Supporting Children’s Resettlement (‘Reentry’) After Custody: Beyond the Risk Paradigm

Increased Attention on Resettlement; Limited Results

The current article aims to resolve an apparent conundrum in youth justice: In spite of a considerable body of evidence, and associated practice guidance, in relation to the transition of children from custodial institutions back into the community, levels of reoffending remain extremely high. The article deals primarily with England and Wales, but the logic of argument and the model developed from it are likely to have wider international relevance, given that a range of jurisdictions continue to have stubbornly poor recidivism rates alongside a plethora of initiatives to improve support for children leaving custody (Development Services Group, 2017; O’Neill, 2018). This conundrum also has relevance for the research–practice nexus and issues of academic impact more generally: Why might a body of research on ‘what works’ that has informed practice guidance from policymakers not see that influence translated into impact on outcomes?

At a policy level, the focus on resettlement (also known internationally as ‘reentry’ or ‘aftercare’) into the community after a period of custody has become more pronounced in recent years, particularly as regards children. Indeed, in England and Wales, better facilitating resettlement was part of the rationale for introducing the detention and training order (DTO) as the standard custodial penalty for children; the youth offending team (YOT) officer responsible for supervising the community element of the order ‘would also be involved in the planning and supervision of the custodial element’ (Home Office, 1997, para 6.17). The DTO was intended to ensure a seamless intervention that offered a continuity of provision from the point of entry to the custodial institution, into the community (Hazel et al., 2002). A clear indicator of the importance attached by central government to making the transition back to the community smoother was the publication by the Youth Justice Board (2005) of a framework for resettlement in 2005. This document was updated 5 years later in the form of resettlement management guidance which attempted to clarify some of the implications for practice of the framework (Youth Justice Board, 2010a). In 2018, the Board confirmed that resettlement remained one of its priority areas (Youth Justice Board, 2018b).

This growing attention was in large part a consequence of a recognition that recidivism rates following a custodial sentence were historically poor and had shown little improvement over time, and that this was particularly true for those below the age of 18 years (Hagell and Hazel, 2001). The Crime and Disorder Act 1998 established that preventing offending by children and young people was the principal aim of the youth justice system. Since 2010, reducing the 12-month rate of reoffending has been one of the three high-level targets by which the performance of the youth justice system has been measured – the other two being reductions in the number of first-time entrants and in the child custodial population. While it should be acknowledged that reoffending is a blunt measure of effectiveness (Bateman, 2010), there is little doubt that, based on this indicator, resettlement provision has been found wanting. In spite of the increased profile of resettlement, any progress against the target has been muted. In 2010, the 1-year reoffending rate for children subject to custodial sentences of 6 months or less (which account for the majority of such disposals) was 74.1 per cent; by 2015, it had risen to 79 per cent. While there was a modest fall over the subsequent 2 years, to 71.5 per cent, there is, given developments hitherto, little reason to suppose that this represents anything other than a statistical fluctuation (Youth Justice Board/Ministry of Justice, 2019). In any event, it would seem unlikely that the recent decline can be attributed to changes in resettlement policy or practice, given that, as outlined above, resettlement has been a high priority for more than a decade.

The Paradox of Enduring Poor Outcomes

Such disappointing outcomes cannot be explained simply in terms of policy not drawing on findings from research; equally, it would appear that they cannot be dismissed as being a straightforward failure of practitioners to implement policy. Over the past 20 years, a considerable body of evidence on what constitutes effective resettlement has been established (for an overview, see Bateman et al., 2013). Moreover, a number of practice initiatives, which have both reflected and added to the existing knowledge base, have demonstrated that high-quality resettlement practice can yield promising results (see for instance, Ellis et al., 2012; Galahad, 2010; Hazel et al., 2010, 2012; Hazel and Liddle, 2012; Wright et al., 2012). To a large degree, guidance published by the Youth Justice Board has reflected the research base and attempted to draw on learning from promising practice. And yet the overall picture remains gloomy; successes have tended to be localised and short-lived, and restricted to pilot initiatives; attempts to mainstream effective resettlement have not yielded the anticipated benefits (Bateman, 2016).

Thus, while research has confirmed that meeting individual needs can make a marked difference to outcomes, an extensive overview of the evidence published in 2013 found that relatively few children leaving custody had access to appropriate education, training or employment. Similarly, although it was clear that a sizable proportion of the child custodial population would not be able to live with their families when they return to the community, alternative accommodation arrangements were rarely in place until shortly before release (Bateman et al., 2013).

Even in some of the pilots, where elements of promising practice were identified, support services were not closely aligned to assessed need. Despite clear evidence of the importance of education, training and employment to successful resettlement, the evaluation of the DTO found that only 40 per cent of young people leaving custody were engaged in any form of educational activities during the custodial element of their sentence (Hazel et al., 2002). More recently, the RESET (Resettlement, Education, Support, Employment and Training) initiative, which ran in three areas between 2005 and 2007, was able to demonstrate the benefits of a dedicated resource to co-ordinate packages of care but children’s identified needs were, nevertheless, adequately addressed in only a minority of cases (Hazel et al., 2010).

In the face of such findings, policy makers’ frustration has become apparent. A thematic inspection of resettlement in the youth justice system, undertaken by HM Inspectorate of Probation in 2015, highlighted what it considered to be a paradox. On the one hand, the high profile of work to support the transition from the secure estate to the community had generated a good deal of research as to what works, and the implications of this knowledge base had been widely disseminated to practitioners through guidance. Yet, with some notable exceptions, resettlement practice was found wanting and outcomes described as poor. Therefore, inspectors concluded that resettlement planning did not commence early enough, and that service provision was disjointed:

The combination of a lack of suitable, settled and supported accommodation; a deficiency in the services to meet mental health and substance misuse issues and an absence of meaningful ETE or other constructive activities did not give [children] an opportunity to make a success of that transition. On the contrary, it made it more likely that they would fail. (HM Inspectorate of Probation, 2015: 18) . . . Some children did not know where they would be living until a few days before their release, and because of this, the rest of the provision could not follow. This does not bode well for a successful reduction or cessation of offending. (HM Inspectorate of Probation, 2015: 4)

In this context, the fact that more than two thirds of children reoffended within 12 months of release and that just one quarter of those, whose cases were reviewed for the inspection, fully complied with the requirements of supervision when they returned to the community was unsurprising. But inspectors took the view that the resettlement evidence base was at a level where such failures were perplexing: ‘We know the solution. . . Why on earth is it not being done?’ (HM Inspectorate of Probation, 2015: 4). This is not only a question for practitioners, but for academics and policymakers looking to ensure that research messages impact on practice and outcomes.

The Nature of the Evidence Base: A Critical Appraisal

In spite of the criticisms in this regard that can be found in the thematic inspection report itself, the key to resolving the paradox identified by the inspectors does not lie simply in pointing to problems of implementation and a failure on the part of youth justice and custodial staff to take forward policy guidance in a consistent manner. While it is no doubt possible to find examples of such failures, these are themselves in need of explanation given the high priority that has been associated with resettlement activities in the recent period. In our view, the origins of the weaknesses in resettlement provision can be traced to limitations in the nature of the evidence base itself.

To a large extent, the research that informs resettlement policy in England and Wales has, in common with much criminological work of the last quarter of a century, been conducted within the traditions of the risk paradigm (Gray, 2011). Studies have drawn predominantly on a succession of evaluations of custodial or resettlement initiatives where effectiveness was narrowly construed in terms of what would reduce the risk of (short term) reoffending, a measure which had come to be accepted as the litmus test of youth justice success (Bateman, 2010; Canton, 2012). Accordingly, findings have tended to be concerned with identifying risk – and to a lesser degree, protective – factors that appear to be correlated with increased incarceration and, within the custodial population, higher levels of recidivism. Given the lens through which the studies have been conducted, investigations have deployed variables for which quantitative data are available, which in turn have tended to be drawn from assessment tools, which have themselves been developed on the basis of risk thinking. In England and Wales, for instance, the standardised assessment framework used across the youth justice system was until recently ASSET, a tool that purported to measure the risk of reoffending for the individual child based on practitioner assessments across a range of 12 domains of risk, including living arrangements, family and personal relationships, education, training and employment, lifestyle, substance use, attitudes to offending and the like (Kemshall, 2008).

It is possible to discern three principal foci around which evidence has been amassed:

From Research to Policy

Attempts to develop an evidence-based resettlement policy first emerged locally in North West England. The 2003 North West Resettlement Strategy and the 2004 Greater Manchester Resettlement Plan developed ‘pathways’ of support to counter particular prisoner characteristics identified in a report by the Social Exclusion Unit (2002). Key areas highlighted as being in need of assessment and intervention were accommodation, personal advice, education training and employment, and health (including drug treatment) (Youth Justice Trust, 2004).

Such local initiatives were in large part an attempt to fill the vacuum left by successive iterations of National Standards for Youth Justice published by the Youth Justice Board (2004, 2010b, 2013). To a large extent, such standards have been concerned with minimum expectations for procedural requirements, an indication perhaps of the extent to which effectiveness in criminal justice has progressively become equated with process, performance targets and time limits (McLaughlin et al., 2001). As late as 2013, the responsibilities of YOT practitioners in relation to resettlement were limited to putting ‘effective and robust plans in place to ensure resettlement into the community on release’ to be finalised as least 1 month prior to release, and making ‘arrangements for the young person to access, as necessary, accommodation, health, education and training/employment services on release’ (Youth Justice Board, 2013: 135).

Further elaboration is, however, provided in a variety of guidance documents, including the aforementioned framework for resettlement and the Board’s case management guidance. The section in the latter dealing with custody and resettlement acknowledges the complexity of the resettlement task and outlines seven ‘pathways’ that should be considered to ensure successful transition of children from custody to the community as follows:

With a single exception, each of these pathways can be readily matched to the risk factors for reoffending on release from custody identified in the literature as described above. The final pathway – case management – is intended to apply to each of the others, to ensure a partnership approach – in recognition of the fact that no single agency has the expertise, or resources, to deliver the range of required services. Again, this focus on multi-agency work can be traced to the evidence base. The development of a ‘brokerage’ function to encourage partnership working, and hold the various agencies to account, has, indeed, been found to make it more likely that services will be put in place to address the child’s assessed needs (Hazel et al., 2010).

The most recent edition of Youth Justice National Standards, published in 2019, represents something of a break with its predecessors, promising to ‘encourage and support innovation and good practice to improve outcomes for children who commit crime’ (Youth Justice Board, 2019: 3). From a resettlement perspective, the principal novelty resides in the recognition that children in the youth justice system may experience frequent transitions in a variety of contexts and that such transitions are potentially problematic. While resettlement pathways are not addressed explicitly, there are clear lines of continuity with previous publications. Practitioners are alerted to the importance of ‘assessment, planning and delivery of interventions’ to mitigate the potentially negative consequences of transitions. YOT management boards and secure establishments are reminded of the requirement for partnership work and of ‘holding services and agencies to account in the event of insufficient planning and delivery of the transition and or resettlement plan for a child’ (Youth Justice Board, 2019: 17).

The Ministry of Justice (2014) has similarly developed proposals for ‘getting resettlement right’ in a youth justice context (p. 11). While the elements are stripped back by comparison with the Youth Justice Board’s framework, it is nonetheless possible to discern the influence of the research base – albeit it mediated through consultation responses – in the headline prescriptions for improving the transition for children from custody to the community. There are three such headlines:

Towards an Explanation of the Paradox: Decontextualised Research Messages

Given the attention that has been devoted to the issue of transitions over the recent period, the mismatch between aspirations and the situation on the ground requires further investigation. If there has been a clear failure to overcome the ‘barriers to effective resettlement’ (Bateman et al., 2013: 11), the question, posed by HM Inspectorate of Probation, as to why evidence from research on resettlement has not been associated with improved outcomes merits additional attention.

The above brief analysis suggests that the disappointing outcomes associated with resettlement cannot be readily attributed to a failure of policy to draw on the evidence. On the contrary, it would appear that in this particular area of service delivery, care has been taken to align policy to research findings to an extent that is perhaps rare in criminal justice. For its part, HM Inspectorate of Probation noted that the requisite structures and processes to ensure successful transitions from custody to the community were in place, and that many of the staff working with children in custody and on release showed ‘tremendous commitment’ (HM Inspectorate of Probation, 2015: 22). Again this would appear to demonstrate some understanding of, and faithfulness to, a resettlement practice consistent with research evidence. Nevertheless, the thematic inspection found plenty to criticise.

Plans were too often an end in themselves, devoid of meaningful content; children, who frequently had little input in developing such plans and recognised their ‘tick box’ nature, became disengaged and disenchanted as a consequence. Work with children in custody was not future orientated, focused on preparation for the transition back into the community, but was restricted to what was routinely available within the institution in terms of educational provision, behaviour management and, in some cases, offending behaviour work. Conversely, resettlement input in the community frequently commenced too late and did not build on whatever progress may have been achieved in detention.

The inadequacy of prescribing processes as a mechanism to guarantee good outcomes is clearly demonstrated in the data provided by the inspectors. An assessment of the child’s needs was completed in 90 per cent of cases reviewed, and in most instances was conducted prior to sentence. More than three quarters of assessments were up to date. But undertaking an assessment was no indication of quality. Of particular significance is the fact that more than half paid insufficient attention to the child’s vulnerabilities, and this failure was especially apparent when those vulnerabilities were considered in the context of resettlement. In more than a quarter of cases, there was no analysis of the information collated, suggesting a lack of clarity on the part of youth justice staff as to the purpose of the assessment. The inspectors concluded that ‘planning was process driven, took little account of individual need and did not involve the child or their parents/carers’ (HM Inspectorate of Probation, 2015: 32).

The lack of coherence between what happened in custody and plans for the community was in part a failure of different agencies to co-operate and exchange information, a consistent theme in public sector reviews. But lack of partnership working was in itself an indication of a deeper underlying difficulty: namely that ‘case workers and case managers did not always understand their roles or the purpose of sentence planning’ (HM Inspectorate of Probation, 2015: 32). This resulted in the development of plans that had a limited focus on resettlement itself: they tended to prioritise arrangements for the child in custody rather than on what would happen to them in the community. Tellingly, for example, 36 per cent of initial planning meetings put in place plans for addressing the child’s physical health needs in custody but none looked at how these would be addressed in the community. A similar, albeit in some instances less pronounced, focus on institutional arrangements at the expense of those that would pertain following release was evident in each of the other areas of need considered in the report, with the exception of accommodation.

While these findings are illuminating and provide some indications of how practice might be improved, they do not, at least explicitly, address the fundamental question posed by the report: why practice has not fully absorbed the lessons from research. The thematic inspection does however provide some clues as to where an adequate answer to that question might lie. The predominance of process over purpose and a lack of understanding by practitioners of what they were aiming to achieve appear to us to be particularly suggestive.

In our view, a potential explanation of the paradox can be found in the partial, and problematic, nature of the evidence base itself. As indicated above, the studies that have contributed to the main body of research knowledge to date have been conducted largely within the framework of the risk paradigm. It is our contention that such risk-focused research, and the Youth Justice Board’s resettlement framework that developed out of it, has encouraged an inadequate conception of resettlement, which views good practice as consisting of discrete interventions to address a raft of disparate issues, that somehow combine to deliver effective transitions. In a sense, this is unsurprising given that the risk-based model of offending tends to present factors that predict recidivism as individual, unconnected and equally weighted considerations. ASSET accordingly invited practitioners to assess separately dimensions of risk that might be integrally connected. For instance, living in a particular neighbourhood might be associated with a significantly higher prospect of school exclusion but the assessment process required that they were considered as distinct issues. Moreover, workers were required to allocate a numeric score to each domain using the same range, thereby implying a non-hierarchical relationship between risk factors which might in reality reflect a rather more sophisticated interaction. For example, a lack of commitment to education might be a consequence of a serious substance misuse problem rather than a distinct and additional risk to be addressed through a separate intervention.

The decontextualised nature of the practice messages has hindered critical reflection and inter-agency cooperation (Hazel et al., 2015; Hazel and Hampson, 2015). Within the risk paradigm, there has been no attempt to identify a high-level aim for resettlement beyond preventing reoffending; and no theory of change has been elaborated that might help to provide an overarching understanding of the resettlement process within which individual strands of activity might cohere (Hazel et al., 2017).

Providing a Theory of Change: Successful Resettlement as Secondary Desistance

Such a high-level aim for resettlement, and explanatory theory of change, has been presented by the authors, among others, as part of the Beyond Youth Custody research programme (BYC). Through re-analysis of data from six previous resettlement projects, and additional surveys and case studies, an alternative vision was developed of how interventions to support transitions might be conceptualised; one that goes beyond the traditional focus on risk understood in terms of criminogenic needs and has greater potential to inform work to support successful transitions (Hazel et al., 2017). Reflecting ‘secondary desistance’ in established adult criminological literature (McNeill and Weaver, 2010; Maruna and Farrall, 2004), BYC found that where the resettlement process is successful, it involves children being able to shift their identity from one that is pro-offending to one that is pro-social. As such, resettlement should not be seen as either a single transition point from custody to community, nor even as following a sentence path; but as a desistance journey from an identity conducive to offending to one that promotes a crime-free life and social inclusion:

I’d always had intelligence and vocabulary to talk to people in a different way and portray myself in a different way, but before, I was ‘street’ and using slang. But it doesn’t get me far in life. . . If you conduct yourself in a good way. . . and portray myself in a good way, people will warm to me more. . . Everybody I know says I seem like a man now – I can’t go round talking like a child in a hoody. (23-year-old ex-offender, now construction worker; cited Hazel et al., 2017: 8)

As the journey in the quotation above suggests, this understanding of identity is social (Jenkins, 2008), seeing the more positive narrative as fostered and reinforced through involvement in constructive activities and interactions, and in the adoption of roles that promote it (Bateman and Hazel, 2013). The role of resettlement services, therefore, should be to facilitate the child making that journey, empowering them to develop an alternative personal narrative that offers them a positive, non-offending, future. It becomes apparent that such facilitation is not primarily about intervening to address identified risks in a linear fashion, but consists of providing support to the child to become agents of their own resettlement and their longer term rehabilitation. Indeed, herein lies a further difficulty with the existing research base. Explaining children’s criminality on the basis of risk, treats children as objects whose fate is largely determined by the risks which they embody, rather than regarding them as active individuals with a capacity to make choices, albeit that their options may be constrained by their socio-economic position (Case and Haines, 2009: 20). Interventions in the United States – where resettlement is known as reentry, have been criticised precisely because they ignore personal agency and individuals’ interpretation of the immediate context:

In reentry models, both control and support narratives assign a passive role to the ex-offender and almost no role to the community outside of correctional and human service agencies, missing the opportunities both to engage the offender in actively reconstructing his or her life and to engage the community in meaningfully accepting the returning ex-offender back. (Barton, 2006: 53).

If it is true that effective resettlement consists of supporting children to exercise agency in a pro-social direction, further clarity is then required as to the nature such support might take. The BYC model proposes that the evidence clearly points to two distinct, but reciprocal forms that are fundamental to enhancing the prospects that children will make the necessary shift: (1) personal support to guide their identity shift and (2) structural support to enable it (Hazel et al., 2017).

Reinterpreting the Risk Paradigm Evidence Base

Understanding effective resettlement as a desistance journey does not mean dismissing the risk paradigm research findings. Nor is it to argue that the statistical correlations between identified risk factors and the likelihood of further offending on release from custody are necessarily spurious. Indeed, it is obviously true that unless a child has stable accommodation, the prospects that they will desist from offending are significantly reduced.

However, a revised framework that sees effective practice as the provision of supporting the child on a journey of identity shift allows a positive reinterpretation of the existing research findings that can help explain them, endow them with meaning, identify interactions between them, contextualise them in a way that is meaningful for policymakers and practitioners, and illuminate what they neglect. For example, the traditional resettlement literature has identified that making provision for the child to take part in constructive activities such as education, training and organised leisure is associated with lower rates of offending, sometimes suggesting that it somehow distracts them from crime (Hazel and Liddle, 2012). Within the proposed model, such constructive activities can be reframed as offering opportunities for children to enjoy positive interactions with others, develop skills for the future, provide them with confidence and an opportunity to take up roles that can help shift to a pro-social identity. It is, in other words, not a question of eliminating a risk but of providing future-oriented structural support that can assist the child to achieve a pro-social identity.

Similar considerations might apply to other issues that have often been conceived through a risk lens. Providing stable accommodation, for instance, can be understood as addressing the real barriers confronting the young person on release if pro-social redemption narratives are to be sustained (Bateman and Hazel, 2014; Rajah et al., 2014).

Such a reconceptualization is not simply a terminological trick however; it has important consequences for how effective provision is delivered. If the provision of accommodation, education, training and employment and so on is understood as a mechanism intended to support the child to develop an alternative future, as opposed to fixing a deficit, it is imperative that such services are planned alongside the child, taking into account their strengths, desires and interests; ensuring what is provided fits with the child’s own view of what is in their interest and their plans for the future. Much of what is regarded as suitable and stable accommodation for present, risk influenced, purposes would accordingly fail that test in the revised model.

Nonetheless, an understanding of the business of resettlement as supporting identity shift can indeed confirm the critical importance of two messages or principles that stand out in the risk orientated evidence base, while helping to provide contextual accounts of why, and in what ways, they assist successful transitions (Hazel and Liddle, 2012). Effective resettlement must be characterised as having the following two qualities:

Beyond the Risk Paradigm

While it is accordingly important to acknowledge the validity of some of the lessons from the literature, albeit with an element of reframing, it is also clear that effective resettlement and practice will need to go beyond the risk paradigm. The tendency to deny agency, and a consequent disregard of issues of personal support, has already been noted as a major weakness in that theoretical approach. The importance of such considerations is made clear in the evidence provided by HM Inspectorate of Probation (2015) in the explanation provided by one child as to why they had successfully managed to stay out of trouble on release:

It hasn’t been difficult because whenever I felt that I wanted to offend, I came here and told [YOT Case Worker] and she just takes me to the park or a drink or home and tells me that if I am going to, you are going to lose everything and end up in and out. She tells me to do the right things. She tells me the good things like college, to be addicted to. She tells me that I’m too lovely, too funny to offend. She tells me that I don’t look like an offender, that I should follow my dreams, forget about the past, about the crime. (p. 29)

The neglect of personal support in the risk paradigm has also resulted in an undervaluing of relationship-based practice (both informal and formal) that plays a vital role in empowering children to weather the storms necessary to forge more positive personal narratives. While such empowerment would appear to be fundamental to all children, evidence suggests that it may be particularly significant for girls – whose small numbers in the resettlement cohort can render them invisible (Bateman and Hazel, 2014). As indicated in the previous section, emotional support from trusted adults is particularly required to help the child frame any lapses in behaviour as temporary setbacks rather than as confirmation of an intractable pro-offending identity (Wright et al. 2015: 5).

Reconceptualising resettlement in a manner that takes resettlement policy and practice beyond the risk paradigm accordingly suggests that it may be necessary to identify other indicators of effective provision. While the precise nature of resettlement support (both structural and personal) will be determined by the individual needs of the child, it is possible to specify three other qualities or principles, in addition to the two described in the previous section, that should characterise all resettlement provision. These relate to recent research within the emerging ‘Children First’ paradigm (Haines and Case, 2015) that speaks particularly to the nature of personal support needed to encourage and guide shifts in identity. Effective resettlement should be

Conclusion: Prospects for Change?

Gill and Wilson (2017), in seeking to explain the disappointing record of interventions for adults leaving custody, have asked whether such outcomes should be interpreted as a theoretical error in the model of service delivery, or a failure of implementation whereby programmes of intervention are not delivered in accord with the model. Our answer in respect of children is both: in particular implementation failure was, at least, in part a consequence of theoretical weakness inherent in the risk paradigm. The lack of an overall framework in which associations between risk factors and outcomes could be located ensured that research findings were decontextualised. As such, policy makers and practitioners not unreasonably viewed tackling risks as a set of discrete activities that became ends in themselves. Rather than holding the child at the centre, such interventions had the potential to bypass them altogether, particularly where non-engagement was seen as a failure on the part of the child rather than a failure of services to respond to the child’s concerns and interest and provide relevant personal support based on trusting relationships.

To the extent that risk-based practice leads to the provision of what would be perceived by the child as a supportive intervention, the focus inevitably tends towards the structural side of the equation – at its best providing opportunities for accommodation, education and training, health input and so on. There is nothing in the risk model itself, however, that requires that this provision should be determined by a wider understanding of the child’s resettlement journey and should as a matter of course reflect their goals, strengths and interests. Put simply, there is little point in offering a child a training course in which they have no interest, but this is not an uncommon scenario. Moreover, the risk model has traditionally had little to say about personal support. Although in some iterations, the importance of ‘responsivity’ – personalising interventions – is highlighted (see, for instance, Ugwudike and Morgan, 2019), this generally falls some way short of seeing the provision of a supportive relationship as one of the keystones that will enable the child to construct a positive future for themselves.

The arguments adduced here suggest that the resettlement knowledge base and the practice that it spawned has yielded less fruitful outcomes than might have been anticipated, in part at least as a function of the limitations of the risk model that was embedded at its heart. Going beyond that paradigm does not imply a wholesale rejection of that evidence or of developments in resettlement practice that have emerged over the past two decades. It does however suggest that improved outcomes require a new conceptualisation of successful resettlement as a journey involving the child’s development towards a pro-social identity. In doing so, the BYC model, including the overall role for services to facilitate the child’s pro-social identity development, provides a framework that is able to capture the helpful insights of the existing evidence base. It makes sense of statistical relationships established by risk paradigm research by interpreting them within the context of secondary desistance; offering an enhanced theory of change by which to co-ordinate policy and practice.

Indeed, we contend that such an interpretation explains, and so reaffirms, the two overarching messages from risk paradigm research findings: Support needs to be consistent across custody and community in order to help the child on a desistance journey rather than a single transition; and it needs to be coordinated in order for all services to address the many structural barriers along the child’s route towards their pro-social identity. However, it also highlights how the risk paradigm research tended to focus on structural support, devaluing child’s agency and largely ignoring the personal support that is also needed to guide identity shift. Consequently, we argue that it is necessary to move beyond the risk paradigm with three additional characteristics, largely drawn from more recent research, that speak to personal support: Support needs to be co-created with the child because identity development is their personal journey; it should be customised because each child’s identity and how it shifts will be unique; and it should be constructive, being strengths-based and future-focused in order to foster rather than undermine a pro-social identity.

There are promising signs that policy in England and Wales has begun to take notice of the criticisms, and limitations, of the risk paradigm and to move beyond it in accordance with the conceptualisation of resettlement proposed here. In late 2018, for instance, the Youth Justice Board published an overview of what it calls ‘constructive resettlement’ which is based explicitly on the BYC model and aims to ‘begin to help the sector make this approach work on the ground’ (Youth Justice Board, 2018a: 5). Similarly, in revising their expectations of the treatment of children in custody, HM Inspectorate of Prisons (2018) has recognised that ‘resettlement underpins the work of the whole establishment’ and that inspections should be designed to explore pro-social strengths and goals (p. 79).

We are conscious that the implementation of such a model is not without challenges. A new conceptualisation, even if accepted at the level of policy does not guarantee meaningful change in practice if the cultural environment remains the same. The recent introduction by the Youth Justice Board of a desistance theory informed assessment tool (AssetPlus), replacing the risk focused ASSET assessment framework, was intended to promote more strengths-based, future-oriented, assessments to underpin youth justice interventions that recognised children as agents of their own rehabilitation. Research has, however, suggested that youth justice practitioners, imbued in the risk paradigm over many years and subject to an inspection framework which has prioritised risk management, continue to make deficit-focused assessments of the risk of offending, while ignoring strengths, future goals and ambitions (Hampson, 2018). Therefore, successful implementation of the BYC model, or ‘constructive resettlement’, will require a radical culture shift that enables practice to move beyond the risk paradigm.

However, the theory of change for children offered by the model presented in this article has already shown some potential to facilitate such a shift in youth justice culture. Policy makers and at least some practitioners have started to explore how the approach proposed in this article might have the potential, with appropriate revisions to suit the context, to provide the terms of a framework for going beyond risk-based practice in order to move towards a practice that embodies the vision of a ‘Child First’ youth justice (Youth Justice Board, 2018a). In particular, the Youth Justice Board (2019) has recently endorsed ‘developing. . . pro-social identity for sustainable desistance’ as the theory of change in its new ‘child first, offender second’ guiding principle for all youth justice practice. Ongoing research will be required to ascertain the extent to which these initiatives at the level of policy filter down to ensure corresponding changes in youth justice intervention and child outcomes. From a wider perspective, given the powerful hold of the risk paradigm on practice with children in trouble across much of the Western industrialised world, it will be important to consider how lessons learned about moving beyond the present obsession with risk can be transferred across jurisdictions.