Moving Forward Making Changes: evaluation of a group-based treatment for sex offenders

1. Introduction and background

1.1 The management of sex offenders in Scotland

Overall, there have been substantial reductions in rates of recorded crime in Scotland, including violent crime, in recent years. However, recorded sexual crime has been on a long-term upward trend since the 1970s. The increase is particularly marked from 2010-11 onwards, with a 47% increase between 2011-12 and 2016-17. There are likely to be a variety of reasons for this increase in both recorded sexual crimes and in the number of Registered Sex Offenders ( RSOs) in Scotland (and England). Police Scotland have indicated that increased reporting, including reporting of historic sexual offences and online child sexual abuse, may be in part responsible for this trend (Scottish Government, 2017b). The implementation of the Sexual Offence (Scotland) Act 2009 is also a significant factor. This created new statutory offences for rape, sexual assault, coercion and new categories of 'other sexual crimes' such as coercing a person to look at a sexual image, which might previously have been prosecuted using common law offences and would not, therefore, have been capture in recorded sexual offence data.

RSOs in Scotland are managed via Multi-Agency Public Protection Arrangements ( MAPPA). These arrangements bring together the Police, Scottish Prison Service ( SPS), Health and Local Authorities to assess and manage the risk RSOs and other serious offenders pose. MAPPA is supported by the Violent and Sex Offender Register ( ViSOR), a national IT system to facilitate inter-agency communication and ensure critical information about offenders is shared. Police Scotland also have Offender Management Units in each of the 13 territorial Divisions, with specially trained staff dedicated to the management of RSOs, restricted patients and serious and violent offenders.

The Scottish Government's justice strategy (2012) has a strong focus on reducing reoffending and rehabilitating offenders, at the same time as protecting the public. This includes those convicted of sexual offences. Moving Forward: Making Changes ( MF:MC) is an intensive treatment programme for sex offenders which aims to "reduce the re-offending of men convicted of sexual offences and increase their opportunities and capacities for meeting needs by non-offending means" ( MF:MC Management Manual, 2014: 6).

1.2 Moving Forward: Making Changes

MF:MC was designed by the Scottish Prison Service and the Community Justice Operational Practice Unit of the Scottish Government. It was introduced in Scotland in 2014, following accreditation by the Scottish Advisory Panel on Offender Rehabilitation ( SAPOR). In line with evidence that intervention intensity ought to be linked to risk level (e.g. Lovins et al, 2009), it is aimed at adult (18 and over) male sexual offenders assessed as medium-high risk (via the Stable 2007 tool ^[1] ). Eligibility is based on risk-level rather than offence type – men who have offended against children and against adult women are both eligible, as are those who have committed internet offences (although a slightly different assessment approach is recommended for this group). Implementation and delivery of MF:MC is supported by a set of manuals, covering the programme theory, structure, content, management requirements, and framework for collecting data for evaluation.

1.2.1 Core principles of the programme

MF:MC is rooted in the 'Good Lives Model' ( GLM) for offender rehabilitation. The GLM assumes that everyone wants certain primary goods in their life but that, for offenders, this desire manifests itself in harmful ways due to a range of deficits in the offender and/or their environment. For example, a paedophile's preference to identify with children rather than adults is a distorted expression of the general desire for 'relatedness'. The programme puts a strong emphasis on helping people to understand themselves and their behaviours from a different perspective, to help them achieve common goals in a more 'pro-social' manner. As the theory manual puts it, the programme aims to "assist individuals to identify what their valued primary goals are, how they have tried to achieve them in the past, and how they could develop ways of achieving them in the future in a way that does not harm others" (2014: 16).

The GLM approach was introduced to address some of the perceived limitations of basing interventions solely on 'Risk-Needs-Responsivity' principles – namely that a risk-management approach did not address offender motivation to participate or provide an incentive to change (Fyfe, 2015). However, while the GLM approach provides the theoretical framework for MF:MC, the programme also draws on a wide range of treatment techniques and approaches to achieve its aim, including: Cognitive Behavioural Therapy ( CBT), Schema-focused work; mindfulness; motivational techniques; and behaviour modification.

1.2.2 How MFMC is delivered

MF:MC is designed to be delivered in a group format, although it can also be delivered 1:1 or 2:1. It is based on ' rolling groups' – men join the group at different times and can be at different stages of the programme at the same time. Men's individual programmes are structured around a number of modules and 'assignments', which they complete outside of the group and then present for discussion at an agreed group session. Different group members may be working on different assignments at the same time, but all contribute to group discussion and feedback on assignments – a feature intended to provide opportunities for 'vicarious learning'.

The programme is designed to be responsive to individual needs – while all men complete a number of 'essential' modules, 'optional' modules will be linked to individual treatment needs identified by the MF:MC team. The modules are linked directly to the Stable 2007 dynamic risk factors, and a small number of additional risk factors believed to be associated with sexual offending. There is no pre-determined timescale for completion – those with greater treatment needs (who are assigned more optional modules) will be likely to spend longer on the programme. The rolling group format distinguishes MF:MC from previous sex offender treatment programmes like the Community Sex Offender Groupwork Programme ( CSOGP), which followed a more fixed structure ( RMA, 2016a).

MF:MC is delivered in both custodial settings (currently in 4 prisons across Scotland) and in the community (currently 11 sites). In prison, the programme is delivered by psychologists and trainee forensic psychologists and by specially trained prison officers. In the community, it is delivered by Criminal Justice Social Work staff. MF:MC teams in each case consist of:

• one or more Treatment Managers, who have responsibility for ensuring the programme is delivered in accordance with its design, that participants are appropriately assessed and selected, and that staff are supported and developed
• a number of Practitioners, who are responsible for facilitating group or 2:1 sessions. Practitioners are assessed as 'Level 1' or 'Level 2' based on assessment of their competency levels at initial MF:MC training, or at subsequent 'transition' training after delivery of the programme for a period of time.

RSOs who are participating in MF:MC in the community will also have an assigned ' Case Manager', a Criminal Justice Social Worker who is responsible for managing their case. The Case Manager will be involved in delivering pre-programme content to men before they join a group, and in discussing their progress throughout their time on MF:MC.

Figure 1.1, below, shows the overall structure of the programme and the essential and optional modules included.

Figure 1.1 MF:MC programme structure (from the MF:MC Theory Manual)

1.3 Evaluation aims

The Scottish Government commissioned Ipsos MORI Scotland to conduct an evaluation of MF:MC to inform considerations on SAPOR's reaccreditation of the programme in 2018, to provide evidence for policy and practice to inform future improvement, and to identify (as far as is possible) potential emerging outcomes for participants. The evaluation is intended to address three main questions:

• 1. How does the programme work in practice, from referral to exit, and to what extent does the programme follow the manuals? Is programme integrity maintained?
• 2. Are there difficulties with delivering any aspects of the programme and are there improvements that could be made?
• 3. What are the main outcome measures and to what extent have outcomes been realised?

1.4 Evaluation design

The evaluation adopted a mixed method design, drawing on a variety of sources of evidence to answer the questions above, including:

• Quantitative analysis of data routinely entered into the MF:MC IT system
• Qualitative interviews with:
  • 19 participants across 5 case study sites (2 custody and 3 community)
  • 18 Treatment Managers across the 15 sites ^[2]
  • 4 Case Managers
  • 3 stakeholders from Police Scotland and 1 psychologist working with community-based MF:MC teams.
• A half-day workshop with 23 Practitioners from 12 MF:MC sites. ^[3]

1.5 Challenges and limitations

In this section, we highlight the key challenges and limitations that apply to this evaluation. When presenting our findings in the remainder of this report, we have tried to strike a balance between drawing out evidence that points to areas that appear to be effective or in need of improvement, while at the same time reminding the reader of the various limitations that apply to the conclusions that can be drawn.

1.5.1 Lack of a control group

The most reliable way of establishing any intervention's impact is to compare outcomes for participants with outcomes for similar individuals who did not go through the intervention, ideally with random allocation to a sufficiently large 'treatment' or 'control' group. However, this was not considered to be an option for evaluating MF:MC at this point (and would not have been feasible within the timescale for reporting to inform SAPOR's reaccreditation decision). In the absence of a control group, any observed changes in participant outcomes cannot be conclusively attributed to MF:MC – it is possible that such changes would have occurred in the absence of the intervention.

Another, related, challenge is separating the impact of MF:MC from that of other elements of the wider MAPPA system for supervising sex offenders in Scotland. In the absence of a control group who have not attended MF:MC, but are otherwise treated identically, isolating the precise impact of MF:MC is not possible. However, it should be noted again that an experimental approach would be challenging to implement in practice. It would require sufficently large sample sizes and men would have to be randomly assigned into the programme or into a control group which could raise both practical and ethical issues. Alternatively, a large matched sample of offenders (for example, who have similar offending histories, sentence lengths, needs and levels of risk) but who were not on the programme would have to be identified as a comparison group.

1.5.2 Timeframe for evaluating intervention impact

Delivery of MF:MC only started in 2014. As noted above, there is no fixed timeframe for participation, but men can be on the programme for relatively long periods. As such, at the time of writing there are still only relatively small numbers of men who have fully completed MF:MC. But perhaps more importantly in terms of assessing impact, all the impact data available for this evaluation relates to the period immediately post-programme. As such, this report is only able to consider short-term outcomes from the programme.

1.5.3 Issues relating to the content and completeness of monitoring data

Quantitative data was provided to the evaluation team from the MF:MC IT system from 2014 to September 2017 by each community site, and by SPS for all 4 prisons delivering MF:MC. However, Treatment Managers and Practitioners have highlighted a number of perceived limitations relating to the content, structure and completeness of this data. The specific issues raised are discussed further in relevant sections of chapters 2-4, but they have implications for:

• The accuracy of the data in terms of identifying how many men have completed MF:MC
• The validity and reliability of some of the data intended as outcome measures – particularly the eight psychometric scales included.

1.5.4 Limitations on interview coverage

The evaluation included interviews with Treatment Managers from every site and practitioners from 12 of 15 sites contributed to the workshop. However, it was not possible within the evaluation timescale or resources to interview men participating in MF:MC in every site. As such, there may be important variations in participants' experiences of MF:MC that are not captured here. In addition, the evaluation team were only able to speak to a very limited number of stakeholders within the available time and resources. Stakeholders from other areas, or from other backgrounds may have had different perspectives on the programme.

Men who had participated in MF:MC were recruited via MF:MC teams. While every effort was made to ensure that they did not 'cherry pick' those participants likely to be most positive about the programme, it is impossible to rule out any element of this having occurred. Moreover, participants who were less engaged with MF:MC are also less likely to be willing to engage with an evaluation interview. As such, interviews with participants are likely to be skewed towards those who were more engaged with MF:MC (although the sample did include a number of men who stated they had not been enthusiastic about taking part at the outset).

1.6 Report conventions and structure

1.6.1 Report conventions

MF:MC participants, staff and stakeholders were interviewed for this evaluation using a qualitative approach. Qualitative samples are generally small, and are designed to ensure a range of different views and experiences are captured. It is not appropriate given the number of interviews conducted to draw conclusions from qualitative data about the prevalence of particular views or experiences. As such, quantifying language, such as ' all', ' most' or ' a few' is avoided as far as possible when discussing qualitative findings.

In order to protect anonymity, participants in MF:MC are identified using anonymous reference numbers only, while quotes from staff are not attributed to specific sites (given the small numbers of staff employed in each site, a job title in combination with the site name could easily be identifying).

Finally, the evaluation was not designed to assess differences in success or impact between sites, nor to compare impact between SPS and community-based sites. Such comparisons are highly problematic, given that we are not able to control for external factors that might impact on 'success' in different sites (for example, the profile of participants will inevitably differ substantially between SPS and community-based sites). Evidence of perceived impact is therefore generally discussed in relation to MF:MC as a whole. However, the report does include some summary information about the overall numbers of participants across different sites and some discussion of variations in practice between sites.

1.6.2 Report structure

The remainder of this report is structured as follows:

• Chapter 2 describes the operation of the MF:MC programme in practice, assessing integrity and consistency of delivery in key areas
• Chapter 3 examines participation in MF:MC, including the operation of the assessment process, numbers participating, attrition, and the level and nature of engagement with the programme
• Chapter 4 considers programme outcomes. As described above, it is not possible to establish conclusive impacts without a control group. However, the report discusses the available evidence on shifts in risk scores and psychometric profiles over the course of the programme. It also explores perceived outcomes and which factors staff, participants and stakeholders believe contribute to or hinder the programme from having a positive impact.
• Chapter 5 discusses the conclusions from the evaluation.

Each chapter is prefaced with a summary of key findings and concludes with suggestions for improvements to the programme. The concluding chapter revisits the three key research questions, as well as considering specifically how to strengthen future evaluation of MF:MC.