Evaluating the legal assumptions of Victoria’s Sex Offender Registration Act 2004 from a psychological perspective

Abstract

The Sex Offender Registration Act 2004 was introduced in Victoria to decrease recidivism and aid in future investigations and prosecutions. This article reviews literature to evaluate four assumptions inherent to the Act: (a) sexual offenders are more dangerous than non-sexual offenders; (b) sexual offenders who target children are more dangerous than those who target adults; (c) recidivism risk can be accurately assessed for sexual offenders who target adults; and (d) the Act is a useful tool for investigations and prosecutions. The findings suggest that there is little evidence that supports the assumptions. Further, given the relatively narrow scope of the Act, it is unlikely to have a positive impact on the safety of the community.

Within the last decade, there has been a trend amongst Western countries to adopt sexual offender registries as a way of preventing and/or reducing sex crimes. The scope and confidentiality of sexual offender registries vary internationally. While some jurisdictions track offenders with GPS technology and release their personal and professional information to the public (Turner, Chamberlain, Jannetta, & Hess, 2015), in other jurisdictions, registries are police databases that can identify sexual offenders if they commit future crimes, but offenders’ personal information is withheld from the public (Powell, Day, Benson, Vess, & Graffam, 2014).

The Sex Offenders Registration Act 2004 (Vic) (Austl.)

In Victoria, the Sex Offenders Registration Act 2004 (Vic) (Austl.) (herein referred to as the Act) governs the Registry, which was introduced with the aim of reducing the likelihood of recidivism and to aid in investigation and prosecution of future offences. The Act categorises offences into four classes. Class 1 and Class 2 offences include bestiality and offences that were committed against a child, whereas Class 3 and Class 4 offences are serious (i.e. two or more) sexual offences committed against any person who is not a child. The Act dictates that offenders convicted of a Class 1 or Class 2 offence must automatically be registered (s.7). Comparatively, offenders who are convicted of Class 3 or Class 4 offences are only to be registered if they are deemed to ‘pose a risk to the sexual safety of one or more persons in the community’ (s.11.3). Notably, the Sex Offenders Registration Amendment (Miscellaneous) Act 2017 (Vic) (Austl.) eliminated mandatory registration for young adult offenders who committed their offence(s) when 18 or 19 years old (s. 7.1).

Under the Sex Offender Registration Act 2004, registered offenders over the age of 18 years old are subject to reporting periods of 8 years, 15 years, or lifetime depending on the class and number of offences (s. 34). During the reporting period, sexual offenders are required to inform police of their whereabouts, and their personal details are stored in a database for future investigations (s. 14). After the reporting period lapses, offenders are considered to be ‘deregistered’. However, there is some controversy surrounding this terminology as registered offenders are forbidden from partaking in child-related employment for life (s. 11.11), meaning that they effectively remain in the registered sexual offender database indefinitely. Notably, this restriction is placed on all offenders, regardless of whether they victimised children.

Compared to other jurisdictions (e.g. Turner et al., 2015), the Victorian registry is relatively passive, as sexual offenders’ whereabouts are not monitored electronically, and their personal information is not made public. However, recent amendments to the Act (Sex Offenders Registration Amendment Act 2014 (Vic) (Austl.); Sex Offenders Registration Amendment Act 2016 (Vic) (Austl.); Sex Offenders Registration Amendment (Miscellaneous) Act 2017 (Vic) (Austl.)), have made the Registry increasingly active, as new powers were introduced enabling police to prohibit offenders from engaging in specific behaviours (e.g. prohibiting alcohol use, excluding them from specific locations, restricting contact with specified others, etc.; Sex Offenders Registration Amendment (Miscellaneous) Act 2017 (Vic) (Austl.) s. 66Q.2). Further, these amendments have been retroactively applied to all registered offenders, raising ethical concern regarding how future amendments may result in added restrictions placed upon offenders well after they have been sentenced, released from prison, or have completed their reporting obligations, as they are never truly deregistered. Although this has not been challenged in Victoria, retrospectively applied sexual offender registry laws in the United States have been met with several legal challenges, some of which have found that the retroactive application of sex offender registry laws is unconstitutional (U.S. Department of Justice, 2018).

As of the end of the 2017–2018 financial year, 7799 sexual offenders had been placed on the Registry in Victoria, with 552 sexual offenders added in the past year (Victoria Police, 2018). More than 1500 sexual offenders (20%) have been placed on the Registry for life. However, since the Registry began, only 989 sexual offenders have been ‘deregistered’ (i.e. completed their reporting obligations; Victoria Police, 2018). The difference between the number of offenders who were registered in a single year and the total number of offenders who have been deregistered in the past 13 years highlights the overwhelming growth of the Registry.

Evaluation of the registry

Within Victoria, the Act has not been evaluated to investigate whether it improves community safety as research on the Registry is forbidden in the legislation (Sex Offender Registration Act 2004 s. 63.1.b). The conditions of the Act were reviewed by the Victoria Law Reform Commission (VLRC; 2012) and 79 recommendations were made. Of those, only eight were taken into consideration in the Sex Offender Registration Act 2014, which allowed judges more discretionary power when placing minors on the Registry (Recommendation 19), enabled individuals who were cognitively or physically impaired to be removed from the Registry if they could not meet their obligations (Recommendation 26), clarified the definition of ‘contact’ with a child (Recommendation 32), and provided greater powers to prohibit contact with children and share information (Recommendations 31, 55–59). However, most of the suggestions to improve the Act were disregarded.

Internationally, there is limited empirical evidence suggesting that sex offender registries have any impact on sexual offending (Bouffard & Askew, 2017; see Terry, 2015, for review). Of the few studies that have found support for sex offender registries, one highlighted that passive registries, such as Victoria’s, are more likely to be successful than active registries (Prescott & Rockoff, 2011). While it is difficult to determine whether the Victorian Registry is effective, it is possible to evaluate whether the existence of the Registry or the registration decision-making process is supported by empirical evidence.

There are many assumptions that underpin the very existence of the Registry, as well as the decision-making process that determines who is registered according to the Act. This article focuses on four legal assumptions to evaluate whether the Act is empirically valid. Specifically, this paper investigates: (a) whether sexual offenders pose a greater risk to the safety of the community than other groups of offenders who are not registered (e.g. violent offenders); (b) whether the mandatory registration of Class 1 and Class 2 offenders (i.e. offenders who victimise children) is because these offenders pose a greater risk to the community than Class 3 or Class 4 offenders (i.e. offenders who victimise adults); (c) whether it is possible to reliably determine whether Class 3 or Class 4 offenders are a risk to the sexual safety of the community; (d) whether the Registry is able to aid in the investigation and prosecution of future sexual offences. Only literature focusing on adult male sexual offenders was reviewed due to the low proportions of female sexual offenders who are registered in Victoria (2%; Victoria Police, 2018) and the different registration rules for juveniles under the Act.

Do sexual offenders pose a greater risk than other offenders?

Sexual offenders are the only type of offenders who are required to register their personal information in a police database in an attempt to reduce recidivism and to aid future investigations and prosecution. The Act aims to register individuals who pose a risk to the sexual safety of the community, but it does not appear to consider the general safety of the community. This emphasis on sexual safety, rather than general safety, implies that sex crimes cause a greater impact than general or violent crimes. However, defining and measuring the impact of crime on victims can be difficult as the operationalisation of ‘impact’ can vary greatly.

In regard to the psychological and physical impact of sexual assault, it is well established that individuals who have been sexually assaulted experience significantly greater somatic and psychiatric disorders than the general population (Chen et al., 2010; Cutajar et al., 2010; Fergusson, Boden, & Horwood, 2008; Papalia, Luebbers, Ogloff, Cutajar, & Mullen, 2017; Paras et al., 2009). However, the relationship is complicated as factors such as the age of victimisation, sex of victim (Papalia et al., 2017), and the relationship to the perpetrator (Golding, Cooper, & George, 1997) can affect the likelihood of experiencing adverse outcomes.

A meta-analysis investigating the relationship between adverse health outcomes and sexual assault found victims of sexual assault in childhood and adulthood did not differ in regard to their negative health outcomes (Golding, Cooper, & George, 1997). However, individuals who were victimised by strangers or romantic partners were more likely to report poor health. In contrast, research examining the prevalence of psychological disorders such as post-traumatic stress disorder (PTSD), depression, or eating disorders found that rape victims were more likely to report these disorders than other sexual assault victims (Chen et al., 2010). Similarly, Kilpatrick et al. (1989) found that, compared to victims of other crimes (including violent crimes), victims of rape were significantly more likely to develop PTSD. These findings suggest that while sexual assault victims have greater physical and psychological distress than the general population, sexual assault in childhood does not necessarily have a significantly greater impact than sexual assault in adulthood. Indeed, rape victims reported significantly higher rates of PTSD than victims of other crimes or types of sexual assault, despite the fact that the Act does not specifically focus on protecting the community from these types of offenders.

Another way to determine the impact of offending is to consider how widespread the offending behaviour is. Both police data (Victoria Police, 2018) and the Personal Safety Survey (Australian Bureau of Statistics, 2016), which investigates the rate of self-reported victimisation, suggest that sexual offending (2.5%; 5–18%, respectively) occurs less often than violent offending (8.2%; 31–41%, respectively) in Australia. While sexual offences undoubtedly have significant consequences for victims, these statistics suggest that crime prevention strategies with the sole focus of preventing sexual offences are likely to have a less overall impact on the community because these offences are less common.

In addition to assessing the prevalence of sexual offending as a proxy for the impact of sexual crimes, it is also useful to consider the rate of sexual reoffending as the Registry specifically aims to reduce recidivism, not first-time sexual offending. Research suggests that the base-rate for sexual recidivism is approximately 10–15% after a five-year follow-up (Barbaree, Seto, Langton, & Peacock, 2001; Hanson & Bussiere, 1998; Hanson & Morton-Bourgon, 2004; Helmus, Thornton, Hanson, & Babchishin, 2012) and up to 24% after 15 years (Harris & Hanson, 2004). Notably, definitions of recidivism vary across research, ranging from rearrest to conviction. Langan and Levin (2002) highlighted the need to consider the definition of ‘recidivism’ when interpreting research findings. For instance, using a more inclusive definition of recidivism (i.e. rearrest within three years), Langan and Levin found that 42% of all sexual offenders released from prison in 15 states recidivated. Comparatively, if convictions were used to define recidivism, only 22% of prisoners in the same sample would be considered recidivists. Hanson and colleagues’ (Hanson & Morton-Bourgon, 2004; Harris & Hanson, 2004; Helmus et al., 2012) meta-analyses of sexual offending research integrated data using different definitions of recidivism. As such, the above estimates of recidivism are aggregated across reconvictions, charges, and rearrests.

Although sexual offenders are singled out in the Act, research suggests that they are less likely to recidivate than violent offenders (Craig, Beech, & Browne, 2006; Stalans, Yarnold, Seng, Olsen, & Repp, 2004). For instance, a study from the United Kingdom that assessed both sexual and violent offenders found that after the 10-year follow-up period, violent offenders were twice as likely to have a new conviction for any offence compared to sexual offenders (Craig et al., 2006). Of the sexual offenders who were reconvicted of another crime, 50% had convictions that were for non-sexual offences (Craig et al., 2006). In fact, research shows that non-sexual recidivism is the most prevalent form of recidivism for sexual offenders (Helmus et al., 2012; Looman & Abracen, 2010). Further, research suggests that 80–95% of sexual offences are perpetrated by individuals who have not previously been detected by the justice system (Rettenberger, Briken, Turner, & Eher, 2015; Sandler, Freeman, & Socia, 2008).

Therefore, regardless of the impact of sexual offending on the community, the Registry is rather narrow in scope as it aims to protect society from a relatively small proportion of individuals. Indeed, it is estimated that only 20% of those who commit sexual offences are detected, and only 1% receive convictions (Sentencing Advisory Council, [SAC], 2007). Of those who are convicted, approximately 5–20% are detected for reoffending (Rettenberger et al., 2015; Sandler et al., 2008). Therefore, despite the significant psychological impact of sexual offences, it is unlikely that registered offenders pose a greater overall risk to the community than violent offenders, who account for a greater proportion of the offender population and have higher rates of recidivism.

Are child sexual offenders more dangerous than those who victimise adults?

In addition to assuming that sexual offenders pose a greater risk to the community, inherent in the Act is the suggestion that certain sexual offenders pose a greater risk than others. The mandatory registration of offenders who victimise children (i.e. Class 1 and Class 2 offenders) implies that child sexual offenders are perceived to be of greater risk than those who offend again adults (i.e. Class 3 and Class 4 offenders). These two offender types are perceived to differ so significantly that as of 2012 only 3% of all registered offenders had committed a Class 3 or Class 4 offence (Victoria Law Reform Commission, 2012). Although more recent data are not available on registered offenders, these statistics suggest that while the Act considers all child sexual offenders as a risk to the community, relatively few sexual offenders who target adults are considered to be a high-risk to the sexual safety of the community. However, research that has compared sexual offenders who target children to those who target adults does not support the assumption that child sex offenders are inherently riskier.

In a New Zealand study examining the recidivism rates of different types of sexual offenders (i.e. child victim under the age of 16, adult victim, or mixed victims; Vess & Skelton, 2010), there were no significant differences in rates of recidivism after a 15-year follow-up between offender types. Similarly, a meta-analysis of 4724 sexual offenders found that overall, child molesters and rapists did not significantly differ in their rates of recidivism (Harris & Hanson, 2004). Although child sexual offences may be less likely to be reported to the police (Australian Bureau of Statistics, 2019; Office of Juvenile Justice and Delinquency Prevention, 2012), these results provide tentative evidence that sexual offenders who target adults are at least as risky as those who target children.

Notably, broadly classifying offenders based upon their victims’ age can obscure important differences in risk according to victim choice. Research suggests that both child sexual offenders and rapists are not homogeneous groups (Robertiello & Terry, 2007; Yates & Kingston, 2006). However, recidivism studies generally only examine variation in recidivism for different types of child sexual offenders, as there are comparatively few rapists (Harris & Hanson, 2004; Reeves et al., 2017).

Child sexual offenders may be classified according to their relationship to victims (intra-familial or extra-familial) or the victims’ gender. With regard to the relationship to victim, a 25-year follow-up study of 320 sexual offenders found that approximately 50% of intra-familial child molesters and rapists were reconvicted of a sexual offence, compared to 71% of extra-familial child molesters (Langevin et al., 2004). Although the recidivism rates in this study are much higher than most (see Doren, 1998 for review), extra-familial offenders have been consistently found to sexually recidivate at higher rates than intra-familial offenders (Hanson, 2002; Hanson & Bussiere, 1998; Nilsson et al., 2014). While this may partially reflect differences in reporting (Goodman-Brown, Edelstein, Goodman, Jones, & Gordon, 2003) or courts’ decision making, as intra-familial cases of sexual abuse may be more likely to result in acquittal than extra-familial cases (Bergh, 2017), Harris and Hanson (2004) found that incest offenders were almost half as likely as the average sex offender to reoffend. This is hypothesised to be because offenders who actively seek out victims (i.e. extra-familial) exhibit more general criminality than offenders who are opportunistic and abuse their family members (i.e. intra-familial; Seto, Babchishin, Pullman, & McPhail, 2015).

When examining the risk of recidivism based on the gender of the victim, research suggests that child sexual offenders who victimise females have markedly lower rates of sexual recidivism than those who victimise males (Friendship & Beech, 2005; Hanson & Bussiere, 1998; Maletzky & Steinhauser, 2002). The difference in gender appears to be particularly strong when considered within the context of the relationship to the victim. In a meta-analysis of 10 studies and 4724 sexual offenders, intra-familial offenders and extra-familial offenders with female victims had similar rates of recidivism after 15 years. However, extra-familial offenders with male victims had significantly greater rates of sexual recidivism than all other groups including rapists (Harris & Hanson, 2004). This research emphasises the heterogeneity of child sexual offenders. While some child sexual offenders (i.e. offenders who target extra-familial males) may recidivate at greater rates than rapists, most child sexual offenders do not.

Additionally, consideration of the general (rather than sexual) risk that sexual offenders pose to society may also be important when determining whether they should be monitored, particularly as sexual recidivism is not the most common form of recidivism for sexual offenders (Craig et al., 2006; Reeves, Ogloff, & Simmons, 2017). This is particularly true for rapists, as research suggests that while they may sexually recidivate at similar rates to child sexual offenders (Prentky, Lee, Knight, & Cerce, 1997; Rettenberger et al., 2015), they are more likely than child offenders to generally or violently reoffend (Rettenberger, Rice, Harris, & Eher, 2017; Vess & Skelton, 2010). Indeed, research highlights that rapists are typically not just sexual offenders and that they generally pose a high overall risk to the community in terms of both recidivism (Rettenberger et al., 2017) and the impact on victims (Chen et al., 2010).

Overall, sexual offender recidivism research is incongruent with the Act’s assumption that child sexual offenders are more likely to reoffend than adult sexual offenders. While the Registry automatically captures relatively high-risk extra-familial child offenders, it also captures the intra-familial offenders who are, as a group, at the lowest risk of recidivism. This is arguably a waste of resources and an unnecessary violation of low-risk offender’s rights. In fact, as of 2012, 81% of all the sexual offenders on the Registry were listed as either low or medium risk (Victoria Police, 2012). Although Victoria Police no longer report the risk level of Registered offenders, this provides insight into just how overly inclusive the Registry is. These results contradict the Risk, Need, Responsivity (RNR) theory (Andrews & Bonta, 2010), a prominent theory in forensic psychology, which posits that resources should be allocated according to risk level for intervention to be effective.

Can we accurately predict recidivism?

As previously discussed, the Act does not require individuals who perpetrate Class 3 and Class 4 sexual offences (i.e. those against adults) to be automatically registered; rather, the offenders must be assessed and deemed a risk to the community. Implicit in this legislation is the assumption that sexual offending behaviour, particularly behaviour that includes adult victims, can be predicted.

Within Victoria, the Static-99 (Harris, Phenix, Hanson, & Thornton, 2003) is the actuarial risk assessment instrument that is used to predict sexual recidivism (Department of Justice and Regulation, 2015). The Static-99 demonstrates moderate predictive accuracy within Australian (Allan, Dawson, & Allan, 2006; Reeves et al., 2017; Smallbone & Rallings, 2013) and international jurisdictions (Hanson & Morton-Bourgon, 2004; Helmus et al., 2012). However, sexual offending risk assessments are generally normed on large samples of sexual offenders, in which most of the subjects have targeted children, not adults (Harris & Hanson, 2004; Reeves et al., 2017). As such, sexual offending risk assessments are typically less accurate for rapists than for child sexual offenders (Bartosh, Garby, Lewis, & Gray, 2003; Looman & Abracen, 2010; Tully, Chou, & Browne, 2013).

Indeed, in a study comparing the utility of four sexual offender risk assessment tools after a five-year follow-up, the predictive validity differed significantly based upon the offender type (Bartosh et al., 2003). The Minnesota Sex Offender Screening Tool, Revised (MnSOST–R; Epperson, Kaul, Hesselton, 1998), the Sex Offender Risk Appraisal Guide (SORAG; Quinsey, Harris, Rice, & Cormier, 2006), the Rapid Risk Assessment for Sex Offense Recidivism (RRASOR; Hanson, 1997), and the Static-99 were effective at predicting sexual recidivism for the sample as a whole, but failed to predict sexual recidivism for rapists (i.e. victims over the age of 16) and hands-off offenders (e.g. exhibitionists). While the sample size of the hands-off offenders was too small to be generalisable, the poor predictive validity of the Static-99 for rapists was a concern given that the Act requires that adult sexual offenders be assessed prior to being placed on the Registry.

Notably, the risk assessment measures discussed above only predict the likelihood of future sexual violence, not the severity of violence. Structured professional judgement (SPJ) risk assessments such as the Risk of Sexual Violence Protocol (RSVP; Hart, Kropp, Klaver, Logan, & Watt, 2003) include risk judgements estimating the severity of future sexual violence. However, there is limited evidence supporting the accuracy of the risk estimates for the severity of violence. The only known study that has investigated the severity ratings on the RSVP has yet to be peer reviewed. Darjee et al. (2016) found that the risk of future severe violence was effective at predicting future physical violence, but not sexual violence. This underscores the issue of distinguishing violent and sexual offending, when both are harmful behaviours. Further, it is important to consider that offences that may not be considered ‘severe’ can still cause considerable psychological and physical health consequences for victims (Cutajar et al., 2010; Papalia, Luebbers, Ogloff, Cutajar, & Mullen, 2017; Paras et al., 2009)

Is the registry useful for improving the investigation and prosecution of sexual offences?

The Act endeavours to improve the investigation and prosecution of sexual offences. International research has found that both police officers and community members perceive that sexual offender registries are useful for preventing reoffending and aiding in investigations (Cubellis, Walfield, & Harris, 2018; Harris, Levenson, Lobanov-Rostovsky, & Walfield, 2018; Koon-Magnin, 2015). However, there is limited empirical evidence suggesting that sexual offender registries have any impact on sexual offending (Bouffard & Askew, 2017; see Terry, 2015, for review), and few researchers have attempted to investigate whether registries have a quantifiable impact on investigations or prosecutions.

Bierie (2016) provided anecdotal evidence of individual cases in which sexual offenders were reported to the police by citizens who recognised them from the sexual offender registry photos. Although this provides some evidence that public registries may work in a handful of cases, it is not directly applicable to the Victoria Registry, because offenders’ identities are not released to the public. Prescott and Rockoff (2011) investigated whether the introduction of a registry improved the number of cases that were cleared for prosecution, finding that registries did not make a difference. This is not surprising considering that relatively few sexual offences are committed by recidivist offenders (Rettenberger et al., 2015; Sandler et al., 2008); as such the potential impact of sex offender registries on the rates of sexual offences that are prosecuted is likely to be small at best.

Amendments to the Act have increased the amount of information that the police can collect to aid their investigations and prosecutions. In 2018, the Act was amended to include powers to collect fingerprints (s.27) to assist with cold cases and future investigations. Within the first year, 2700 registered offenders were identified for fingerprinting, and 400 provided their fingerprints (Victoria Police, 2018). As this is a new initiative, there are no data regarding whether fingerprinting has translated into solving cold cases or more successful prosecutions.

Notably, the Victoria Police’s annual report (2018) suggested that only 2700 sexual offenders had been identified for fingerprinting. It is unclear how cases were prioritised, given that most registered sexual offenders were not placed on the Registry due to their likelihood of reoffending. Research in other jurisdictions suggested that despite police feeling that the sexual offender registries were useful, they noted that the Registry’s ability to assist with sex crime investigations was hindered by difficulties with distinguishing between low- and high-risk offenders (Harris et al., 2018), which would presumably generalise to the Victorian Registry as well. Further, one of the few studies that found support for sexual offender registries in the United States suggested that the registry only reduced recidivism when the stringent conditions were reserved for high-risk individuals as assessed by validated risk tools (Duwe & Donnay, 2008).

Although there is no research regarding whether the Registry has improved police investigations, research suggests that it has had a significant impact on the courts due to the volume of secondary offences (i.e. offences resulting from breaching conditions) being prosecuted (SAC, 2017). Between 2011 and 2016, secondary offences across all offence types increased from 6% to 10% of the Magistrates’ Court’s caseload, with the number of secondary offences for sexual offenders rising 37% (SAC, 2017). However, of the 4990 secondary offences heard at the Magistrate’s Court for breaches to the Act, 97% were for failing to comply with reporting obligations. Typically, these breaches were relatively minor, resulting in fines (54%) or an ‘other’ outcome (i.e. suspended sentences, adjourned undertakings, diversion, etc.; 20%; SAC, 2017).

Although the public may fear that breaches of the Registry are motivated by deviant desires, as one American politician stated, ‘sex offenders often fail to register precisely so they can evade detection and in many cases, find new victims …’ (Blumenthal, 2011, p. 1), research does not support this conclusion (Duwe & Donnay, 2010; Levenson & Harris, 2012; Walfield, Levenson, Cubellis, Harris, & Lobanov-Rostovsky, 2017). Instead, often-cited reasons for registry non-compliance include complex reporting requirements that offenders are either unaware of or unable to comply with, fear of retaliation by public, and unstable accommodation (Walfield et al., 2017). Further, research has found that indicators of general antisociality (i.e. past arrest for non-sexual offending, greater diversity of offending, and history of breaches) were the best predictors of non-compliance, while sexual rearrest was not related to failure to comply for child sexual offenders (Duwe & Donnay, 2010; Levenson, Sandler, & Freeman, 2012). Considering the relatively minor sanctions for secondary offences and the overwhelming proportion of low- or moderate-risk offenders on the Registry (Victoria Police, 2012), it is unlikely that the increase in secondary offences has translated into greater public safety.

In fact, the increase in the number of secondary offences may have had significant negative consequences by contributing to the delayed prosecution of cases in the Magistrates’ Court. In Australia, the Magistrates’ Courts are expected to finalise all cases within 12 months and only allow approximately 10% of cases to go on for longer than six further months (Steering Committee for the Review of Government Services Provision, 2006). However, as of June 2017, 10% of cases in the Victorian Magistrates’ Court were more than 12 months old, and 26% of cases were more than six months old (Australian Government Productivity Commission, [AGPC], 2018). Indeed, the total number of backlogged cases doubled since 2011 (AGPC, 2018). The rise of secondary offences since 2011 may have contributed to the backlog in Victoria, which can have unintended consequences such as sending the wrong message to offenders, increasing the hardship for victims, and eroding the confidence of the public and morale of criminal justice practitioners (Australian Institute of Criminology, 2006).

Moreover, from an RNR (Andrews & Bonta, 2010) perspective, not only do secondary offences drain the Courts’ resources, but they have the potential to inflate low-risk offenders’ risk of recidivism by unduly exposing them repeatedly to the Court system and other higher risk offenders. Receiving additional convictions for secondary offences also has the potential to affect offenders’ self-image by being further labelled as a criminal or limit their protective factors by negatively impacting relationships or employment (Davies & Tanner, 2003; Mingus & Burchfield, 2012; Tweksbury & Lees, 2006). Although difficult to measure, there is theoretically a concern that the Registry may negatively impact offenders who would have otherwise been low risk.

Aside from public opinion (Cubellis et al., 2018; Harris et al., 2018; Koon-Magnin, 2015), there is limited evidence to suggest that sexual offender registries improve investigations and prosecutions. The recent amendments allowing fingerprinting may assist police to investigate cases or provide greater evidence for prosecution, but without evidence it cannot be known for certain. Instead, the Registry has arguably had a significant negative impact through the volume of minor secondary offences that may have contributed to delayed prosecutions in the Magistrates’ Court and unduly exposed some lower risk offenders to the justice system repeatedly.

Discussion

The Sex Offender Registration Act 2004 (Vic) was developed with the aim of reducing recidivism and improving the investigation of sexual offences. This article has sought to evaluate the Registry by considering whether psychological research supports the assumptions implicit in the legislation. The aims of the Registry appear to be somewhat misguided. Considering that the vast majority of sexual offences are perpetrated by first-time offenders (Rettenberger et al., 2015) and that 80% of registered offenders do not pose a high risk to the community (Victoria Police, 2012), the Registry has the potential to prevent a relatively small proportion of sexual crimes. Further, the mandatory registration of child sexual offenders, but required assessment for adult sexual offenders such as rapists, is not supported by empirical evidence or RNR theory. In addition to the concerns regarding the misattribution of government resources and infringement on the rights of low-risk offenders, the Director of Police Integrity highlighted that the Registry may have the opposite effect to what was intended as ‘truly dangerous offenders may be overlooked in the cast sea of registrants’ (Ombudsman Victoria, 2010, p. 24). Finally, although police have been granted additional powers over time to aid in their investigations, it is unclear whether these have been effective. Instead, the Registry has placed considerable pressure on the Magistrates’ Courts through the increase in secondary offences that are unlikely to be related to the sexual safety of the community.

Although there has been no empirical evaluation of the Registry, its overly inclusive nature suggests that it would be unlikely to be effective in reducing sexual violence in Victoria. In order to more effectively prevent sexual offending, the Registry should be reserved for individuals who have been assessed as being a high risk of sexual recidivism as determined by a sex offender risk assessment that has good to strong predictive validity for the specific offender type. This will allow police to focus on the offenders who need to be monitored, while not further exposing low-risk offenders to the justice system. If the Registry were to focus solely on high-risk offenders, this should in turn reduce the sheer volume of secondary offences and may increase the likelihood that charges for secondary offences may relate to more deviant behaviour.

Further, the conditions placed upon registered offenders should be determined by judicial decision making, rather than universally applied. For example, not all registered offenders pose a risk to children, yet all offenders are prohibited from child-related employment. Similarly, requirements to disclose new internet profiles or changes in internet providers may not be relevant to all offenders, particularly if they do not have a history of viewing child exploitation material or luring victims on the internet. While these conditions would undoubtedly be useful for certain cases, generalising them to all registered offenders increases the likelihood of secondary sentences for non-deviant behaviour. This may inadvertently increase the risk of registered offenders through exposure to the justice system, which can negatively impact upon the factors that protect them from reoffending (Davies & Tanner, 2003; Mingus & Burchfield, 2012; Tewksbury & Lees, 2006).

The VLRC (2012) argued that individuals who target adults should not be registered under the Act. Instead, the VLRC suggested that adult offenders would be better monitored by what is now the Serious Offenders Act 2018 (Vic) (Austl.), which subjects offenders who present as an unacceptable risk to the community to ongoing detention or supervision. Under this legislation, violent and sexual offenders are not differentiated, which may be more appropriate given the harm both groups pose to the community and the overlap between the two (Helmus et al., 2012; Looman & Abracen, 2010). Further, although there are core conditions of serious offender orders (e.g. attend appointments, report changes in employment, report personal details, not engage in dangerous behaviour; s.31), any further conditions are applied at the Court’s discretion.

The VLRC’s suggestion that the Serious Offenders Act 2018 (Vic) (Austl.) may be more useful for supervising adult sexual offenders has merit. However, it perpetuates the flawed approach to differentiating types of sexual offenders based upon the age of their victim rather than their relationship to the victim, motivation for offending, or risk of reoffending. Ongoing supervision orders, regardless of the legislation that mandates it, would have greater benefit if they were informed by evidence-based risk assessments and judges’ decision making rather than being based upon broad sweeping legislation that disregards empirical literature, violates the rights of low-risk offenders, and places undue burden on government resources.

Conclusion

Like many Western jurisdictions, Victoria has introduced sex offender registration legislation to protect the safety of the community. Due to the restrictions on research (s. 63.1.b), it is unknown whether the Registry has had an impact on community safety. However, considering the broad sweeping nature of the legislation, the overwhelming number of low- and moderate-risk offenders who are registered, and the lack of empirical evidence to support its underlying assumptions, there is considerable doubt whether the Registry is likely to have been useful for reducing recidivism or enhance community safety. At best, it may assist with police investigations, but without research on this topic, this cannot be said for certain. Instead, the Registry has burdened the court system, which costs taxpayers money and can delay the prosecutions of more significant matters (SAC, 2017). The Registry could be improved by revising who is registered and what their conditions are. Alternatively, other legislative options may be more useful for supervising high-risk offenders, particularly if evidence guides the assessment, supervision, and interventions of these offenders.

Ethical standards

This article was written in accordance with 2018 Australian Code for the Responsible Conduct of Research.

Declaration of conflicts of interest

Melanie L. Simmons has declared no conflicts of interest.

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.