identified, and agentic thinking (self-efficacy) is encouraged by facilitators to ensure those goals can be met. This is similar to many social work and psychological approaches that have been employed to work with people in overcoming negative personal experiences and their effects on social functioning (Saleebey, 2002). This enables the offender to develop the knowledge, skills, social competence and personal confidence to meet his goals to lead a productive life that will inhibit offending.

Whereas previous models viewed men identifying themselves as sex offenders as a positive sign of responsibility, Marshall et al. (2005) argued that this was negative for offenders as they would then view themselves solely within this frame of reference and see themselves as unable to change. The good lives or strengths approach attempts to break down the dichotomy between sex offenders as being different from the rest of the community who are not committing such offences.

According to research on good lives, all people seek sexual satisfaction, feelings of comfort, and some degree of power and control in their lives. Nonaggressive people may not want to strike back at someone who has offended them, but they will certainly want some form of redress even if just an apology. Thus, each of the goals that sexual offenders seek in their abusive behaviours is a goal they share in common with other people, it is simply the pathways they choose to obtain these goals that are dysfunctional and hurtful to others (Marshall et al., 2005, p. 1104-1105).

However, the approach raises a number of concerns. The above quote implies that sex offenders have experienced some form of sexual rejection as offensive, they understandably want redress, and it is just their means of obtaining redress that are inappropriate. It appears from the quote that somehow victims have done sex offenders an injustice and the crimes are the redress. We are not convinced that everyone is so similar that they have universal needs in their lives. We are also concerned at the suggestion that offender programs should not include a strong focus on responsibility, for fear of it alienating the offender. Although the use of strengths-based and empowerment approaches have a long history in the human services with voluntary clients, the application of such an approach with sex offenders is more contentious and represents a shift from previous thinking. It is not clear from the literature available whether this approach considers any accountability to the victims as being relevant or desirable.

Another issue which this approach raises is the importance of the individual therapist/worker in the success of the program. Mann (2004) argued that, traditionally, the influence of the therapist on program outcomes was not considered a significant variable and that the role of the therapist was to confront the offender about their beliefs and descriptions of events, and to facilitate him taking responsibility for his offending. In the shift from psycho-educational to psycho-therapeutic approaches, such as the good lives model, the role of the therapist is evidently more critical. Mann suggested that important skills for the therapist to demonstrate in order to improve program outcomes are empathy, warmth, directive approaches, and being supportive and rewarding of offender progress.

Using this approach with sex offenders is relatively new, and it will be important for it to be evaluated, particularly as it stands against much of the previous orthodoxy on both program logic and approach to sex offender treatment.

4.3.8 Risk assessment as good program practice

As sexually violent offenders pose a threat to victim and community safety, there has been considerable effort in developing risk assessment measures to predict the likelihood of sexual violence recidivism (Abracen & Looman, 2005). There is a range of psychological risk assessment scales which have been developed in the past ten years for incorporation into treatment programs with individual participants. There are two types of risk assessment scales; 'measures that incorporate dynamic (that is changeable) as well as static (relatively enduring qualities) variables' (Abracen & Looman, 2005, p. 14). An example of a dynamic risk factor is denial and minimisation, which may change over the course of treatment. Leading authors in the area advocate the importance of the advancement of these scales in the assessment and treatment of sex offenders (Abracen & Looman; Mann, 2004). Mann highlighted the importance of using both static and dynamic risk assessment measures to ensure that the safety of potential victims and the community is foremost in treatment, as offenders' levels of risk may change over time. Risk assessment measures are increasingly being used in the assessment of offender treatment needs and as clinical data in the evaluation of the offender's progress while in the program.

This is similar to developments in domestic violence perpetrator programs, which are increasingly developing and using risk assessment tools that are specifically targeting domestic violence offender risks and their changing risk due to their situations and their types of offending (O'Leary, Chung, & Zannettino, 2004).

4.4 Evidence about the effectiveness of sexual violence treatment programs

In this section, evidence about the effectiveness of sex offender treatment programs is examined and identified good practice is discussed. Australian programs are described in terms of program content and evaluations of their effectiveness.

Debate exists as to whether sex offender treatment programs reduce sexual recidivism (Lievore, 2004; Marques, 1999). However, a recent meta-analysis of program effectiveness showed that treatment programs can be effective in reducing recidivism (Hanson, Gordon, Harris, Marques, Murphy, Quinsey, & Seto, 2002). Given the dominance of the psychology discipline in such treatment programs, the evaluations are understandably positivist in design and are grounded in 'objective' quantitative data. Lievore outlined factors that can have an impact on judgements about the effectiveness of sex offender treatment programs:

• the measurement of recidivism (or criteria for success and failure)
• definitional differences (for example, whether recidivism is counted as reconviction or reported re-offending, as the latter will provide a higher rate of recidivism)
• measurement error in criminal justice data
• variability between official and unofficial data sources
• differences between comparison groups on important predictor variables, including offence and offender characteristics
• the duration and context of treatment provisions
• the use of appropriate statistical methods to account for differences in time at risk of offending (p. 90).

Methodological differences can also have an impact on the variability of recidivism rates when evaluating sex offender treatment programs (Lievore, 2004; Marques, 1999). Marques argued that very few studies evaluating sex offender treatment programs are well designed. Ideally, studies should compare a group of offenders who willingly agree to participate in the treatment to a control group comprised of untreated sex offenders (Lievore, 2004). However, this design, while methodologically sound, can be considered unethical, as the control group is being denied treatment. Thus, in practice, control groups usually comprise untreated offenders, non-sexual offenders and perpetrators who started the treatment program but, for whatever reason, did not complete it. Importantly, Lievore highlighted that, with the high rates of unreported sexual violence and sexual offences, some participants in the non-sexual offender control group may actually be sex offenders.

Such evaluations in the areas of gendered violence and criminology are often fraught with these methodological and ethical hurdles and challenges; however, the importance of evaluation continues to dominate debates in the area. Undertaking sex offender treatment program evaluations is important to a range of professionals and policy-makers who make decisions about sex offenders (Marques, 1999). For example, Marques noted that outcome data are required for future decision-making and policies about the sentence options to be considered, types of sentences and orders, which perpetrators are suited to which types of treatment programs, the levels of community supervision post-release from custody and which programs will continue to receive funding.

One longitudinal evaluation of a North American sex offender treatment program, the Sex Offender Treatment and Evaluation Project (SOTEP), incorporated elements of best-practice research design (Lievore, 2004) and has been implemented in California. The study sought to evaluate the effectiveness of treatment in reducing recidivism in sex offences against adults and children (Marques, 1999). The program's theoretical orientation was cognitive behaviour therapy, situated within a relapse prevention framework. The key features of the SOTEP evaluation included:

• an experimental design that included the random assignment of volunteers to either treatment or no-treatment (control) conditions and another control group of non-volunteers
• an intensive cognitive behavioural inpatient treatment program designed specifically to prevent relapse among sex offenders
• a one-year aftercare program in the community
• a comprehensive evaluation of both in-treatment changes and long-term effects (including a follow-up period in which recidivism rates for treated and untreated participants are measured for at least five years) (Marques, 1999, p. 439).

Preliminary findings for the study did not conclusively determine whether the treatment program was effective (Marques, 1999), although there were some data suggesting that treated sex offenders' subsequent re-offences were less serious. The preliminary findings also showed that treated child molesters with only male victims or with both female and male victims had a 13 per cent sexual re-offence rate, compared to 16 per cent and 29 per cent for the two untreated control groups, and their treatment appeared to be more effective than for child molesters who offended exclusively against female victims.

Hanson et al.'s (2002) meta-analytic review of treatment evaluations found recidivism rates for those in sex offender treatment programs were lower (12.3 per cent) than for those in comparison groups (16.8 per cent). The type of program was important to this outcome; specifically, they found that cognitive behavioural programs had the best results. This has been subsequently confirmed by other authors such as Abracen and Looman (2004). In the United Kingdom, Manderville-Norden and Beech (2004) reported that cognitive behavioural programs in conjunction with a probation service were more effective in reducing recidivism than probation services as the sole intervention. Importantly, Hanson et al. also found that treatment approaches implemented prior to 1980 were not valuable.

Obviously, the long-term evaluation of sex offender treatment programs is important, but costly. How recidivism is measured must be taken into account when determining a program's success, as previous research tells us that conviction rates for sexual offences are low and are therefore not an accurate indicator of actual recidivism. There has been debate among researchers about the level of sex offender recidivism. There is a strong argument for long-term follow-up of sex offenders, as there is evidence to suggest re-offending occurs for up to twenty years after incarceration and/or treatment. Cann, Falshaw and Friendship (2004), in a 21-year follow-up of convicted sex offenders, found 22 per cent had been reconvicted. It should be noted, however, that not all of these offenders had access to programs and, of those that did, programs were variable.

In short, there is evidence to suggest that treatment programs for sex offenders, particularly cognitive behavioural programs, are likely to reduce recidivism.

4.5 Australian evidence about sex offender treatment programs

In this section of the paper, we distinguish between adolescent and adult sex offenders, as the program approaches can differ for important reasons related to the development of sexual offending and subsequent risk factors..

4.5.1 Australian intervention programs with young men

Providing sex offender intervention programs for adolescents is logical, as research indicates that the majority of high-risk adult sex offenders start offending in adolescence (Abel & Rouleau, 1990; Manderville-Norden & Beech, 2004; WACRC&DR, 2001), and they often commit multiple sexual offences over a lifetime (Sapp & Vaughn, 1990). Typically, adolescent sex offenders abuse younger children and, to a lesser extent, children their own age. Recent research into the prevalence of sexual violence in Australia has highlighted the need for preventative initiatives, including treatment programs, with young men. This is particularly important, as this paper highlighted earlier, because young women are the group most likely to experience sexual violence.

Adolescence is a timely intervention point, as it is then that young people's use of violence in intimate relationships often commence (WACRC&DR, 2001). Furthermore, intervention at this stage has the capacity to reduce the acceptability of violence among adolescents and offers them strategies for change, before the violence becomes entrenched and more severe. There is also potential at this stage to offer alternatives based on non-violent, healthy relationships.

In Australia, treatment for young men who perpetrate sexual violence or who are at risk of using sexual violence has been developing since 1991, and most Australian states and territories now have programs targeting young sex offenders⁵ (Flanagan, 2003). The following is a summary of the ACT Young Sex Offender Program and the Victorian Male Adolescent Program for Positive Sexuality. These two programs are discussed because they have been the most rigorously evaluated and comprehensively documented.

The Young Sex Offender Program is part of the Australian Capital Territory's Corrective Services Offender Intervention Program Unit (Flanagan, 2003). The program was established in 2001 and targets young sex offenders between the ages of 12-24 years. Offenders are referred to the program from the juvenile justice system; however, the program will also accept referrals from other sources, as long as the sexually offending behaviour has been reported to the police. This program is delivered in a community-based setting (Flanagan, 2003).

Community safety and the needs of victims and potential victims are the key priorities for the Young Sex Offender Program (Flanagan, 2003). The program operates from a cognitive behavioural therapy model and utilises a group therapy format. Family therapy and individual therapy sessions are also offered by the program.

The Male Adolescent Program for Positive Sexuality (MAPPS) targets adolescent sex offenders in Victoria. It was established by the Victorian Department of Human Services in 1993 as an alternative to incarceration (Flanagan, 2003; Lievore, 2004). The program focuses on rehabilitation and is premised on the notion that incarceration is not a conducive environment in which to effect changes in the adolescent's behaviour and attitudes. Lievore noted that the program places emphasis on the offender taking responsibility for his behaviour and 'making the necessary changes for controlling his behaviour, with the goal of building an offence-free future' (p. 101).

Adolescents referred to MAPPS have been found guilty of perpetrating a sexual offence by the Victorian Juvenile Justice system (Lievore, 2004). The majority of the offenders in the program are aged between 14 and 17 years. While participation is meant to be voluntary, many offenders are required to undertake the program as a condition of their court orders. Participation is usually for 11 months of weekly sessions; however, the duration of attendance varies, depending on the individual's court order. Lievore expressed concern about the young offenders' risk of re-offending and lack of accountability when they do not participate in the program or disengage from it.

MAPPS is a group treatment program that moves through the following stages:

• assessment
• basic groups
• transition program
• advanced group
• relapse prevention (Lievore, 2004, p. 101).

Four-and-a-half years after its inception, MAPPS underwent an evaluation (Curnow, Streker, & Williams, 1998). In comparing those who completed treatment with non-completers, the evaluation showed that the treatment reduced sexual recidivism rates in convicted adolescent sex offenders. Of the 138 offenders who completed treatment between 1993 and 1998, 5 per cent committed further sexual offences, and treatment completers were found to be eight times less likely to re-offend compared to adolescents who did not complete the program. The evaluation also reported that offenders who completed the program were more likely to take responsibility for their offending behaviours (Curnow et al., 1998).

While the evaluation results are encouraging, Lievore (2004) cautioned against drawing causal links between MAPPS and reduced sexual recidivism rates. She suggests that the evaluation could be strengthened by:

• increasing the follow-up period
• comparing recidivism rates with those of a randomly assigned control group
• further examination of the differences between the offenders who completed the program compared to non-completers (p. 102).

In comparison to the Australian findings, an evaluation of a North American adolescent sexual offender treatment program (SOTP) compared the effectiveness of the SOTP with another therapy intervention (not solely focused on sex offending) (Lab, Shields, & Schondel, 1993). In terms of further sexual offending, the results were similar for both the experimental and control groups: one of the 46 perpetrators (2.2 per cent) assigned to the SOTP sexually re-offended, and four of the 109 offenders (3.7 per cent) who undertook a treatment not specific to sexual offences sexually re-offended. For all cases of recidivism, both sexual and non-sexual, eleven juveniles (24 per cent) assigned to the SOTP re-offended, compared to 18 juveniles (18 per cent) assigned to a treatment not specific to sexual offences who re-offended. The evaluation results suggested that the SOTP was no more effective at reducing recidivism for juvenile sex offenders than a treatment not specific to sexual offences (Lab et al., 1993).

The evaluation results differed for a range of reasons, which have previously been outlined in terms of considering differences across methodologies and programs (Lievore, 2004). However, when using recidivism rates that are based on re-convictions for sexual offences, low rates could either be a promising result or could be a sign of offenders not getting caught again. However, given that research shows that when sexual violence offending begins at an early age, there is a higher risk of offending in adulthood, treatment efforts for sexually violent adolescent males should be continued and evaluated.

4.5.2 Australian treatment programs with adult male sex offenders

Treatment programs with adult men aim to reduce their sexual recidivism and to protect the community and potential future victims. Thus, treatment programs are victim-centred, with the welfare of the offender regarded as secondary to the interests of the community (Howells, Heseltine, Sarre, Davey, & Day, 2004; Lievore, 2004). These programs target men known to have perpetrated sexual violence against women and/or children and are delivered in prisons⁶ and within the community.⁷ A conviction is often required to compel offenders to attend treatment, so participation in treatment is generally part of an order or sentencing conditions. The following is a summary of two treatment programs delivered by the New South Wales Department of Correctional Services, as well as the Sexual Offender Treatment Program (SOTP), which is run by the Queensland Department of Corrective Services.

The New South Wales Department of Correctional Services delivers both prison- and community-based programs to convicted sex offenders (Lievore, 2004). Two different prison-based programs are offered: Custody-based Intensive Treatment (CUBIT) and Custody-based Intensive Treatment Out Reach (CORE). Both programs offer modules of varying lengths covering offence-related and offence-specific topics, including anger and stress management (Lievore, 2004).

CUBIT is run over eight months (moderate-intensity program) or ten months (high-intensity program) and is targeted to offenders assessed as being of moderate to high risk (Lievore, 2004). The program is predominately run in a group setting and administered by a multidisciplinary team, including psychologists, human service workers and custodial workers. During the program, the offenders are accommodated in a self-contained unit at the Long Bay Correctional Centre. Lievore argued that this physical environment is more conducive for the offenders to work on long-term behaviour and attitude change.

The CORE program is for convicted sex offenders assessed as being low risk and covers the same educational modules as the CUBIT program (Lievore 2004). CORE is facilitated in a group format and is run on two half-days per week, over a period of approximately five months. During the program, prisoners continue their regular institutional activities (that is, education and work duties).

Mammone, Keeling, Sleeman, and McElhore (2002) evaluated the impact of the therapeutic intervention of the Australian CUBIT and CORE programs, based on 55 offenders who completed either of these programs in New South Wales between 1999 and 2002. Therapeutic interventions utilised in both CORE and CUBIT target cognitive distortions, and offenders are required to undertake a range of psychological tests upon entering and exiting the program to facilitate program evaluation (Mammone et al.). The results of the evaluation showed that offenders' general and offence-specific cognitive distortions were significantly lower at the post-treatment stage, compared to when the offenders entered the program (Mammone et al.).

The evaluation found that the residential environment in which moderate- to high-risk offenders undertook the CUBIT program facilitated a reduction in cognitive distortions (Mammone et al., 2002). Offenders in the non-residential CORE program, who remained in the general prison population for the duration of the treatment program, also exhibited significant reductions in their cognitive distortions. Mammone et al. reported that sex offenders' risk of sexually re-offending may be decreased if their reduced cognitive distortions can be maintained long term. Further research is required to evaluate the long-term impact of the CUBIT and CORE programs on convicted sex offenders upon their release from prison and following their parole period.

Another prison-based sex offender treatment program in Australia is the Queensland prison-based Sexual Offender Treatment Program. SOTP provides pre-release assessment and treatment for adult male sex offenders and is facilitated by a multidisciplinary team, including psychologists, educational officers, activities officers and program support workers (Schweitzer & Dwyer, 2003). The program is offered over 45 weeks and is divided into three phases: assessment and treatment phase (15 weeks), treatment planning phase (5 weeks) and intensive treatment phase (25 weeks). Using a group therapy format, the following modules are undertaken by offenders during the intensive treatment phase:

• effective relationships (including a focus on inappropriate communication styles)
• cognitive distortions
• control of deviant sexual arousal
• social issues
• victim issues
• relapse prevention (Schwietzer & Dwyer, 2003).

The programs described are based on the cognitive behavioural approach, which emphasises changing cognitive distortions and developing pro-social skills and behaviours that aim to reduce sexual re-offending.

An evaluation of the Queensland SOTP was undertaken by examining the sexual recidivism of convicted adult male sex offenders who undertook some or all of the SOTP while in prison. The offenders were released from custody between 1992 and 2001 and were monitored for recidivism for five years and one month post-release into the community (Schweitzer & Dwyer, 2003). The sample consisted of 445 men: 196 SOTP completers, 85 SOTP non-completers and 164 matched controls.

The outcomes of the SOTP evaluation showed that there was no statistically significant reduction in sexual recidivism following participation in this treatment program (Schweitzer & Dwyer, 2003). In this evaluation, recorded past histories of violent crimes and sex crimes were the strongest predictors of recidivism. The evaluation did not show that the program effects were statistically significant; however, Schweitzer and Dyer noted that this should be considered in light of a considerable amount of missing data for the evaluation, and variations and limitations in follow-up periods post-release. They asserted that evaluations of sex offender treatment programs need to continue to be undertaken.

Similar to the results on adolescent sex offender programs, the Australian evaluation findings on programs for adult sex offenders are mixed. The small sample size of Australian studies limits what can be concluded about the effectiveness of programs at present.

4.5.3 Good practice in sex offender treatment programs

In relation to good practice, program evaluation findings have had mixed results. Cognitive behavioural therapy has been the most commonly used model of treatment for sex offenders and the evaluations of such treatments showed mixed results. The Australian evaluations of sex offender treatment programs have shown that some have been successful, others less so; however, evaluations from North America and the United Kingdom, with larger samples, indicate that they have been effective (Hanson et al., 2002; Manderville-Norden & Beech, 2004). Lievore (2004) argued that international best-practice treatment programs for sex offenders are based on cognitive behaviour therapy that incorporates relapse prevention. Interventions underpinned by cognitive behaviour therapy focus on altering sexual behaviours and attractions, modifying cognitive distortions and addressing social difficulties (Marshall & Barbaree, 1990). In relation to relapse prevention, Pithers (1990) asserted that it provides offenders with 'tools' to enhance internal self-management skills for dealing with high-risk situations. The intention of relapse prevention is to promote not offending beyond the intervention period.

In general, treatment based on cognitive behaviour therapy within a framework of relapse prevention are delivered by a multidisciplinary team consisting of, but not limited to, social workers, psychologists, legal practitioners and workers from correctional services (Lievore, 2004).

Lievore (2004) outlined three common core issues which should be addressed in cognitive behavioural treatment programs for sex offenders:

1. Challenging beliefs that support offending. Offenders are required to:
   1. identify and challenge cognitive distortions and factors associated with offending
   2. acknowledge and take responsibility for the offending behaviour. This requires full and open disclosure of sexual offending
   3. understand offending in the broader life context and gain insight into the cycle of offending, the precise nature of the problem, its antecedents and its consequences.
2. Developing empathy. As affective deficits facilitate offending, it is essential that offenders:
   1. understand the impact of sexual offending and its consequences for victims, offenders and the community
   2. develop victim empathy.
3. Relapse prevention. To control offending behaviour, offenders must:
   1. learn to manage inappropriate sexual fantasies, thoughts and arousal patterns
   2. develop relapse prevention plans to manage the risk of future offending. This is a highly individualised approach to therapy that involves identifying high-risk situations and offence precursors and forestalling the threat of relapse by implementing coping responses or avoidance strategies (p. 78-79).

Group therapy is preferred over individual therapy sessions. Lievore (2004) argued that group therapy is more effective and cheaper compared to individual therapy.

Some of the strengths of group therapy in working with sex offenders are:

• group formats break the secrecy of sexual violence. Offenders are no longer in a position where they can justify, defend and rationalise their offending behaviour, as others become aware of the sexual violence they have perpetrated
• groups can provide a richly therapeutic environment in which group members are able to comment and provide insight into other members' problems. This can facilitate group members to utilise the opportunity to implement new ways of thinking and interacting
• group formats, when facilitated by a female and male worker, provide workers with the opportunity to model desirable ways of relating, thus facilitating attitude change in the offenders (Lievore, 2004, p. 79).

Abracen and Looman (2005) argued that individual approaches to treatment are relevant in cases where the offender has a serious mental illness or intellectual disability that impairs their capacity to operate in a group and therefore requires individual intensive treatment. Importantly, they noted that the use of individual treatment does not imply that they are higher-risk offenders. Rather, it is the result of being assessed as being unsuitable for a group program.

Lievore (2004) argued that sex offender treatment programs should aim to provide detailed and integrated systems of assessment, management and intervention. Similar arguments about the importance of the integration of services to ensure a comprehensive approach that promotes safety have been made for the organisation of domestic violence services (Shepard & Pence, 1999). Domestic violence services appear to have developed further in this regard, by involving justice agencies, corrections, women's support services, legal services and others. A key purpose of such approaches is to collaboratively promote the safety of women and children. Further collaboration in sexual violence responses could provide closer integration in developing both treatment and prevention approaches to sexual violence, which currently tend to run independently of each other.

Cognitive behavioural approaches to treatment programs have been identified as the most effective of the psychological approaches to date, based on much evaluation research. Unlike the field of domestic violence, which contains a mixture of treatment programs ranging from psychological psycho-therapy to pro-feminist psycho-educational programs, the sex offender treatment field has been narrower in its range of approaches. The pro-feminist psycho-education approaches adopted in domestic violence perpetrator programs are unlikely to have any influence on the treatment of sex offenders, given the dominance of psychology in this field of intervention. Relevant gender theory concepts in the area of sexual violence include: heterosexual dominance, gendered power relations and the social construction of masculinity. Concepts such as these provide insight into the link between the occurrence of sexual violence in the community and its impacts on both offenders and victims of such gendered violence.

Such concepts have been usefully utilised in the treatment and support of victims of sexual assault and adult survivors of child sexual assault (Reavey, 2003). A consideration of how these concepts impact on the behaviour, identity and the ways in which men relate to both women and other men can provide useful insights into the treatment of sex offenders (Flood, 2002-2003; Pease, 2004-2005). This would also lead to a closer alliance in approach between sexual violence services that support victims and those that treat offenders. However, the difference in the mandates of such organisations is acknowledged as being critical to how they work with victims or offenders. This is highlighted by Lancaster and Lumb (1999) in describing social workers' practice with male child sex offenders in the United Kingdom's Probation Service:

All practitioners we have spoken to recognise the socio-cultural influence within their work, but, in terms of what they can achieve on a day-to-day basis with individual offenders, they utilise techniques of intervention which are based on increasingly sophisticated knowledge of the psychological processes involved in effecting change in someone's behaviour (p. 125).

The dilemma of social structural explanations not being able to provide a readily translatable link to individual work on behaviour change has been documented. However, to focus solely on the individual psychology of men convicted of sexual offences ignores some important facets of the problem of sexual violence: it provides no theory for its primary prevention, it does not consider the social conditions, values and mores which determine what gets counted as sexual violence and who gets convicted, and it does not necessarily provide accountability to the victims of sexual violence.

In the next section on the prevention of sexual violence, the greater use of gender-based explanations of sexual violence will be examined, as will be the shift away from psychological underpinnings in the approach.

4.6 Prevention of sexual violence

This section of the paper describes some of the various education programs and campaigns that have been developed to prevent and stop sexual violence. Prevention programs aimed at stopping sexual violence generally target young men and women of particular age groups and within particular institutions, such as schools. These programs often include activities such as education programs, public awareness campaigns and programs to challenge attitudes. They tend to be universal in their approach in as much as they do not often target young people considered to be 'at risk' of either offending or being victims of sexual assault. Venues for such programs have traditionally been schools, universities and sporting clubs. Unlike sex offender treatment programs, many of the prevention programs have not been developed within a correctional or criminal justice setting. They have tended to emerge from partnerships with women's, youth and health services that have a concern with the prevention of sexual violence in order to stem the large numbers of women seeking help for sexual assaults.

4.6.1 Approaches to sexual violence prevention

Recent research indicates the urgent need for primary prevention initiatives related to sexual violence for young people (Mulroney, 2003). Adolescence is a unique time for primary prevention, as it is the stage when young people are working out their identities and values and beginning their own intimate relationships (Mulroney; Smith, Huxley, & McKernan, 2004).

The Western Australia Crime Research Centre and Donovan Research (2001) conducted a national study of young people's attitudes and experiences of domestic violence involving 5000 participants. One in three young people revealed that violence occurred within their own relationships. Fourteen per cent of young women reported they had been sexually assaulted. The study also found that 14 per cent of young women reported that their partner had attempted to force them to have sex. One in twenty of the young people considered forcing a partner to have sex was part of normal conflict in relationships. Due to the high prevalence of sexual violence reported in young people's relationships, the researchers concluded that sexual violence in young people's relationships should be a primary concern of governments.

International literature confirms sexual violence is a common experience among young people. For example, Sundermann and Jaffe (1995) reported that 23 per cent of American female high school students revealed that sexual violence was part of their dating relationships. Therefore, primary prevention is required with young people to address sexual violence and the development of healthy intimate relationships (Carmody, 2003a, 2005; Mulroney, 2003).

One approach to sexual violence prevention that emerged in the 1980s was the popularly known 'Just Say No' approach. This approach encouraged women to take responsibility for stopping male sexual violence by individually telling men that they did not want sex with them. This prevention strategy was limited in many respects. Firstly, it placed responsibility on the woman for controlling and stopping the man's behaviour. Secondly, it inadvertently blamed women for not being able to communicate clearly, which conflated the issue of sexual violence with sexual intimacy. In regard to the rationale of the 'Just Say No' approach, Kitzinger and Frith (1999) stated that sexual violence is often presented as a result of miscommunication between a man and a woman, where the man does not understand that the woman is refusing sex.

The miscommunication idea is highly problematic on a range of levels: it asserts that the woman is responsible for controlling the man's behaviour and for stopping the sexual violence; it is ignorant to the gendered power relations within heterosexual relationships; it ignores power differences between the genders; it assumes that the man will act respectfully to the woman's request to not engage in sexual intercourse/sexual activities; and it presents men as not being responsible for their behaviour and actions. Kitzinger and Frith (1999) further critiqued the 'Just Say No' strategy and its notions of miscommunication by arguing that in Western cultures, refusing invitations and requests is generally difficult, especially when there is a vested interest in continuing the relationship and where there is a power imbalance. However:

both men and women have a sophisticated ability to convey and to comprehend refusals, including refusals which do not include the word 'no', and we suggest that male claims not to have 'understood' refusals which conform to culturally normative patterns can only be heard as self-interested justifications for coercive behaviour (Kitzinger & Frith, p. 295).

In contrast to the severe limitations of this approach, Moira Carmody (2003a, 2005) conceptualised an important advancement in sexual violence prevention approaches and argued that people should think about alternative prevention education initiatives to sexual violence, namely ethical sexuality. To date, much focus and efforts into sexual violence prevention have been on tertiary intervention with 'known' sex offenders (Carmody, 2003a, 2005). The continued high prevalence rates of sexual violence suggest that these tertiary initiatives are having little impact on preventing the sexual violence from occurring (Carmody, 2005). Carmody argued for shifting the 'focus from teaching refusal skills and awareness-raising to a focus on promoting and developing ethical non-violent relating' (p. 478).

Carmody's (2003a, 2005) alternative approach focuses on providing the space to highlight both the pleasures and dangers in intimate sexual relationships and to discuss sexual negotiation. Young people's experiences of sex and of sexual negotiation are complex issues, due to the many forces that shape sexual negotiation (Smith et al., 2004). Young people may have a limited understanding of consenting to sex and of the differences in consent, including giving free consent, forced or coerced consent and non-consent (Smith et al., 2004). This approach centres on creating an understanding of ethical negotiation in all aspects of intimate relationships, both sexual and non-sexual (Carmody, 2003a). Carmody stated that this enables a shift from punitive education initiatives which 'achieve prevention through threat of punishment or controlling risk and promotes a pessimistic view of gender relations with men as always dangerous and women as passive victims' (p. 4).

4.6.2 School-based sexual violence education programs

A recent approach to sexual violence has been the development of prevention/education programs in Australian schools (Keel, 2005). School-based programs target young people who are at the stage of developing intimate relationships; the orientation is universal, with all students included (Urbis Keys Young, 2004). Most recently, an anti-violence education resource pack has been developed by the Australian government and distributed to all secondary schools (Australian Government Office for Women [AGOFW], 2005). This pack is aimed at students in Years 11 and 12 and equips teachers with the information and strategies for teaching about healthy relationships and identifying abusive behaviours (Keel, 2005). While the primary focus is physical violence, Keel noted that issues of sexual violence are discussed. Young men are primarily targeted by this prevention strategy, with the intention of encouraging them to take responsibility for their behaviour and to act appropriately in the context of intimate relationships (AGOFW, 2005; Keel, 2005). Carmody's (2005) work on ethical sexuality is more comprehensive than the approach described above as she includes the critical step of ethical and healthy alternatives, which moves beyond just letting people know what is unacceptable.

Another school-based prevention strategy specifically aimed at sexual violence prevention is offered in the north-west region of Melbourne by the Centre Against Sexual Assault (CASA) (Keel, 2005). The program involves workers from CASA attending schools and conducting an initial 90-minute session with staff and between three and five workshops with students from years 9, 10 and 11. The staff session focuses on introducing the topic of sexual violence and aiding the staff to appropriately respond to disclosures of sexual violence.

Initially, CASA offered students one session aimed at education about sexual violence (Keel, 2005). Findings from evaluations of the program have led to extending the program to include three to five sessions. Topics covered in the sessions include:

• an introduction to sexual violence
• the meaning of consent and social pressures that influence communication
• the impact of sexual violence on victims
• social action strategies to prevent sexual violence, including an opportunity for the students to create banners and slogans for a mock anti-sexual violence campaign (Keel, 2005, p. 22).

Female and male students are separated during the workshops until the final session, when they are brought together and encouraged to discuss their opinions with each other (Keel, 2005). While the curriculum is not dissimilar for male and female students, evaluations of the program revealed that separating the students into single-sex groups allowed students to communicate better with their facilitator. Keel noted that CASA does not initiate contact with schools; rather schools approach CASA, and the curriculum is tailored to suit the needs of each individual school. To date, 700 students and 200 teachers have participated in the program.

A criticism often directed at primary prevention programs is that the evaluation of their effectiveness is often inadequate (Schewe & O'Donohue, 1993; Weisz & Black, 2001). In comparison to treatment program evaluations, prevention evaluation is at an early stage of development. Some of the reasons accounting for the inadequacy of the evaluations have been small sample sizes, lack of medium- to long-term follow-up, high attrition rates in follow-up, reliance on the male participants to self-report behaviour, and the evaluation design not being able to exclude other variables which influence sexual violence attitudes and behaviours.

A large North American evaluation of a dating violence prevention initiative, the Safe Dates Program, overcame many of these limitations. The Safe Dates Program was conducted in fourteen schools in North Carolina and involved 33 agencies (Foshee, Bauman, Arriaga, Helms, Koch, Linder, & Fletcher, 2004). The program comprised a 10-session curriculum, with various activities directed at young people, parents and service providers, and the study had a four-year follow-up period, with a sample of 460, including 201 controls. Data were collected at post-intervention intervals of one month, one year, two years and four years. The evaluation indicated that Safe Dates was successful in reducing aspects of physical and sexual victimisation and perpetration of dating violence, with '56% to 92% less dating violence victimization and perpetration at follow up' (Foshee et al., p. 623). A smaller study by Weisz and Black (2001) evaluating a dating violence and sexual assault prevention program also showed an increase in knowledge and improved attitudes at the six-month follow-up stage. A weakness of both studies was the reliance on self-report data to identify changes.

4.6.3 University-based sexual violence education programs

Published studies of sexual violence programs in universities have been based on North American experiences and have indicated that sexual violence on North American college campuses is a significant problem. Koss, Gidycz, and Wisniewski (1987) found that 54 per cent of college women had been subjected to some form of sexual violence. More recent studies have reported even higher victimisation rates. For example, a survey of 1160 college women found that 59 per cent had been subjected to sexual violence (Abbey, Ross, McDuffie, & McAuslan, 1996). In this study, 8 per cent of women reported experiencing attempted rape and 23 per cent of women reported having been raped.

In response to studies highlighting the prevalence of sexual violence on college campuses, many North American colleges and universities have initiated sexual violence prevention programs (Breitenbecher & Scarce, 1999; Heppner, Neville, Smith, Kivlighan, & Gershuny, 1999). Breitenbecher and Scarce outlined common features of such prevention programs:

• provision of information regarding the prevalence of sexual assault among college students
• debunking of rape socialisation practices
• identification of risk-related dating behaviours
• discussions of dating expectations and communication of sexual intentions
• induction of empathy for rape survivors (p. 460).

Breitenbecher and Scarce (1999) evaluated the effectiveness of a sexual assault education program delivered to university students, focusing on two domains: increasing knowledge about sexual assault and decreasing the incidence of sexual assault during a seven-month follow-up period. Of the 275 women recruited from the university student population to participate in the study, 224 (82 per cent) returned for the follow-up session seven months later. The women were randomly assigned to either the treatment or control groups. The evaluation found that the sexual assault education program was effective in increasing the women's knowledge about sexual assault, but was found to be ineffective in reducing the women's risk of being subjected to sexual assault during the seven-month follow-up period.

Breitenbecher and Scarce (1999) suggested four reasons for the ineffectiveness of the program in reducing women's risk of being subjected to sexual assault during the follow-up period. Firstly, the one hour initial session may not have been 'powerful enough to change women's risk-related behaviors or responses to unwanted sexual advances' (Breitenbecher & Scarce, 1999, p. 471). Secondly, the authors argued that the ineffectiveness of the program may be attributed to the program content. Thirdly, the program's focus on potential victims' cognitions and behaviours was linked to the program's failure to reduce the incidents of rape among the women. The program focused on women reducing their risk of sexual victimisation by altering their own behaviours. It could be argued that the program would have been more effective at reducing the incidence of sexual victimisation if the program had been targeted at male students, addressed potential perpetrators' cognitions and behaviours, and challenged their behaviours.

Other college sexual assault programs focus on date rape prevention. These programs, while acknowledging that men are always responsible for their sexually offending behaviour, have also been developed to educate women about lowering their risk of victimisation (Hanson & Gidycz, 1993). These date rape prevention programs generally include such features as consciousness-raising information about sex-role stereotyping and debunking rape myths, films depicting acquaintance rape scenarios and education regarding available rape counselling and crisis services (Parrot, 1991). Literature on rape prevention education highlights a need for further evaluations of the effectiveness of these programs (Hanson & Gidycz; Schewe & O'Donohue, 1993).

In response to the dearth of evaluations of college acquaintance rape education programs, Lonsway, Klaw, Berg, Waldo, Kothari, Mazurek and Hegeman (1998) evaluated the Campus Acquaintance Rape Education (CARE) program. The CARE program is a detailed course, offered in a large midwestern North American university, that trains undergraduate students to facilitate peer workshops in various campus settings. The program curriculum includes debunking rape myths from a feminist perspective, providing sexuality education, generating participant interaction and teaching non-confrontational approaches.

The evaluation collected pre- and post-test data on a range of attitudinal measures, which showed that CARE program participants experienced comprehensive attitude changes towards anti-rape ideology (Lonsway et al., 1998). The results also suggested that both male and female students who undertook the CARE program became 'more willing and able to directly express themselves and assert their needs in ways that facilitated increased sexual communication' (Lonsway et al., 1998, p. 73) and were less accepting of rape myths.

The North American university campus programs documented are a mix of awareness raising, education and prevention. The prevention approach has often been very literal in its interpretation, as it has presumed that sexual assault will be perpetrated by men and cannot be stopped, and that it is only through changing the behaviour of women that it can be reduced. This type of approach - while it may seem 'common sense' and practical - can come close to victim-blaming and viewing men's sexual violence as normal. It does not seek to address the social factors that ignore or condone sexual assault; rather, sexual violence is presented as a problem of individuals and that individuals should learn to avoid being sexually victimised.

4.6.4 Sexual violence education and sport

Recent reports of sexual violence perpetrated by high-profile male sporting identities have added to the call for education programs. Sporting associations have introduced player conduct codes and compulsory education programs. To some extent, these measures have also focused on how men can avoid allegations of sexual violence rather than substantially challenging aspects of masculine culture that support the sexual exploitation of women. Media coverage of police action against sporting identities has raised the profile of debates concerning the prevention of sexual violence. However, rather than viewing male sporting identities as more at risk of committing sexual crimes, there is a need to see these incidents of male violence as representative of an all-too-common occurrence in the wider community. It will be interesting to see whether this remains an issue for education and 'risk management' in professional sporting associations or a quick fix to 'bad press'.

4.7 Conclusions about sexual violence treatment and prevention approaches

The evaluations of sex offender treatment programs indicate that cognitive behavioural approaches are, to date, the most effective compared with other intervention approaches. The Australian programs described in detail earlier in this paper are all based on cognitive behavioural approaches and incorporate relapse prevention. In this respect, the programs are drawing on approaches deemed to be most effective in the published literature.

The majority of sexual violence prevention initiatives have originated from women's organisations and gender specialists in mainstream services that have histories of responding to the needs of victims of sexual assault. A component of many of these initiatives still include an emphasis on the historical approach of women avoiding the risks of sexual victimisation. Education initiatives have been targeted at females and males to dispel rape myths, to make women more aware of the threat of sexual assault, to provide women with information about support services if they are needed, and to educate men and women about the effects of sexual violence.

Overall the evaluation evidence of such initiatives indicate that knowledge and awareness of sexual assault and its effects are increased; however, this does not necessarily lead to a decrease in sexual assaults.

Most prevention initiatives rarely offer alternative ways for young people to enjoy ethical sexuality and healthy relationships (Carmody, 2003b). It could be argued that a major limitation of most prevention has been that the sources of the problem - both individual men who sexually assault women and the social conditions that overlook or minimise it - are not addressed by such initiatives. In examining the theoretical underpinnings of the prevention of sexual violence, Michael Flood (2002-03) argued that, ultimately, sexual violence reduction lies in undermining those aspects of culture that lend social support to violence; in particular, the ways in which masculinity is positively associated with a sense of entitlement, dominance and power. While Flood has been critical of many of the current initiatives for merely promoting the idea that violence against women is unacceptable, he has argued that prevention must be an essential component in the overall response to sexual violence:

Formal prevention and control strategies such as sound laws and integrated criminal responses are important. They can make a difference to victims' recovery and to the likelihood of perpetrators' recidivism, and they have symbolic value. But formal control strategies have little to work with in a climate where women and men do not formally report abusive events, most survivors remain silent...and dominant beliefs about violence convince many women that their experience was not rape or assault at all or that it was their fault...' (p. 25).

One of the most advanced approaches to prevention in Australia at present, which takes account of such complexities and previous criticisms, are the ethical sexuality concepts being developed by Moira Carmody (2003b, 2005). They represent a significant shift in the approach to sexual violence prevention for a number of reasons:

• the approach does not assume women are responsible for the prevention of their own sexual victimisation
• it moves beyond condemning those that commit sexual assault to developing alternatives which are non-violent and respectful, thus offering the development of a new cultural script to young people for relationships
• the underlying assumption is a positive view of sexuality for women and men which profoundly differs from sexual violence, and is not merely perceived as being about men's inability to communicate and inappropriate attempts at intimacy (Carmody, 2003b, 2005).

Finally, it is evident that there are very different theoretical ideas and concepts which underpin victim services and prevention initiatives compared with treatment programs for sex offenders. Whereas the sex offender treatment programs are heavily influenced by the individual psychology of sex offenders, prevention has frequently been based on a feminist analysis of sexual violence, which implies that sexual offending is an abuse of male power and act of gender inequality. It does not tend to pathologise either the victims or perpetrators of sexual assault and is based on the assumption that the social structures that privilege masculinity will continue to influence what is named as sexual violence and how those perpetrating it and who are victimised by it are consequently treated. In a further elaboration of Table 1 presented earlier in the explanations section of the paper, Table 2 below shows how the theoretical explanations of sexual violence have influenced prevention and intervention in stopping sexual violence.

* The darkness of the shading represents the strength of the explanatory influence in the area.

As Table 2 shows, there is, overall, a considerable gulf - theoretically and in practice - between the treatment of sex offenders and initiatives aimed at the prevention of sexual violence. Similarly, a theoretical gulf exists in sexual assault services for women victims and programs for sex offenders. These gaps restrict the levels at which a coordinated intersectoral response to sexual violence can effectively develop.

5. Conclusions and future directions

In the last 35 years, research into sexual violence has identified it as a major form of gender oppression. This has been highlighted by the emphasis on the criminalisation of male sexual violence against women. Importantly, in sexual violence, it has been recognised that the perpetrator is more often known to the person than a stranger. Nevertheless, perceptions of what constitutes sexual violence and how to respond have been overshadowed by morality and punishment paradigms. Concerns about sexual offender numbers and recidivism have led to treatment programs targeted at men as the main means to stop sexual violence. The focus of these programs has been on men who were found guilty of sexual offences in the criminal justice system, but such men represent a very small proportion of the total number of those committing sexual violence against women.

Criminal justice system definitions of sexual assault, while focusing on perpetrator behaviour, vary between Australian state and territory jurisdictions. Offenders are commonly categorised by the types of sexual offences they perpetrate and by victim characteristics; for example, 'child molester', 'paedophile' and 'rapist'. Such behavioural interpretations do not adequately take into account common unwanted experiences of sexual violence that many women experience, including coercion, harassment and intimidation. This is indicated by Australian and international studies of the self-reporting of sexual violence by victims, where hidden forms of such violence are shown to be a common experience for many women from a person well known to them. Surveys also reveal how few sexual assaults are reported to the police.

Traditional approaches to preventing sexual violence have been critiqued for placing blame - and responsibility - with the victim rather than the perpetrator of the violence. Further, rape avoidance strategies, under the banner of prevention, can perpetuate the myth that sexual violence occurs mainly in public places, rather than in private spaces.

Concerns about the dangerousness of sex offenders, particularly child sex offenders, have led to interest in treatment programs for such perpetrators. Treatment programs are primarily group programs with adolescent and adult men who are known to have committed sexual offences against women and/or children. In general, the treatment orientation of sex offender programs is the same whether sexual offences are committed against children or adults. The program evaluations show varying levels of success in reducing recidivism, compared to non-treatment offenders. How recidivism is measured and over what time period all impact on determining 'success'. Schweitzer and Dwyer (2003) argued that evaluations of the effectiveness of sex offender treatment programs need to follow convicted sex offenders for 10 years post-release from custody. Hall (1995) took this further by stating that sex offenders may continue to be at risk of sexually re-offending for more than 20 years post-treatment. It is evident that there is a need to follow the offenders over a long period of time to evaluate the effectiveness of sex offender treatment programs.

In seeking to answer the question, 'Does sex offender treatment work?', Marques (1999, p. 438) acknowledged that there has been progress in determining the effectiveness of sex offender treatment programs. Marques argued that there are some specific questions about treatment programs targeting child and adult sex offenders that need to be answered:

• Which treatments work with which kinds of offenders (identified using factors such as who the victims were, the types of offences committed, the personal characteristics of the offender and the social circumstances of the offender)?
• What is the optimal combination of inpatient and aftercare services?
• How do we determine when offenders are ready for less restrictive treatment environments?

Other areas needing greater attention and development in stopping sexual violence include responses to male-to-male sexual violence and sexual violence among same-sex couples.

It does appear that there is some value in continuing to invest in the popular cognitive behavioural programs and in the monitoring and surveillance of sex offenders to deter their violence. We would caution against the strengths-based or good lives approach being adopted at this stage. Although it has worked effectively with those who have been de-powered and victimised for a long time, there is limited evidence that such an approach is valid for those who have chosen to abuse their power through sexual violence towards women and children.

Contemporary gender theorists have attempted to explain sexual violence as the product of unequal power relations, rather than a psychopathological condition. Such explanations of sexual violence have been applied in prevention programs, but treatment programs for sex offenders still overwhelmingly rely on pathological and psychological explanations. Treatment programs generally are geared towards incarcerated and convicted offenders. Thus, it is important to conceptualise intervention programs beyond pathological and conviction approaches. In this respect, social theory, in areas such as masculinity, that have been associated with sexual violence can be better utilised for education, prevention and treatment programs.

In terms of the organisation of tertiary responses to sexual violence, the development of greater links between sex offenders' treatment programs and services for women and children affected by sexual violence could serve to improve the overall strength of response to sexual violence in the community. These organisations are often vastly different in mandate and theoretical underpinnings; however, collaboration does offer a more complete picture of sexual violence.

Prevention programs include strategies such as education programs and public awareness campaigns, although evaluating their effectiveness is often inadequate due to small sample sizes and reliance on male participants to self-report behaviour. To some extent, the federal government campaign 'Violence Against Women: Australia says No' has challenged men more broadly to consider acts of violence. In addition, men need to be given alternatives to the message of 'don't do it'. There remains a need for greater attention to be given to attitudes and values that contribute to the tolerance of sexual violence and exploitation of women. Future programs should not only focus on explicit acts of sexual violence against women, but also on a range of attitudes, behaviours and acts of sexual violence that remain hidden or accepted as normal heterosexual practice. In order to address these more subtle forms of sexual violence, programs should more critically address social conditions which condone and minimise such violence. Furthermore, they need to offer alternatives of what constitutes respectful and non-violent relationships.

Importantly, prevention activities must also offer alternative tangible and sustainable ways of having an intimate relationship in a society where gender inequality in heterosexual relationships is presented in popular culture as normal. In this respect, Carmody's (2003b) innovative ideas about ethical sexuality offer the possibility of a model that develops a script for equality and non-violence in intimate relationships. Carmody argued that this involves young people having discussions about what they want in their relationships, as opposed to just having conversations about what they do not want. She argued that encouraging ethical sexuality among young people has the potential to reduce sexual violence.

A dichotomy has continued in thinking about male perpetrators of sexual violence, in particular that men who perpetrate sexual violence only include those who are legally convicted 'sex offenders'. All other men are, by inference, not sexually violent or abusive. This ignores those men, often in intimate heterosexual relationships, who coerce, pressure, harass and assault women. Because these behaviours are often considered 'normal sexual relations', they can be legitimised and overlooked, but these micro-practices of sexual violence should also be the focus of change. While there is much work to be done in reform to stop sexual violence, preventing some forms of male sexual violence requires social and personal change that is broad ranging.

There continues to be a strong legacy of attributing blame to the victim. Placing responsibility on the offender and making visible the social conditions which perpetuate this continue to be a challenge, particularly when social attitudes and social policy trends favour the individualisation of social problems. This requires a fundamental shift in focus to consider the centrality of gender inequality as the key problem in the continuation of sexual violence.

In this paper, our aim was to stimulate debate about how the elimination of sexual violence requires attention to be given across the spectrum of sexually violent behaviours and the social conditions which perpetuate sexual violence. This paper has been based on reviewing literature about male sex offenders' violence towards women. The scope of the paper has not included male-to-male sexual violence and sexual violence in same-sex couples, but these are areas requiring future research and program development.

Current prevention strategies and services are heavily weighted towards the very small number (around 15 per cent) of sexual offences reported to police and the offenders pursued by the criminal justice system. Important to this debate is the need for greater recognition of how sexual violence is routinely condoned within dominant and popular representations of masculine culture. These representations have a profound effect on 'ordinary men' and their 'everyday' relationships with women, whether they be a father, partner, relative, friend or trusted professional. In taking this into account, intervention to stop sexual violence at all levels cannot be separated from critical approaches that take into account gender power relations and constructions of masculinity.

In conclusion, we believe this paper makes a convincing argument for a more comprehensive and intersectoral approach to the treatment and prevention of sexual violence. Such an approach requires a greater collaboration between victim services, offender treatment programs, prevention initiatives and the criminal justice system. Importantly, it has underlined the need for further discussion about how to develop social policy and programs that can simultaneously work towards the elimination of sexual violence through changes at the individual, institutional and social levels.

6. Appendices

Appendix A - Australian tertiary sex offender treatment programs for young offenders

Adapted from Flanagan (2003).

Appendix B - Australian tertiary sex offender treatment programs for adult offenders

Adapted from Lievore (2004).

7. References

• Abbey, A., Ross, L., McDuffie, D., & McAuslan, p. (1996). Alcohol and dating risk factors for sexual assault among college women. Psychology of Women Quarterly, 20, 147-169.
• Abel, G., & Rouleau, J. (1990). The nature and extent of sexual assault. In W. Marshall, D. Laws, & H. Barbaree (Eds.), Handbook of sexual assault: Issues, theories, and treatment of the offender (pp. 9-12). New York: Plenum Press.
• Abracen, J., & Looman, J. (2004). Issues in the treatment of sexual offenders: Recent developments and directions for the future. Aggression and Violent Behaviors, 9, 229-246.
• Aljazireh, L. (1993). Historical, environmental and behavioural correlates of sexual offending by male adolescents: A critical review. Behavioural Science and the Law, 11, 423-440.
• American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders, 4th edn., American Psychiatric Association, Washington.
• Australian Bureau of Statistics. (1996). Women's Safety Australia (Cat. No. 4128.0). Canberra: Australian Bureau of Statistics.
• Australian Bureau of Statistics. (2002a). National Crime and Safety Survey (Cat. No. 4509.0). Canberra: Australian Bureau of Statistics.
• Australian Bureau of Statistics. (2002b). Prisoners in Australia (Cat. No. 4517.0). Canberra: Australian Bureau of Statistics.
• Australian Bureau of Statistics. (2004a). Recorded Crime: Victims, Australia (Cat. No. 4510.0). Canberra: Australian Bureau of Statistics.
• Australian Bureau of Statistics. (2004b). Sexual assault in Australia: A statistical overview (Cat. No. 4523.0). Canberra: Australian Bureau of Statistics.
• Australian Government Office for Women. (2005). Education resource pack: Identifying and avoiding abusive and violent relationships, Years 11 and 12. Canberra: Violence Against Women: Australia Says No Campaign, An Australian Government Initiative.
• Bart, p. (1981). A study of women who both were raped and avoided rape. Journal of Social Issues, 37(4), 123-137.
• Bart, P., & O'Brien, p. (1984). Stopping rape: Effective avoidance strategies. Signs: Journal of Women in Culture and Society, 10(1), 83-101.
• Bateman, p. (1991). The context of date rape. In B. Levy (Ed.), Dating violence: Young women in danger. Seattle: Seal Press.
• Beck, A., & Mumola, C. (1999). Prisoners in 1998. Washington, DC: US Department of Justice, Bureau of Justice Statistics.
• Becker, J. V., Cunningham-Rathner, J., & Kaplan, M. (1986). Adolescent sexual offenders: Demographics, criminal and sexual histories, and recommendations for reducing future offences. Journal of Interpersonal Violence, 1(4), 431-435.
• Becker, J. V., & Johnson, B. R. (2001). Treating juvenile sex offenders. In J. Ashford, B. Sales, & W. Reid (Eds.), Treating adult and juvenile offenders with special needs. Washington, DC: American Psychological Association.
• Benedict, J. (1998). Athletes and acquaintance rape. Thousand Oaks, CA: Sage.
• Bradford, J. (1990). Can paedophilia be treated? The Harvard Mental Health Letter, 10(8), 8.
• Bradford, J. (1993). The pharmacological treatment of the adolescent sex offender. New York: Guilford Press.
• Brehm, S., & Kassin, S. (1989). Social psychology. Boston, MA: Houghton Mifflin Company.
• Breitenbecher, K., & Scarce, M. (1999). A longitudinal evaluation of the effectiveness of a sexual assault education program. Journal of Interpersonal Violence, 14(5), 459-478.
• Bromfield, L., & Higgins, D. (2005). National comparison of child protection systems (Child abuse prevention issues No. 22). Melbourne: National Child Protection Clearinghouse, Australian Institute of Family Studies.
• Brownmiller, S. (1975). Against our will: Men, women and rape. New York: Simon and Schuster.
• Byers, E. S. (1996). How well does the traditional sexual script explain sexual coercion? Review of a program of research. In E. S. Byers, & L. O'Sullivan (Eds.), Sexual coercion in dating relationships. New York: Haworth Press.
• Cann, J., Falshaw L., & Friendship, C. (2004). Sexual offenders discharged from prison in England and Wales. Legal and Criminological Psychology, 9, 1-10.
• Carmody, M. (2003a). New approaches to sexual assault prevention. Practice and Prevention: Contemporary Issues in Adult Sexual Assault in NSW Conference, 12-14 February 2003, Sydney.
• Carmody, M. (2003b). Sexual ethics and violence prevention. Social and Legal Studies, 12(2), 199-216.
• Carmody, M. (2005). Ethical erotics: Reconceptualizing anti-rape education. Sexualities, 8(4), 465-480.
• Carmody, M., & Carrington, K. (2000). Preventing sexual violence? Australian and New Zealand Journal of Criminology, 33(3), 341-361.
• Cautela, J. (1967). Covert sensitization. Psychological Reports, 20, 459-468.
• Chung, D. (2005). Violence, control, romance and gender equality: Young women and heterosexual relationships. Women's Studies International Forum, 28, 445-455.
• Colson, C. (1972). Olfactory aversion therapy for homosexual behavior. Journal of Behavior Therapy and Experimental Psychiatry, 3, 185-187.
• Connell, R. (1995). The men and the boys. Sydney: Allen and Unwin.
• Connell, R. (2000). Masculinities. Cambridge: Polity Press.
• Connolly, M., & Wolf, S. (1995). Services for juvenile sex offenders: Issues in establishing programs. Australian Social Work, 48(3), 3-10.
• Cook, B., David, F., & Grant, A. (2001). Sexual violence in Australia (Australian Institute of Criminology Research and Public Policy Series No. 36). Canberra: Australian Institute of Criminology.
• Cossins, A. (2000). Masculinities, sexualities and child sexual abuse. The Hague: Kluwer Law.
• Curnow, R., Streker, P., & Williams, E. (1998). Juvenile justice report: Male adolescent program for positive sexuality. Melbourne: Department of Human Services, Victoria.
• Davis, G., & Leitenberg, H. (1987). Adolescent sex offenders. Psychological Bulletin, 101(3), 417-427.
• Daws, L., Brannock, J., Brooker, R., Patton, W., Smeal, G., & Warren, S. (1995). Young people's perceptions of and attitudes to sexual violence, Hobart: National Clearinghouse for Youth Studies.
• Doren, D. (2002). Evaluating sex offenders: A manual for civil commitments and beyond. Thousand Oaks, CA: Sage Publications.
• Fernandez, J., & Loh, N. (2001). Crime and justice statistics for Western Australia, 2001. Perth: Crime Research Centre, University of Western Australia.
• Finkelhor, D., & Lewis, I. (1988). An epistemological approach to the study of child molestation. Annals of the New York Academy of Sciences, 528, 64.
• Flanagan, K. (2003). Intervention with sexually abusive young people in Australia and New Zealand. Journal of Sexual Aggression, 9(2), 135-149.
• Flood, M. (2002-2003) Engaging men: Strategies and dilemmas in violence prevention education among men. Women Against Violence, 13, 25-32.
• Foshee, V., Bauman, K., Arriaga, X., Helms, R., Koch, G., Linder, G., & Fletcher, G. (2004). Assessing the long term effects of the Safe Dates Program and a booster in preventing and reducing adolescent dating violence victimization and perpetration. American Journal of Public Health, 94(4), 619-624.
• Gebhard, p. H., Gagnon, J. H., Pomeroy, W. B., & Christenson, C.V. (1965). Sex offenders: An analysis of types. New York: Harper and Row.
• Glaser, B. (1997). Paedophilia: The public health problem of the decade. In M. James (Ed.), Paedophilia, policy and prevention (Research and Public Policy Series No. 12). Canberra: Australian Institute of Criminology.
• Grant, A. (2000). The historical development of treatment for adolescent sex offenders (Trends and Issues in Crime and Criminal Justice No. 145). Canberra: Australian Institute of Criminology.
• Greenberg, D., Da Silva, J., & Loh, N. (2002). Evaluation of the Western Australian Sex Offender Treatment Unit (1987-1999): A quantitative analysis. Perth: Forensic Research Unit, University of Western Australia.
• Greenfeld, L. (1997). Sex offences and offenders: An analysis of data and sexual assault. Washington, DC: US Department of Justice, Bureau of Justice Statistics.
• Hall, G. (1995). Sex offender recidivism revisited: A meta-analysis of recent treatment studies. Journal of Consulting and Clinical Psychology, 63, 802-809.
• Hanson, K., & Gidycz, C. (1993). Evaluation of a sexual assault prevention program. Journal of Consulting and Clinical Psychology, 61(6), 1046-1052.
• Hanson, R. K., Gordon, A., Harris, A. J. R., Marques, J. K., Murphy, W., Quinsey, V. L., & Seto, M. (2002). First report of the collaborative outcome data project on the effectiveness of psychological treatment for sex offenders. Sexual abuse: A journal of research and treatment, 14(2), 169-194.
• Harrison, P., & Beck, A. (2003). Prisoners in 2002. Washington, DC: US Department of Justice, Bureau of Justice Statistics.
• Heath, M. (2005). The law and sexual offences against adults in Australia (Issues No. 4). Melbourne: Australian Centre for the Study of Sexual Assault, Australian Institute of Family Studies.
• Heppner, M., Neville, H., Smith, K., Kivlighan, D., & Gershuny, B. (1999). Examining immediate and long-term efficacy of rape prevention programming with racially diverse college men. Journal of Counselling Psychology, 46(1), 16-26.
• Hird, M. (2000). An empirical study of adolescent dating aggression in the United Kingdom. Journal of Adolescence, 23, 69-78.
• Hoffman, L. (1993). Exchanging voices: A collaborative approach to family therapy. London: Karnac.
• Howells, K., Heseltine, K., Sarre, R., Davey, L., & Day, A. (2004). Correctional offender rehabilitation programs: The national picture in Australia. Adelaide: Forensic Psychology Research Group, University of South Australia.
• James, M. (1996). Paedophilia (Trends and Issues in Crime and Criminal Justice No. 57). Canberra: Australian Institute of Criminology.
• Jenkins, A. (1990). Invitations to responsibility: The therapeutic engagement of men who are violent and abusive. Adelaide: Dulwich Centre Publications.
• Jones, E. (1993). Family systems therapy: Developments in Milan Systematic Therapies. Chichester: Wiley.
• Keel, M. (2005). Prevention of sexual assault: Working with adolescents within the education system. Aware: ACSSA Newsletter, 8, 16-25.
• Kitzinger, C., & Frith, C. (1999). Just say no? The use of conversation analysis in developing a feminist perspective on sexual refusal. Discourse and Society, 10(3), 293-316.
• Kleck, G., & Sayles, S. (1990). Rape and resistance. Social Problems, 37(2), 149-162.
• Koss, M., Gidycz, C., & Wisniewski, N. (1987). The scope of rape: Incidence and prevalence of sexual aggression and victimization in a national sample of higher education students. Journal of Consulting and Clinical Psychology, 55, 162-170.
• Lab, S., Shields, G., & Schondel, C. (1993). Research note: An evaluation of juvenile sexual offender treatment. Crime and Delinquency, 39(4), 543-553.
• Lancaster, E., & Lumb, J. (1999). Bridging the gap: Feminist theory and practice reality in work with the perpetrators of child sexual abuse. Child and Family Social Work, 4, 119-129.
• Law Reform Commission of Victoria. (1988). Sexual offences against children (Report No. 18). Melbourne: Law Reform Commission of Victoria.
• Laws, D. (2003). The rise and fall of relapse prevention. Australian Psychologist, 38(1), 22-30.
• Laws, D., & Marshall, W. (2003). A brief history of behavioral and cognitive behavioral approaches to sexual offenders: Part 1, Early developments. Sexual Abuse: A Journal of Research and Treatment, 15(2), 75-92.
• Lievore, D. (2004). Recidivism of sexual assault offenders: Rates, risk factors and treatment efficacy. Canberra: Australian Institute of Criminology.
• Lievore, D. (2005). Thoughts on recidivism and rehabilitation of rapists. University of New South Wales Law Journal Forum, 11(1), 293-298.
• Lonsway, K., Klaw, E., Berg, D., Waldo, C., Kothari, C., Mazurek, C., & Hegeman, K. (1998). Beyond 'No Means No': Outcomes of an intensive program to train peer facilitators for campus acquaintance rape education. Journal of Interpersonal Violence, 13(1), 73-92.
• Malamuth, N. (1989). The attraction to sexual aggression scale: Part two. Journal of Sexual Assault, 26, 324-354.
• Mammone, N., Keeling, J., Sleeman, V., & McElhore M. (2002). A preliminary evaluation of the impact of an intensive treatment programme on the cognitive distortions of sexual offenders. Second Biennial International Conference of the Australian and New Zealand Association for the Treatment of Sexual Abusers, 4-6 April 2002, Sydney.
• Mandeville-Norden, R., & Beech, A. (2004). Community-based treatment of sex offenders. Journal of Sexual Aggression, 10(2), 193-214.
• Mann, R. E. (2004). Innovations in sex offender treatment. Journal of Sexual Aggression, 10(2), 141-152.
• Marques, J. (1999). How to answer the question: Does sex offender treatment work? Journal of Interpersonal Violence, 14(4), 437-451.
• Marshall, W. (1993). The role of attachment, intimacy and loneliness in the etiology and maintenance of sexual offending. Sexual and Marital Therapy, 8, 109-121
• Marshall, W. (1999). Current status of North American assessment and treatment programs for sexual offenders. Journal of Interpersonal Violence, 14(3), 221-239.
• Marshall, W., & Barbaree, H. (1990). Outcome of comprehensive cognitive-behavioural treatment programs. In W. Marshall, D. Laws, & H. Barbaree (Eds.), Handbook of sexual assault: Issues, theories, and treatment of the offender. New York: Plenum Press.
• Marshall, W., & Laws, D. (2003). A brief history of behavioral and cognitive behavioral approaches to sexual offender treatment: Part 2, The modern era. Sexual Abuse: A Journal of Research and Treatment, 15(2), 93-120.
• Marshall, W., Ward, T., Mann, R., Moulden, H., Fernandez, Y., Serran, G., & Marshall, L. (2005). Working positively with sexual offenders: Maximizing the effectiveness of treatment. Journal of Interpersonal Violence, 20(9), 1096-1114.
• McGuire, J., & Priestly, p. (1985). Offending behaviour: Skills and stratagems for going straight. London: Batsford.
• McIntosh, I., & Griffin, A. (2001). Teenage tolerance: A study of young people's experience and responses to violence and abuse. London: Women's Aid.
• Mouzos, J., & Makkai, T. (2004). Women's experiences of male violence: Findings from the Australian component of the International Violence Against Women Survey. Canberra: Australian Institute of Criminology.
• Mulroney, J. (2003). Prevention programs for young people that promote healthy relationships. Practice and Prevention: Contemporary Issues in Adult Sexual Assault in NSW Conference, 12-14 February 2003, Sydney.
• Neame, A. (2003). Differing perspectives on 'preventing' adult sexual assault. Aware: ACSSA Newsletter, 2, 8-14.
• Neame, A., & Heenan, M. (2003). What lies behind the hidden figure of sexual assault? Issues of prevalence and disclosure (Briefing No. 1). Melbourne: Australian Centre for the Study of Sexual Assault, Australian Institute of Family Studies.
• Office of Crime Statistics and Research. (2002). Crime and justice in South Australia 2001: Offences reported to police, the victims and alleged perpetrators, A statistical report. Adelaide: Attorney-General's Department.
• O'Leary, P., Chung, D., & Zannettino, L. (2004). National comparative assessment of programs for men who are violent towards their female partners. Canberra: Office for Status of Women, Department of the Prime Minister and Cabinet.
• Parissa, S. (2002). Boys behaving badly: Popular literature on the mis/behaviour of male team sports athletes in North America. Sociology of Sport, 27(1), 97-102.
• Parrot, A. (1991). Institutional response: How can acquaintance rape be prevented? In A. Parrot & L. Bechhofer (Eds.), Acquaintance rape: The hidden crime. New York: Wiley.
• Patton, W., & Mannison, M. (1995). Sexual coercion in high school dating. Sex Roles, 33(5), 447-457.
• Pease, B. (2004-2005). Rethinking profeminist men's behaviour change programs. Women Against Violence: An Australian Feminist Journal, 16, 32-40.
• Pithers, W. (1990). Relapse prevention with sexual aggressors: A method for maintaining therapeutic gain and enhancing external supervision. In W. Marshall, D. Laws, & H. Barbaree (Eds.), Handbook of sexual assault: Issues, theories, and treatment of the offender. New York: Plenum Press.
• Plummer, K. (1991). Understanding childhood sexualities. In T. Sandfort, E. Brongersma, & A. van Naerssen (Eds.), Male intergenerational intimacy: Historical, socio-psychological, and legal perspectives. New York: Harrington Park Press.
• Queensland Crime and Misconduct Commission. (2003). Seeking justice: An inquiry into how sexual offences are handled by the Queensland criminal justice system. Brisbane: Crime and Misconduct Commission.
• Queensland Police Service. (2001). Annual statistical review 2000-2001. Brisbane: Queensland Police Service.
• Quinsey, V., & Earls, C. (1990). The modification of sexual preferences. In W. Marshall, D. Laws, & H. Barbaree (Eds.), Handbook of sexual assault: Issues, theories, and treatment of the offender. New York: Plenum Press.
• Quinsey, V., & Marshall, W. (1983). Procedures for reducing inappropriate sexual arousal: An evaluation review. In G. Greer, & I. Stuart (Eds.), The sexual aggressor: Current perspectives on treatment. New York: Van Nostrand Reinhold.
• Reavey, p. (Ed.). (2003). New feminist stories of child sexual abuse. London: Routledge.
• Russo, L. (2000). Date rape: A hidden crime (Trends and Issues in Crime and Criminal Justice No. 157). Canberra: Australian Institute of Criminology.
• Saleebey, D. (2002). The strengths perspective in social work practice (3rd ed.). Boston, MA: Allyn and Bacon.
• Sapp, A., & Vaughn, M. (1990). Juvenile sex offender treatment at state-operated correctional institutions. International Journal of Offender Therapy and Comparative Criminology, 34, 131-146.
• Schewe, P., & O'Donohue, W. (1993). Rape prevention: Methodological problems and new directions. Clinical Psychology Review, 13, 667-682.
• Schwartz, B. K., & Cellini, H. R. (1997). The sex offender: New insights, treatment innovations and legal development. Kingston, NJ: Civic Research Institute.
• Schweitzer, R., & Dwyer, J. (2003). Sex crime recidivism: Evaluation of a sexual offender treatment program. Journal of Interpersonal Violence, 18(11), 1292-1310.
• Serber, M. (1970). Shame aversion therapy. Journal of Behavior Therapy and Experimental Psychiatry, 1, 213-215.
• Shepard, M., & Pence, E. (Eds.). (1999). Coordinated community responses to domestic violence: Lessons from Duluth and beyond. Thousand Oaks, CA: Sage.
• Smallbone, S., & Ransley, J. (2005). Legal and psychological controversies in the preventive incapacitation of sexual offenders. University of New South Wales Law Journal Forum, 11(1), 299-305.
• Smith, C., Huxley, J., & McKernan, K. (2004). Kinks, cubes and cyberspace: Working with young people to prevent sexual violence. Home Truths: Stop Sexual Assault and Domestic Violence: A National Challenge, 15-17 September 2004, Melbourne.
• Smith, S., Wampler, R., Jones, J., & Reifman, A. (2005). Differences in self-report measures by adolescent sex offender risk group. International Journal of Offender Therapy and Comparative Criminology, 49(1), 82-106.
• Sundermann, M., & Jaffe, p. G. (1995). Violence prevention in secondary (high) schools. In E. Peled, p. Jaffee, & J. L. Edleson (Eds.), Ending the cycle of violence: Community responses to children of battered women. Thousand Oaks, CA: Sage Publications.
• Taylor, C. (2004). Court licensed abuse: Patriarchal lore and the legal response to intrafamilial abuse of children. New York: Peter Lang.
• Tomison, A., & Poole, L. (2000). Preventing child abuse and neglect: Findings from an Australian audit of prevention programs. Melbourne: National Child Protection Clearinghouse, Australian Institute of Family Studies.
• Urbis Keys Young. (2004). National framework for sexual assault prevention. Canberra: Commonwealth of Australia.
• Victorian Law Reform Commission. (2003). Sexual offences interim report. Melbourne: Victorian Law Reform Commission.
• Victorian Police. (2002). Crime statistics 2000-01. Melbourne: Victorian Police, Melbourne.
• Vivian-Byrne, S. (2002). Using context and difference in sex offender treatment: An integrated systematic approach. Journal of Sexual Aggression, 8, 59-73.
• Vivian-Byrne, S. (2004). Changing people's minds. Journal of Sexual Aggression, 10(2), 181-192.
• Ward, T., & Beech, A. (2006). An integrated theory of sexual offending. Aggression and Violent Behavior, 11, 44-63.
• Weisz, A., & Black, B. (2001). Evaluating a sexual assault and dating violence prevention program for urban youths. Social Work Research, 25(2), 89-100.
• Western Australia Crime Research Centre and Donovan Research. (2001). Young people and domestic violence: National research on young people's attitudes and experiences of domestic violence. Canberra: Western Australia Crime Research Centre and Donovan Research.
• White, M. (2000). Reflections on narrative practice: Essays and interviews, Adelaide: Dulwich Centre Publications.
• Wortley, R., & Smallbone, S. (2006). Applying situational principles to sexual offences against children. In R. Wortley, & S. Smallbone (Eds.), Situational prevention and child sexual abuse. Annandale: Federation Press.
• Young, J. E. (1990). Cognitive therapy for personality disorders: A schema-focused approach. Sarasota, FL: Professional Resource Press.