Abstract
This article analyses the use of criminal justice measures aimed at the prevention of sexual offending across England and Wales. Specifically, it focuses on measures such as the ‘sex offenders register’ and sexual offences prevention orders (SOPOs) and the use of sanctions for their breach. Following a discussion of the apparent tensions between individual rights and public protection measures we present an original analysis of data collated from Multi-Agency Public Protection Arrangements (MAPPA) area reports over a nine-year period. Our analysis reveals considerable variation between areas in both the risk-level allocation of cases, the imposition of SOPOs and sanctions for non-compliance with MAPPA. We argue that these disparities raise issues concerning both the rights and autonomy of those subjected to public protection measures and highlight the need for further detailed research into MAPPA practices.
Introduction
Over the last two decades a raft of punitive legislation intended to manage the risks posed by ‘known sexual offenders’ has developed. This has included the introduction of supervisory oversight and monitoring of individuals in the community post-sentence through the ‘Sex Offenders Register’ 1 and Multi-Agency Public Protection Arrangements, or MAPPA 2 (Stone, 2012; Thomas, 2011). The general trend has been for interventions to be ‘strengthened’, ‘toughened’ and ‘tightened’ (Thomas, 2008), notably through the Sexual Offences Act 2003 and more recently by the Sexual Offences Act 2003 (Notification Requirements) (England and Wales) Regulations 2012. 3 Part 2 of the Sexual Offences Act 2003, for example, introduced a range of new civil preventative orders. These included: the sexual offences prevention order (SOPO), which replaced the sex offender order (introduced by s. 2 of the Crime and Disorder Act 1998); restraining orders (introduced by s. 66 of the Criminal Justice and Court Services Act 2000); foreign travel orders, which impose travel restrictions on registered sex offenders (RSOs); 4 and risk of sexual harm orders, which can be imposed on persons who have no convictions of any kind (Shute, 2004). Additionally, the Child Sex Offenders Disclosure Scheme allows for information about RSOs to be shared with the public through controlled disclosure (see Chan et al., 2010; Kemshall and Wood, 2010; Kemshall et al., 2012).
Recent changes made in the Anti-Social Behaviour, Crime and Policing Act 2014 widen the use of such orders so that they can be used to manage the risk against adults as well as children. For example, a new sexual harm prevention order will replace the sexual offences prevention order and foreign travel order; and the sexual risk order (SRO) has replaced the risk of sexual harm order. Due to the civil nature of these orders, the standard of proof required for their imposition is based on the balance of probabilities (their jurisdictional application and related debates surrounding the subject’s rights will be discussed later in the article). They then essentially entail the use of what Kemshall and Maguire (2002: 14) have termed ‘negative conditions’ to restrict and control the subject’s behaviour in the community. All carry criminal penalties should they be breached, including up to five years’ imprisonment (Knock et al., 2002; Shute, 2004). While designed to limit and monitor the behaviour of particular individuals in the interests of public protection, the gravity of these potential consequences for RSOs clearly underscores the importance of accurately evaluating their future risk of offending.
Currently, MAPPA oversees the assessment and management of ‘high risk’ individuals, notably RSOs (Stone, 2012; Thomas, 2011). MAPPA utilizes an inter-agency approach to the operational practices involved in the risk assessment, supervision, surveillance, intervention with, enforcement, compliance, and breach of, purportedly ‘high risk’ cases. A three-tiered risk management system (ranging from Levels 1 to 3) is used to target resources at those perceived as posing the highest level of risk (Level 3). The risk level of the individual therefore determines the degree of restrictive, rehabilitative and protective interventions put in place to manage risk, and the extent to which information is shared between the agencies involved in MAPPA. The Home Office and latterly the Ministry of Justice (MoJ) have issued various guidance manuals to MAPPA (e.g. MoJ, 2007, 2012) and subsequent evaluations of the public protection arrangements have been conducted which outline the risk management procedures used in relation to RSOs (e.g. Kemshall et al., 2005; Maguire et al., 2001; Wood and Kemshall, 2007; see also Kemshall, 2008; Wood and Kemshall, 2008). An area of concern that has been given increasing recognition is the need for consistency in practice across the different 42 MAPPA areas in England and Wales. In particular, the need to ensure that all RSOs are allocated to the appropriate level of risk management and that the imposition and use of SOPOs during this time was appropriate. Similar concerns over consistency were also raised by Hudson et al. (2015). These findings highlighted an uneven distribution of RSOs and variations in risk-level allocations both geographically (between the 42 MAPPA areas) and over time (from 2004/2005 to 2010/2011), again raising a number of concerns about the fairness of MAPPA practice.
The primary purpose of this article is to examine further the incidence of RSOs alongside the use and consistency of preventative measures against further offending. Specifically this article is interested in the use and consistency of the SOPO, 5 as well as the use of sanctions relating to the breach of SOPOs and the notification requirements. The primary data used in this analysis were derived from the annual reports for the 42 designated MAPPA areas in England and Wales over a nine-year period, from 2004/2005 to 2012/2013. We have therefore been able to examine consistency in the use of SOPOs over time and geographically.
The remainder of this article is structured in the following four ways. First, we examine in more detail a number of human rights concerns over the use of such preventative measures, specifically the notification requirements and the SOPO. Second, we describe how the data were collated and prepared for analysis. Third, we present the findings from fairly straightforward analyses of the data before, fourth, discussing important insights into future practice and research.
Just Social Policies versus Public Protection
Numerous legislative responses have been introduced to protect the public and curb rising community concern about the release of convicted sexual offenders. Yet the rate of sexual reconviction for sexual offenders is known to be comparatively low, typically 10 per cent to 15 per cent after five years (Bates et al., 2003; Hanson and Morton-Bourgon, 2005; Harris and Hanson, 2004; Hoare and Jansson, 2008; Kershall et al., 2008). Reconviction as an outcome measure, of course, fails to take into account unofficial levels of sexual re-offending (Falshaw et al., 2003), high attrition rates and/or difficulties in prosecution proceedings (Kelly et al., 2005; Taylor and Gassner, 2010). Accordingly researchers now recommend the use of other outcome measures to supplement reconviction data (Falshaw et al., 2003). Similarly, it is accepted that research examining repeat offending of convicted sexual offenders should also take into account the offence characteristics of different subtypes of sexual offenders (Butler et al., 2012; Craig et al., 2003; Finkelhor, 2009; Smallbone et al., 2008) as well as the impact of any intervention or management strategies (Brown, 2010; Butler et al., 2012; Hanson and Morton-Bourgon, 2009; Peck, 2011; Stalans, 2004) which have been shown to account for differences in recidivism rates.
Despite these methodological challenges, the observed recidivism rate is ‘less than commonly believed’ (Mann et al., 2010: 192). Thus MAPPA and the associated measures that supervise, control and monitor RSOs in the community have been extensively criticized for responding to an over-exaggerated level of risk reflecting the ‘Othering’ of those convicted of sexual offences. This has largely been attributed to the concept of ‘populist punitiveness’ (Bottoms, 1995: 40) in addition to the influence of the media who stand accused of inflating public awareness, concern and fear of sexual offending within the narrow stereotypes of ‘stranger’ and ‘predator danger’ (see inter alia Brayford and Deering, 2012; Cobley, 2000; Greer, 2003). Indeed, the increased policy focus on sexual offending during the 1990s led Soothill et al. (1998) to coin the phrase ‘criminal apartheid’ to account for the way in which sexual crimes were differentiated from other forms of offending. This trend has evidently continued with few other offence-groups 6 having been singled out in such ways (Blandford and Beech, 2011). The appropriation of criminal justice resources based on public demand and/or media recommendations (see Critcher, 2002) has been extensively criticized for failing to meet the real needs of current or future victims (see Greer, 2003; Kitzinger, 1999, 2004a, 2004b). Of interest to this article, is how policies based on so-called ‘precautionary logic’ (see Ericson, 2007; Hebenton and Seddon, 2009) neglect to take into account an RSO’s basic rights.
Many of the measures cited above have previously been regarded as acceptable due to their aim at ex ante prevention of crime rather than ex post punishment (see Schauer, 2013). However, this argument is being challenged due to the increasing severity of measures being regarded as effectively imposing or increasing further punishments which extend the original sentence. Thomas (2008) questions whether the ‘sex offenders register’, which is supposedly not a ‘punishment for a sexual offence [nor] part of the system of penalties’ (Home Office, 2012a: 9), now, due to numerous changes since its introduction, constitutes a punishment in its own right. Indeed, when introducing recent changes, the Home Office (2012b: 1, emphasis added) stated that current registrants were ‘required to comply with all requirements of the notification regime, including the new matters set out in the Sexual Offences Act 2003 (Notification Requirements) (England and Wales) Regulations 2012’. This included a reduction in the number of days an individual had to make initial and subsequent notifications, a requirement to provide identification at each notification (including finger prints and photographs), the requirement to notify any address at which they stayed for more than 12 hours where a child aged under 18 resided and the requirement to provide details of their bank accounts. Failure to comply with these new requirements constitutes a separate criminal offence punishable by up to five years’ imprisonment (Home Office, 2012b: 1).
This opens up current legislation to a number of human rights concerns and possible breaches of the European Convention of Human Rights (ECHR), which is enshrined into domestic law through the Human Rights Act 1998. The retrospective nature of the changes outlined above raises important questions regarding their ex post facto nature and whether they in fact breach article 7 (on ‘retroactivity’) and article 8 (on ‘the right to respect for privacy and family life’) of the ECHR. Following article 8.2, where measures aimed at improving ‘public safety’ or ‘the prevention of disorder or crime’ are not ‘necessary in a democratic society’, they may become susceptible to challenge. 7 Domestic courts appear to have reached the position that while inclusion on the ‘sex offender register’ engages article 8 it does not necessarily breach it (see Thomas, 2011: 76–79). However, more recent case law has challenged the requirement that those sentenced to 30 months or more for a qualifying offence are automatically placed on the register for life. In April 2010, the UK Supreme Court ruled 8 that individuals so affected must be given the opportunity to apply for their names to be removed from the register, arguing that lifetime inclusion without the right of review was a violation of article 8. Consequently, the Home Office has introduced a review mechanism, albeit only after 15 years’ inclusion on the ‘register’ and within the context of the new more stringent measures just discussed. While the lawfulness of the review mechanism may well be questioned further, the government’s position remains that the new amendments do not breach human rights law (Home Office, 2012b: 1).
Similar rights and proportionality arguments arise when discussing the use and impact of the civil order introduced in the Sexual Offences Act 2003. A SOPO is available to the courts when sentencing an individual for an offence listed in schedule 3 or schedule 5 9 of the Sexual Offences Act 2003, or on the application of the police in respect of a person who has previously been dealt with for such an offence. Under a SOPO an individual will (if not already) become subject to the notification requirements and be subject to restrictions on their behaviour – for instance, prohibition from having unsupervised contact with children or from refusing police access to Internet-enabled devices in their possession in order to examine the content. The minimum duration of a SOPO is five years, but orders can be made for an indefinite period of time and therefore can remain in place until discharged, varied or renewed by the courts. While a court cannot discharge an order within five years (without the consent of both its subject and the chief police officer from the relevant area) it has become the accepted position of the appeal courts 10 that the duration of a SOPO should match the duration of the notification requirements imposed. This applies particularly where an order is being made at the same time as the sentencing for an offence.
Section 104 of the 2003 Act, under which applications for a SOPO are granted, contains the statutory ‘test’ that the imposition of an order should occur only when ‘there is a reasonable cause to believe that an order is necessary to protect the public, or individual members of the public, from serious sexual harm’ 11 (Home Office, 2012a: 44, emphases added). Moreover, ‘a key factor to be considered is the risk presented by the defendant’ with the risk, in this context, based on ‘the likelihood of the offender committing a sexual offence; the imminence of that offending; and the seriousness of the harm resulting from it’ (Home Office, 2012a: 44). Thus a SOPO, and ultimately its intrusive content, is imposed based upon a prediction of future behaviour. This risk assessment is undertaken throughout a RSO’s sentence using risk assessment tools such as the Offender Assessment System (OASys), actuarial measures such as Risk Matrix 2000, Structured Assessment of Risk and Need (SARN) for those in treatment programmes and, in the case of juveniles, ASSET – Young Offender Assessment Profile. The most widely used actuarial risk-assessment tool for sexual offenders in the English and Welsh Prison and Probation Services is the Risk Matrix 2000 (Thornton et al., 2003).
The accuracy and predictive validity of actuarial instruments and the appropriateness of their use in predicting future offending has been widely debated and researched (see, for example, Barnett et al., 2010; Craig et al., 2006, 2008; Hanson and Morton-Bourgon, 2009). Current findings suggest that actuarial assessments outperform unstructured clinical judgement with regard to assessing risk of recidivism (Hanson and Morton-Bourgon, 2009). Accordingly, Barnett et al. (2010: 465), as with previous studies, found that the RM2000/S had ‘moderate predictive accuracy as a measure of relative risk, with a range of sexual offenders’. However, actuarial methods are not without their critics. Concerns over fairness immediately emerge from the possibility of false-positive outcomes when attempting to predict serious future offending among those subject to public protection measures. False-positives, or wrongly predicting that an individual poses a ‘high risk of re-offending’, arguably threaten the article 8 ‘right to private and family life’ when such predictions lead to the imposition of measures (e.g. SOPOs) aimed at reducing any perceived risk. These arguments can equally apply to the imposition of notification requirements. It is worth remembering that the penalty for breaching such measures is a custodial sentence of up to five years.
While actuarial measures are favoured, a number of other sexual offender risk measures continue to be used. Questions therefore remain as to how to interpret different results from various risk assessment tools (Hanson and Morton-Bourgon, 2009) and the role of professional judgement. Given the high stakes for RSOs, it is imperative that the assessments being used are as accurate as possible. It is therefore essential that research continues to examine each tool’s ability to measure relative risk, the purpose of risk assessments and the role of professional judgement, in order to ensure that restrictive measures put in place to manage those considered a higher risk of reoffending are applied fairly, consistently and only where strictly ‘necessary’.
Data and Methodology
The data used within this article were derived from MAPPA annual reports. These reports provide the public with information on how MAPPA operates and present statistical information on the individuals monitored across various risk-level categorizations. For the period 2004/2005 to 2010/2011 the annual reports were accessed via the MoJ website and from the websites of the relevant police force or probation service for each area. Since 2010/2011, this information has been collated centrally by the MoJ therefore centralized data were used to extend the time period to 2012/2013. The data used within this article thus cover a nine-year period from 2004/2005 to 2012/2013.
Statistical information relating to ‘Category One’ cases (RSOs) was collated for each of the 42 MAPPA areas. This included the number of RSOs under MAPPA and the number of RSOs managed at Levels 2 and 3 of the arrangements (from which the percentage managed at these ‘high’ and ‘critical risk’ levels was calculated). It also includes the number of RSOs convicted or cautioned for breach of the notification requirements, the number of full SOPOs imposed by the courts and the number of Level 2 and Level 3 RSOs sent to custody for breaching a SOPO (although as we explain below these are not all presented as ‘raw’ figures).
It is important to note that the data represent the number of cases included on 31 March for each corresponding year, like a form of census data. The total number of cases that were managed may have fluctuated between each census date. The same also applies to the risk-level allocation of RSOs, SOPOs and breach figures. Crucially however, the same census date was used consistently across all 42 MAPPA areas. Slight differences were observed for the period 2004/2005 to 2010/2011 when comparing data extracted from individual MAPPA area reports with official national statistics provided by the MoJ. These minor differences can possibly be explained by timing differences in the recording of these data, and due to the outcome of some convictions or cautions being overturned at a later date on appeal. For the majority of statistics, however, the number of RSOs over the nine-year period was broadly consistent with the figures provided within annual MoJ statistical bulletins.
Reported MAPPA data only provide total figures for each area across different measures such as the actual number of SOPOs imposed or RSOs convicted and cautioned for breaching notification requirements in each area. Given the large disparity in the number of RSOs managed in different areas these figures were not directly comparable between or across areas. This is because certain MAPPA areas cover much larger numbers of people in the population generally, for instance London and the West Midlands areas. In this article, we reproduce the total number of RSOs managed by MAPPA each year across England and Wales as a whole. However, we also reproduce nationwide data on risk-level allocation, as a percentage of RSOs managed at Levels 2 and 3 of MAPPA, per year. Furthermore, data concerning the use of preventative measures against sexual offending and sanctions for their breach are presented as relative annual rates, that is, per 100 RSOs, per year. The use of proxy measures based on relative rates allowed for correlations to be calculated between these measures using Pearson product-moment correlation coefficients (r).
Our research does not therefore report actual numbers of RSOs, preventative measures and sanctions for their breach in each area. The reasons for this are twofold. First, without standardizing the data in this way, as stated above, comparisons between the different MAPPA areas are meaningless. Second, by not providing the actual numbers of RSOs in each area (or numbers of SOPOs imposed) we avoid ‘sensationalizing’ the data since it is important to bear in mind that the actual number of RSOs in any given area is very small in relation to the size of the general population. While these standardized figures, or rates, remain mostly descriptive, they allow for the comparison of inter-area consistency (and fairness) in the use of preventative justice measures aimed at reducing sexual offending.
Analysis of MAPPA Data on ‘Registered Sexual Offenders’
The following section presents findings from fairly straightforward analyses of the data which explore both the use of public protection measures over time and inter-area (in-) consistency. Table 1 provides data on the total number of RSOs across England and Wales between 2004/2005 and 2012/2013, risk-level allocation of RSOs, 12 the rate at which SOPOs have been imposed and sanctions for breaching them (in the case of ‘higher risk’ RSOs) 13 as well as the rate of convictions and cautions for those that breach the notification requirements. These data show trends across all 42 MAPPA areas in England and Wales combined. In Table 2 we then compare the management of RSOs between the 42 MAPPA areas in England and Wales across the nine-year period since the implementation of the Sexual Offences Act 2003. Table 2 also provides data on risk-level allocation, the relative rates at which SOPOs were imposed and the relative rate of sanctions for breaching the notification requirements on an area-by-area basis. These are presented as the nine-year mean rates based on the end of year calculations of those rates from 2004/2005 to 2012/2013.
Numbers of RSOs and relative rates of the use of public protection measures in England and Wales over time (2004/2005 to 2012/2013).
Relative rates in the use of public protection measures across MAPPA areas in England and Wales (2004/2005 to 2012/2013).
Table 1 clearly shows that the overall number of RSOs in England and Wales has steadily risen over the nine-year period in which the data were recorded (2004/2005 to 2012/2013). This is arguably indicative of the cumulative effect of long notification periods which in some cases are indefinite. Predictably, there are fewer individuals leaving the ‘register’ than are joining it. However, the data also show a marked decrease in the proportion of RSOs being managed at the ‘high’ and ‘critical’ levels of risk (Levels 2 and 3). Accounting for this decrease in the overall number of ‘higher risk’ RSOs in England and Wales over a seven-year period, Hudson et al. (2015) acknowledge that this may have been the result of central guidance, including changes to the definition of the ‘critical few’ in 2003 (Home Office, 2003) and/or a break from the initial ‘precautionary principles’, ‘better safe than sorry’ and ‘defensive response’ to the perceived risks of sexual offenders (Kemshall and Wood, 2008: 611). A further positive interpretation is that high risk RSOs are being effectively managed through MAPPA, resulting in case management decisions to reclassify them as lower risk (Hudson et al., 2015). However, reiterating the questions posed in Hudson et al. (2015: 67); first, if any of these explanations were valid, would the decline in numbers be as dramatic as the analysis suggests? And second, why would the rate of decline vary so dramatically between MAPPA areas? Indeed, looking at these data over a nine-year period (see Table 2) the average rate ranges from 29.64 in London down to only 4.66 in (neighbouring) Essex – or a proportion of ‘higher risk’ cases six times greater in London.
What is important to this study is that despite this consistent (and expected) increase in the total number of RSOs, the proportion of cases managed at Levels 2 and 3 has shown a marked decrease at the same time as a considerable increase in the rate at which SOPOs have been imposed across England and Wales. Table 1 clearly shows that the rate at which SOPOs have been imposed by courts has increased dramatically following their introduction in the Sexual Offences Act 2003. Critically, this increase is not attributable to overall changes in the number of RSOs managed under MAPPA as our figures show the relative rate at which SOPOs were imposed (per 100 RSOs, per year).
This trend may be a consequence of greater familiarity with the processes of applying for and imposing SOPOs among police and Crown prosecutors. It could also reflect the different points of time at which an order can be obtained, pointing to an increase in the use of a SOPO under s. 104(1)(a) of the Sexual Offences Act 2003. Here, the ‘Relevant Chief Officer’ can apply for an order after the date of conviction if ‘satisfied that the defendant’s behaviour since the appropriate date makes it necessary’. However, the relationship between the annual figures for the proportion of RSOs managed at Levels 2 and 3 of MAPPA and the rate at which SOPOs were imposed by the courts shows a statistically significant negative relationship (r = -0.91, p < .01). While much needed clarification of the boundaries of ‘necessity’ when imposing SOPOs was provided in July 2012 by the Court of Appeal, 14 our data clearly underline the need for further research into their use. Indeed, given that a SOPO lasts for a minimum period of five years, a large number of these orders are likely to be in place with lower risk RSOs managed at MAPPA Level 1.
Table 1 also shows that the rate at which RSOs were convicted or cautioned for breaching the notification requirements increased steadily up until 2006/2007 before beginning to decline. Historically, a higher proportion of RSOs being subject to more intensive ‘risk management’ seems to bear some relation (if not a strictly ‘significant’ one, r = 0.66, p = .053) to the rate at which convictions and cautions for breaching the notification requirements were dispensed. It should be noted that the highest rate of convictions and cautions coincided with the period when the highest proportion of RSOs was managed at Levels 2 and 3, with a lower rate observed in recent years (when far fewer RSOs were managed at these levels). From 2008/2009 onward a significantly lower proportion of RSOs have been managed at the more intensive management Levels 2 and 3 which may account, to some extent, for the drop in the annual rate since breaches might be being detected less often due to the more infrequent contact with police public protection units (PPUs) among those RSOs previously managed at Levels 2 or 3.
Table 1 also shows the rate at which Level 2 and Level 3 RSOs were sent to custody for breaching a SOPO. The perceived increase in the rate at which custodial sentences have been passed for SOPO breaches from 2009/2010 onwards is the result of a much smaller number of RSOs being managed at MAPPA Levels 2 and 3. This has the effect of magnifying the rate in comparison to previous years. For this reason, it is difficult to ascertain whether responses to proven breaches have simply become more punitive or whether the reduction in the proportion of RSOs managed at Levels 2 and 3 has enabled PPUs to be more effective in their policing of compliance with SOPOs. Given these difficulties, we have not included data on sanctions for SOPO breaches in the comparisons of inter-area consistency to which we now turn.
When comparing the rate at which SOPOs were used between the 42 MAPPA areas in England and Wales across the same time period (in Table 2) there is also evidence of inter-area inconsistency. Teesside for example, imposed orders at an average rate of 10.74 (per 100 RSOs, per year) while the comparable rate for London was only 1.64. Therefore RSOs were, on average, more than six times more likely to have been made the subject of a SOPO (which could result in their imprisonment if breached) if they lived (or were sentenced) in Teesside than if they were in London. Rather incongruously given that they were statistically least likely to be subjected to a SOPO, RSOs in London were over six times more likely to receive a conviction or caution for breaching their notification requirement (8.65 sanctions per 100 RSOs, per year) than their counterparts in Avon and Somerset who were the least likely to be sanctioned for breaches (1.32 sanctions per 100 RSOs, per year).
There was no correlation between the average rates at which SOPOs were imposed across MAPPA areas with the average rates at which RSOs were convicted or cautioned for breaching the notification requirements (r = -0.04, p = .80). Nor was there any correlation between the average percentage of RSOs managed at the ‘higher risk’ Levels 2 and 3 of MAPPA and the rate of breaches (r = -0.03, p = .85). Therefore, when comparing practice across areas, higher rates of sanctions for non-compliance with notification requirements did not appear to be linked to higher use of SOPOs or higher levels of RSOs assessed as posing a greater risk to the public. Interestingly, the rates at which SOPOs were used showed a significant but moderate negative correlation (r = -0.43, p < .01) with higher reported average percentages of RSOs managed at Levels 2 and 3. The opposite relation might have been expected with more ‘high risk’ RSOs being associated with more orders to protect against ‘serious sexual harm’.
Concluding Remarks
This research has examined the incidence of RSOs and the use of civil prevention orders with the intention of examining the extent to which these measures are being used consistently. Collating and utilizing data published by the 42 MAPPA areas of England and Wales from 2004/2005 to 2010/2011, and MAPPA data collated centrally from 2010/2011 to 2012/2013, we were able to highlight variations over time (over a nine-year period) and geographically across the different MAPPA areas in a number of ways: first, in the risk allocation of RSOs; second, in the administration of SOPOs; and third, in the use of sanctions for non-compliance with MAPPA.
A number of tentative arguments have been put forward to account for this variation. For example, the variation in the way that risk is being assessed most likely reflects institutional differences (see also Hudson et al., 2015). The imposition of a SOPO, similarly, appears to be strongly influenced by localized preference across MAPPA areas, rather than due to the risk management level to which RSOs have been allocated there. Both explanations may reflect the cost of resourcing MAPPA practices over time. Resourcing issues could, for example, account for the apparent downwards trend in the use of higher risk categories in the risk-level allocation. Our data certainly raise questions as to whether the increase in the rate at which SOPOs have been used reflects the problem of shrinking resources and a need to find a more efficient and/or cheaper alternative to ‘managing’ RSOs previously monitored at MAPPA Levels 2 and 3.
Clearly a limitation of using these kinds of data and approach is that such explanations can only be considered to be proxies for the underlying variations. Only further detailed research, such as that undertaken in the USA (for example, Tewksbury and Mustaine, 2013), which includes local analysis, taking into account direct comparisons of offender types and practitioner attitudes and decisions across MAPPA areas would reveal this. However, the tentative findings presented here would suggest that further exploration and study of these issues is warranted. Crucially, we would argue that this is necessary from a human rights perspective. To reiterate, the level of risk at which a RSO is managed has implications in terms of their rights and autonomy, determining both the degree of restrictive, rehabilitative and protective interventions put in place to manage risk, and the extent to which information is shared between agencies involved in MAPPA. The apparent variations in the risk-level allocation of RSOs reinforce existing concerns surrounding the accuracy and predictability of different risk assessment tools and professional decision-making techniques. The variation in the rates of Level 2 and Level 3 RSOs and the use of SOPOs between MAPPA areas suggests an acceptability and/or tolerance of ‘false-positive’ outcomes, which disregard an RSO’s basic rights (Ericson, 2007; Hebenton and Seddon, 2009). While MAPPA guidance aims to negate any bias, our data clearly question whether the guidance is clear enough to ensure that all MAPPA areas follow it in the same way.
Human rights concerns become even more pertinent when you consider the punishment for non-compliance with crime prevention measures such as SOPOs and the notification requirements. Indeed, while the compliance rate for registration is high, this may actually reflect the fact that breach of notification requirements is an offence in its own right, attracting up to five years’ imprisonment. Similarly, breach of court orders can demand the same sentence. While our data show that the number of RSOs convicted or cautioned for breaching the notification requirements has declined, the apparent variations between the proportion of RSOs managed at MAPPA Levels 2 and 3 and the sanctions for non-compliance in each MAPPA area, raises questions about fairness and legitimacy.
Recent changes made in the Anti-social Behaviour, Crime and Policing Act 2014, as well as the recommendations made in the Davies Report (Davies, 2013) on civil prevention orders widens the net of who will be subject to the measures discussed in this article. It has been argued that the changes do not pose any fundamental human rights objections (including those discussed earlier in relation to article 8 of the ECHR, which concerns the ‘Right to Respect for Private and Family Life’) and that any additional restrictions or decisions to widen the net to include individuals with ‘no qualifying conviction’ (see, for example, Davies, 2013) are necessary and justifiable in order to protect children and adults from sexual abuse. While this article comes from the same purposive starting point, it raises questions over the precautionary logic that may be driving such decisions. More importantly, it recognizes that while there is clearly a need to protect the public from sexual abuse and to ensure that the legislation covers all eventualities, the measures administered need to be reasonable, proportionate, just or if nothing else, appropriately targeted rather than universally applied.
It is therefore imperative that we have accurate data on MAPPA practice. At the very least, our findings remind us that the practices of managing (and controlling) RSOs are a social process and are thus susceptible to variations in those practices which may not necessarily reflect their intended design and guidance. The level of variations in practice shown in this article supports the need for further research into the consistency of MAPPA practice to ensure that the need to manage RSOs is matched appropriately to an accurate prediction of future risk.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.