Experiences of Supervised Bail

Introduction

1.1 This report outlines the findings of a series of research interviews with people who have been on supervised bail in Scotland. These interviews sought to gain an insight into their experiences of supervised bail, their views on supervised bail, and the ways in which supervised bail impacted upon their lives. This research forms one strand of a wider project examining supervised bail in Scotland.

1.2 This research aims to fill a gap in the evidence base by exploring how supervised bail works, thus informing future supervised bail policy and practice.

Supervised Bail in Scotland

1.3 Supervised bail is an alternative to remand (when people are imprisoned while they await trial), whereby those accused of a crime are released on bail on the condition that they meet with a social work or third sector^[1] supervisor up to three times a week. The aim of supervised bail is outlined in the Scottish Government guidance of 2008:

Bail supervision schemes are designed to minimise the numbers of accused held on remand in custody pending trial or for reports after conviction, who subject to safeguards in respect of public safety, could be released on bail to the community pending their further court hearing (Scottish Government 2008: 1)

1.4 In other words, the aim is to support bailees to comply with bail conditions, enabling them to remain in the community while awaiting trial. On an individual level the aim is to avoid the disruptive experience of a period of remand, and give bailees support to deal with any problems they may have, and at a system level it aims to reduce breach of bail and the prison population.

1.5 Supervised bail was initially piloted in Glasgow and Edinburgh from 1994. Following an evaluation of 9 months of the pilots (McCaig & Hardin 1999), funding was made available across Scotland, which areas could choose to use for bail supervision, or for bail verification schemes, which confirm addresses for potential bailees. As a result, bail supervision was not introduced in all Local Authority areas.

1.6 Research commissioned by the then Scottish Executive in 2004 on bail/custody decisions for females described a general consensus among judges interviewed that bail supervision "was an effective tool with some accused" (Brown et al 2004: 27). Many of the study's respondents also felt remand numbers could be reduced by increased funding of bail schemes (Ibid: 31), however the study also found that there was a lack of awareness amongst judges and prosecutors about services provided by social work and the third sector.

1.7 There is currently a supervised bail scheme in place in most Local Authority areas in Scotland. It is targeted at those who would otherwise be remanded, especially 'vulnerable' groups as defined by the 2008 Scottish Government guidance - young people, women, carers, drug misusers, those with mental health problems, and more generally those "who would suffer extreme difficulties if remanded to custody" (Scottish Government 2008: 2)^[2].

The theory

1.8 The logic model^[3] above outlines (from left to right) the theoretical links between the resources invested in supervised bail, the activities that take place, and the short, medium and long term outcomes for bailees.

1.9 Each column in this model is reliant on the one before it, such that activities can only take place if inputs are in place, and outcomes can only be realised if activities take place as envisaged, and if preceding outcomes have been realised. Moreover, outcomes will only be realised if the theory that the activities will or can lead to the short, then medium, then long term outcomes is correct.

1.10 Existing research literature gives us some evidence as to the existence and nature of the links between the activities and outcomes outlined in this theoretical logic model. There are two bodies of literature which are of use here - literature on desistance from offending and criminal justice interventions, and literature on the theory of planned behaviour.

1.11 Now, while supervised bailees can not all be described as 'offenders', as bailees do not always have previous convictions, and have not yet been tried for the case in hand, the aims and nature of bail supervision have much in common with other criminal justice interventions in the community, or with a supervision element, such as probation and mentoring.

1.12 And so, a recent review of the literature on what works to reduce reoffending drew two main conclusions that are relevant to supervised bail. First, it was found that community sentences are more effective than short prison sentences, in the long term, in reducing reoffending, and there is evidence that this is due to there being increased access to rehabilitative services in the community (Sapouna et al 2011). There are clear parallels between bail supervision and community sentences here, especially considering the emphasis placed in supervised bail guidance on signposting bailees to relevant services. There are also parallels between remand and short prison sentences, in that both involve a short spell of time in prison, and indeed it is even less likely that remand prisoners will have access to rehabilitative services due to the relatively short average length of remand.

1.13 Second, the review found that 'respectful, participatory and flexible contact with a supervisor can trigger positive change in offenders' (Ibid). Specifically, research has found that desistance from offending is a process, by which people 'discover' their own agency, and gradually build and live out a pro-social identity in place of a 'criminal' one (see McNeill 2009 and Farrall 2005). This can be influenced by life events such as marriage and employment (see 'external factors' in the logic model), but also by relationships with those around them, including family, friends, and 'lightly engaged professionals' such as supervisors (Farrall 2005).

1.14 Specific features of relationships with supervisors which have been found to lead to reduced reoffending and the development of pro-social identity include supervisor assistance to envisage a new identity and alternative new future (McNeill 2009, Farrall 2005), and help to find employment and mend family rifts (McNeill 2009). More specifically, supervision has been shown to be most effective when those being supervised experience supervision as active and participatory, perceive that the supervisors are personally and professionally committed to them, and demonstrate encouragement, fairness, and a genuine concern for their wellbeing. This in turn motivates in those being supervised a sense of loyalty to their supervisors, and a commitment to desist (Rex 1999).

1.15 This literature on desistance and supervision provides theory and evidence for the links between supervision activity, and our short, medium and long term outcomes by demonstrating that supervision activity can lead to the long term outcome of reduced reoffending when there is a good relationship between supervisor and supervised, and when this leads or contributes to a reassessment of behaviour, and a change in behaviour and identity.

1.16 The second body of literature which is relevant here is the literature on the theory of planned behaviour. This theory was developed in psychological literature, and has been successfully used to understand, explain and predict human behaviour across a multitude of spheres. For example it has been used to examine problem drinking, electoral voting, and condom use (see Azjen 1991, Kiriakidis 2010) to name just a few. To date, however, it has only ever been used to try to build a predictive model of reoffending (see Kiriakidis 2008 & 2010, and Forste et al 2011), rather than to explain and further understand relevant factors in behaviour change.

Figure 1 (Azjen 1991, page 182)

1.17 The theory of planned behaviour proposes that behaviour is primarily driven by intentions to behave in a certain way, which in turn are driven by three motivational factors (see Figure 1):

• Attitude towards the behaviour - the extent to which a person has positive or negative views of the behaviour
• Subjective norm - Perceived social pressure felt by a person in regards to the behaviour
• Perceived behavioural control - the extent to which a person feels they have agency in regards to the behaviour. This will be related to, but not necessarily match, actual control over behaviour.

1.18 The structure of this theory is useful in the current context, as it provides a framework for understanding the links between supervised bail activities, and their impact on bailee behaviour, as outlined in the short and medium term outcomes in the logic model.

1.19 The findings outlined in this report will focus on the experience of the supervised bail activities (being screened and selected for supervised bail, and the frequency and content of meetings), and the links between the activities and the outcomes, particularly the short and medium term outcomes. More specifically, the analysis of the interviews will explore the mechanisms and circumstances by which the activities involved in supervised bail can lead to the outcomes identified, as well as identifying situations where the activities do not lead to the outcomes.

The research

1.20 The interviews form one strand of a wider evaluation of the impact of supervised bail, which also included analysis of operational data, a workshop with bail workers, surveys of the judiciary and Procurators Fiscal and economic analysis. A report on the findings of the evaluation as a whole will be published alongside this report.

1.21 These other strands of the research found that supervised bail schemes are in place in most, but not all, Local Authority areas in Scotland, but uptake of the service has been declining in recent years. This decline could be addressed by ensuring good processes are in place for supervised bail screening and by tackling lack of awareness and buy in among the judiciary, Procurators Fiscal and defence agents.

1.22 Three quarters of bail supervision orders studied were completed successfully, and bail workers felt that breach rates were low. Members of the judiciary who believed that breach rates were low said that this influenced their use of supervised bail. Evidence also suggested that successful completion of supervised bail encourages the use of community sentences over prison sentences.

1.23 As outlined above, supervised bail is designed to only be used as an alternative to remand, not to regular bail. The research found that there is little 'net-widening' of supervised bail to cases where accused would not have otherwise been remanded.

1.24 Finally, the economic analysis found that the net benefits of supervised bail as an alternative to remand over the three years examined were between £2million and £13million, and that the average cost of a supervised bail case would need to rise by between 75% and 560% for supervised bail to no longer be cost effective.

1.25 The purpose of the interviews with those who had been on supervised bail was to add to this research by exploring the 'on the ground' experience of supervised bail, to find out bailees' experiences and attitudes in regards to supervised bail and its perceived impact on their behaviour and wider lives. Specifically, the aim was to explore the circumstances under which supervised bail does and does not lead to its intended outcomes, as outlined in the logic model above. It is hoped that this will inform future policy and practice, for example aiding in the development of best practice toolkits for relevant justice professionals.

1.26 This report will give a descriptive account of interviewed bailee's experiences of supervised bail, from activities to outcomes, before returning to the logic model and exploring how the findings, and desistance literature, contribute to understanding the links between inputs, outputs and outcomes for supervised bail.

Methods

1.27 Ten people who had recently been on supervised bail were interviewed between October 2010 and March 2011. In line with requirements of the Data Protection Act 1998^[4], interviewees were recruited with the assistance of three bail supervision schemes across Scotland, who either asked those who were about to finish supervised bail if they would be willing to be interviewed; contacted those whose supervised bail had recently ended to ask permission to send contact details to the researchers; or posted opt in letters to those bailees who were now in prison, on the researcher's behalf.

1.28 It had been intended to use a snowballing technique to boost recruitment, that is, to ask interviewees if they knew any other people who had been on supervised bail who might be willing to speak to us. However, the first interviewee indicated that he was currently trying to avoid contact with anyone who had recent contact with the justice system, and on the basis of this the researchers decided not to ask this of other interviewees.

1.29 It should therefore be noted that because the interviewees were recruited through their bail workers, generally these were people who had engaged with supervised bail, and had good relationships with their bail workers. It is likely that there are other bailees who did not engage, who were less likely to be contacted by bail workers in regards to this research, and also less likely to agree to participate.

1.30 There is therefore likely to be selection bias in the sample, meaning that those interviewed are not representative of supervised bailees more widely, and may tend to have more positive experiences than many of those who did not participate.

1.31 Nonetheless, the sample does include a range of age, gender, experience with the justice system, and location, meaning that the findings below cover a range of experiences and reactions to supervised bail, though the range would have undoubtedly have been wider if the sample had included those kinds of bailees who would be less likely to participate, such as those who did not complete, those who did not like their bail supervisors, and those without stable contact details.

1.32 The findings in this report should thus be regarded as by and large exploring the positive stories of supervised bail, and cannot examine the potentially less positive stories of bailees who did not participate.

1.33 To ensure their comfort, interviewees were (when possible) given a choice as to where the interview would take place. As a result, four interviews took place in the interviewees' homes, four in bail offices, and two in prisons. A note taker accompanied the interviewer to each interview, both for safety and to take comprehensive notes in case the recorder failed. It is possible that the presence of two researchers may have influenced the dynamic of the interview by creating a less 'natural' interview setting, which may not have been fully counteracted by the interviewer's assurances, informal and friendly approach, and requests for honesty. Clearly, there is no way to know whether and to what extent this influenced interviewee responses.

1.34 The interviews were semi-structured, and lasted between 15 and 50 minutes, with an average duration of 30 minutes. The topic guide can be found in Annex A.

1.35 Interviews were, with the interviewees' permission, recorded, and then transcribed, and analysed using the framework mode of analysis, with the assistance of Framework software.

1.36 Case studies of the particular circumstances and experiences of individual bailees can be found throughout the next chapter of the report.