‘It’s a Hard Balance to Find’: The Perspectives of Youth Justice Practitioners in England on the Place of ‘Risk’ in an Emerging ‘Child-First’ World

In recent years, there has been a shift in youth justice central policy narratives in England and Wales away from risk assessment and management and towards child first. However, this shift is meeting with a number of challenges on the ground. The reasons for this have been conceptualised as resistance and reticence, contradiction and bifurcation and confusion about competing narratives emerging from different UK government departments about how to meet the statutory requirement to ‘prevent’ youth offending. The article emphasises the importance of meaningfully engaging with youth justice practitioners in debates about how to meet this challenge.


Introduction
The Youth Justice Board (the non-departmental public body with responsibility for overseeing the Youth Justice System in England and Wales) recently published an updated Strategic Plan for 2021-2024 (YJB, 2021: 10) which identified child first as their central guiding principle. Notably, this marks a shift from the approach outlined in their strategic plan of 2019-2022 of 'child first, offender second' (YJB, 2019: 7), presumably in response to criticisms that the 'offender second' element continued to have a labelling effect and deficit focus (Bateman, 2020). This represents a significant policy shift for the Youth Justice Board (YJB) away from the risk factor prevention paradigm (RFPP) (Bateman, 2020;Case and Haines, 2016;Wigzell, 2021). It has been a journey of several years, which gained significant momentum following UK Government commissioned Taylor Review of the Youth Justice System (Taylor, 2016), and subsequent appointment of Charlie Taylor as Chair of the YJB in 2017. However, despite this shift, there remains within national legislation in England and Wales the principle aim of youth offending teams to 'prevent' offending (s37 (1) The Crime and Disorder Act 1998). Risk assessment and management have become the means by which the statutory responsibility to prevent offending has been executed. As a result of this and other key pieces of legislation, a 'risk culture' (Case and Haines, 2016;Hampson, 2018) has dominated both youth justice and wider criminal justice practice for the past quarter of a century. However, as moves away from risk and towards child-first approaches emerge from central policy narratives, it is important that we gain an understanding of the perspectives of the youth justice practitioners responsible for navigating this shift, and the impact this is having on front-line practice.
This article is based on findings from an evaluation of the YJB's Constructive Resettlement Pathfinder Project. Data are drawn from 14 interviews with youth justice practitioners and operational managers, which were conducted as part of the evaluation. This article will argue that the policy shift away from risk narratives is meeting with a number of challenges on the ground. Previous research has argued that despite attempts by the YJB to move towards desistance-based and child-first approaches, the risk culture continues to dominate front-line practice (Hampson, 2018). This article builds on this narrative by offering further insight into the potential barriers to the cultural shift on the ground, and emphasises the importance of engaging with and hearing the perceptions of front-line staff about the current and future direction of youth justice policy and practice.

Setting the scene
It is important to first set out what is meant by child first, the risk factor prevention paradigm and desistance within a youth justice context. The risk factor prevention paradigm has been the dominant discourse across criminal justice since the 1990s and emerged in tandem with a general rise in actuarialism across public services (Smith, 2006). The actuarialist perspective seeks to identify causes within the context of risk factors, without seeking to explain why the link is there or a theory of change. It also focuses firmly on preventing and controlling the crime 'problem' through risk assessment and management. Various studies such as 'The Cambridge Study ' (West and Farrington, 1973) and 'The Rochester Youth Development Study' (Loeber et al., 1996) claimed to be able to identify a series of risk factors that increase the probability of a person or child committing a criminal offence. Therefore, by identifying and addressing these key risk factors, it was concluded that a person's risk of reoffending could reduce if they received the correct 'intervention' to address this (Farrington, 2002).
The risk paradigm has been heavily criticised in recent years for its flawed methodology (Case and Haines, 2009), its labelling effect (Bateman, 2020), and the adulterisation and responsibilisation of children : responsibilising measures, while simultaneously controlling the central-local relationship and the practice of YOTs and their staff via prescriptive and technicised responses to restricted conceptions of offending behaviour by children. (Haines and Case, 2015: 29) As part of their critique, Haines and Case argued that a youth justice system could prevent offending by taking a child first, offender second approach. The more recently renamed child-first approach would involve 'the total abandonment of risk-based assessment and intervention, but not the abandonment of assessment and intervention per se' (Case and Haines, 2016: 69). Rather, a child-first system would focus on 'Positive Promotion' (Case and Haines, 2009), rather than the negative, deficit focus of risk. Positive Promotion would be based on the concepts of 'universalism, diversion and normalisation, pursued through (non-criminal justice) practice that is inclusionary, participatory and legitimate' (Case and Haines, 2020: 9). Crucially, this approach would involve a holistic, individualistic, tailored-approach based on a child's welfare needs that focuses on their strengths and future aspirations. It is envisaged that this would be delivered by universal services, recognising that children in conflict with the law are still children first and do not need to be separated, labelled, responsibilised and criminalised.
The YJB has, in recent years, embraced this approach in their policy documentation, which will be outlined in detail below. However, at the same time, they have also adopted elements of desistance-based approaches to working with children in conflict with the law. Wigzell (2021:4) notes that the desistance paradigm differs from the risk paradigm in the following two key ways: first, it focuses on how people stop offending and second, desistance is a process, rather than an intervention delivered 'to' children with the aim of reducing the risk of offending. Maruna and Farrell (2004) break the desistance process into two phases -primary (stopping offending) and secondary desistance (maintenance of ceasing offending by achieving a shift in identity from pro-offending to non-offending or pro-social). Elements of desistance-based approaches have emerged within youth justice policy and practice in England and Wales in the last 10 years. First, the YJB responded to criticisms that the main youth justice risk assessment tool, Asset, was too focused on historical deficits by replacing it with AssetPlus in 2014. The new risk assessment sought to include consideration of a child's strengths, and identification of desistance factors. Second, the YJB have adopted a 'Constructive Resettlement' approach to working with children on release from custody. Constructive Resettlement is defined by the YJB (2018) as 'collaborative work with a child in custody and following release that builds upon their strengths, to help them shift their identity from pro-offending to prosocial' (p. 8). Within this approach, the clear overall role for all agencies is to facilitate the child's identity shift from 'offender' to 'pro-social' by focusing on the provision of individualised personal and structural support. Constructive Resettlement builds upon the work of the Beyond Youth Custody (BYC) programme in the United Kingdom (Hazel et al., 2017). Beyond Youth Custody was a 5-year programme ending in 2018 that sought to identify best practice in helping children resettle from custody into the community. BYC identified the following five key characteristics of effective and sustainable resettlement, namely, that all work with children should be constructive, co-created, customised, consistent and co-ordinated.
There are evident overlaps between child-first and desistance approaches, and Wigzell (2021) concludes that central policy narratives may be moving towards 'Child First Desistance' as elements of both approaches appear to be shaping the future direction of youth justice policy in England and Wales. It is, however, important to understand how this shift has occurred.

The policy shift from risk to child-first narratives
It is important that the YJB's move away from its focus on risk to child first is briefly outlined. This article does not seek to chart all historical developments within youth justice over the last 20-30 years. Indeed, there are many papers that do this extremely well, (see, for example, Case and Bateman, 2020;Case and Hampson, 2019;Cuneen et al., 2018;Goldson, 2020). The 'new' youth justice system was created under the 1998 Crime and Disorder Act, reflecting a wider 'punitive turn' (Muncie, 2008) towards children in trouble with the law in Western Europe and the United States. The main aim of the system in England and Wales was the prevention of offending by children (s37 (1) CDA 1998), with the Government viewing risk assessment-based intervention and management as the means by which to achieve this .
It has been argued that the UK Government's continued commitment to the risk factor prevention paradigm (Case and Hampson, 2019) in the face of sustained criticism from many quarters reflects, at worst, a contempt for children in trouble with the law; and, at best, a disregard for their needs. Some of the key criticisms of the risk factor prevention paradigm (RFPP) have been documented in detail elsewhere (see, for example, Bateman and Pitts, 2010;Case and Haines, 2009;Smith, 2006) and are outlined above. When Every Child Matters was published (Department for Education, 2004), Bateman and Pitts concluded that 'every child matters unless you have broken the law' (Bateman and Pitts, 2010: 56). Successive UK governments have tended to seek reforms to the YJS that, it has been argued, have amounted to 'stochastic features of statecraft' (Case and Hampson, 2019: 27) -a tinkering around the edges, rather than seeking wholesale reform.
An opportunity for major reform presented in 2016, with Charlie Taylor's (2016) government-commissioned review of the Youth Justice System. Taylor, inspired by the child first, offender second approach , made a number of significant recommendations that, if acted upon, could have brought about the reform sought by many. However, the UK Government ignored the majority of recommendations, with the main commitment being to replace a large part of the children's custodial estate (Young Offender Institutions and Secure Training Centres) with Secure Schools. Despite ignoring many of Taylor's recommendations, the UK Government appointed Taylor as Chair of the YJB in 2017. A number of child-first advocates joined the Board, and have commenced a work programme that has moved the YJB's central policy narrative away from the RFPP and towards child first. Revisions and updates to strategic documents, youth justice standards, pathfinder initiatives, terminology and training events all reflect this change in recent years (Bateman, 2020).

Confusion at the macro, meso and micro levels of youth justice
Ideological shifts in policy often seek to trigger equally sudden shifts in practice. Such shifts, however, can often lead to confusion among practitioners. Indeed, similar patterns of confusion emerged among front-line practitioners following the inception of the Crime and Disorder Act 1998, which, as part of the 'punitive turn', marked a shift in practice away from welfare and towards the management of risk (Burnett and Appleton, 2004). The confusion surrounding the current policy shift is evident at the macro, meso and micro levels. For example, at the macro level, there are competing narratives about how best to 'do' youth justice among academics and the wider research community. Advocates of child first  have clearly influenced the direction of YJB policy, while others draw attention to the importance of rights-based (e.g. Goldson and Randazzo, 2021),participatory (e.g. Smithson and Jones, 2021), or trauma informed (e.g. Evans et al., 2020) approaches to working with children in conflict with the law.
Furthermore, there is confusion and a lack of consensus between key UK Government agencies over the meaning of key criminal justice terms such as desistance (Maruna and Mann, 2019), prevention and diversion (Case and Hampson, 2019). This has resulted in different interpretations of key terms in key policy documents, assessment tools and inspection criteria, which for reasons of political expediency or otherwise, are leading to mixed and confusing messages to those on the youth justice 'coal face' (Case and Hampson, 2019: 33). In particular, the YJB and Her Majesty's Inspectorate of Probation's (the body responsible for inspecting local youth justice teams in England and Wales) understanding of desistance appear to be 'at odds' (Bateman, 2020) with one another, leading to further confusion and mixed messages for youth justice managers and practitioners. Moreover, although the YJB have stated that their approach to youth justice will be on the basis of child first, there are strong elements of desistance-based approaches (see, for example, 'Making Resettlement Constructive', YJB, 2018), while features of the RFPP remain in the YJB (2019) Case Management Guidance.
The lack of consensus and confusion is also evident at the meso level, with many different iterations and models of youth justice teams across England and Wales. Smith and Gray (2019) contend that the 'monolithic view' of youth justice presented at the centre does not reflect the application of youth justice practice on the ground. The localism agenda (the decentralisation of power away from London and into the hands of local councils) has contributed to this, reducing the influence of the centrally based YJB on local youth justice teams (Bateman, 2020), leading to multiple iterations of how best to 'do' youth justice in different local authorities. As new principles and approaches emerge, they often complete with existing principles and practices on the ground. Goodman et al. (2017) argue that penal policy and practice rarely changes from one thing to the other. Rather, established approaches blend with emerging ideas. Smith and Gray (2019), in their analysis of 34 youth justice plans, found a number of different models of practice displaying a range of different 'blends' of ideas ranging from traditional and distinct youth offending teams, that prioritised risk assessment and management of children's criminal behaviours, through to child-first models where the youth offending team has been absorbed into a wider targeted youth support provision that prioritises welfare needs over criminal behaviours. The differing models of youth justice practice across England and Wales align themselves with differing theoretical approaches including RFPP, trauma informed practice, desistance or child first models. Some appear to attempt a hybrid of some, or all of, the above (Smith and Gray, 2019).
Further compounding this, confused and contradictory messages are contained in HMIP (Her Majesty's Inspectorate of Probation) youth justice inspection reports. For example, Hampson (2018) notes that the Youth Justice Inspectorate commenced a desistance-themed inspection of youth offending teams in 2016. However, subsequent inspections 'virtually ignored' desistance, seeking alternatively to focus on risk assessment and management, and offending behaviour interventions focused on reducing a child's risk of offending and harm (Hampson, 2018: 30). As Hampson (2018) notes, This surely gives mixed messages to YOTs regarding what is expected of them -how can they pursue a desistance-based agenda if the criteria upon which they will be judged by the inspectorate is still (for general inspections) firmly risk-focused? (p. 30) The confusion and mixed messages are having an impact on the micro level among youth justice practitioners. Hampson (2018) found that, despite the YJB developing a new assessment tool, AssetPlus, with an increased focus on desistance, the risk culture continued to dominate youth justice practice. Equally, Wigzell (2021) found that the concept of 'identity shift', a core part of desistance-based approaches in youth justice, was 'irrelevant' to children subject to youth justice supervision. Wigzell (2021) also highlighted a growing number of first-time entrants into the youth justice system, who may not have an established 'pro criminal identity', meaning that the desistance-based initiatives promoted by the YJB as part of their shift to child first, may have limited impact.
Variations in front-line practice are also influenced by individuals' diverse backgrounds, cultures, training and perspectives. For example, Weston and Mythen (2020) found that practitioners' understandings of child sexual exploitation were 'markedly influenced by their own personal experiences, moral codes and social values'. Similarly, Bovarnick (2010) found that the approach adopted by professionals impacted on their perceptions of the children. For those who adopted a child-centred approach, children were viewed as in need of protection.
The youth justice literature has suggested that there appears to be a 'resistance and reticence' among practitioners to move away from risk-based approaches (Case and Haines, 2020;Hampson, 2018). Lack of adequate training, mixed messages at the macro and meso levels, confusion about key terminology, investment in established forms of practice and the reducing influence of the YJB have all been cited as possible reasons for the limited impact of changes in practice (Bateman, 2020;Case and Haines, 2020;Hampson, 2018).
It is clear from the literature that the impact of the shift in central policy narratives away from risk-based approaches and towards child first on front-line practice is difficult to discern (Bateman, 2020:6). This article seeks to explore this by focusing on the perceptions of youth justice practitioners on whether, and how, these shifts are having an impact 'on the ground'.

Methods
The study from which this particular article is drawn adopted a qualitative approach using 1:1 semi-structured interviews. They were conducted with 14 youth justice practitioners and operational managers from South and West Yorkshire Resettlement Consortium as part of a wider, YJB-funded, evaluation of the Constructive Resettlement pathfinder.
Prior to conducting the research, ethical clearance was secured from Keele University, and due to COVID restrictions, all interviews were conducted online through Microsoft Teams, after participants were initially asked to participate through email. The information sheet and consent form were attached to the initial email. Once participants responded positively by email, a date and time was arranged to conduct the interview.
Immediately prior to commencing the interview, the researcher checked that the participant still consented, and made it clear that the process was entirely voluntary. An interview schedule containing 'prompts' and 'probes' was used to facilitate a conversation, the pace and direction of which was largely determined by the interviewee. The aim in the interviews was for the exchange to feel like a 'conversation with a purpose' (Burgess, 1984) and less like a structured question and answer session. All interviews were recorded on Teams and deleted immediately after transcription. All interviews were anonymised at the point of transcription. Youth justice workers and managers roles are highlighted in brackets below each quote.
Adopting inductive reasoning, data analysis of the interviews highlighted several recurring themes. The approach recognises that data are produced as a result of the interaction between interviewer and interviewee (Charmaz and Bryant, 2007). It is therefore recognised that the nature of the interview, the relationship between the interviewer and interviewee, and the setting can all impact on the data. A number of themes were identified, which are outlined below.

Findings and discussion -On the ground -Resistance, contradiction and confusion
A number of themes emerged during the interviews with youth justice practitioners and operational managers including evidence of resistance and a culture of fear; contradiction and emerging bifurcated practice; and confusion about the meaning of key terms and how to negotiate a 'balance' between competing approaches.

Resistance and a culture of fear
Evidence emerged that staff feel that risk should form a central part of their work with youth justice. This is synonymous with the literature that referred to the continued dominance of the risk culture (Hampson, 2018), and a perceived reticence on the part of practitioners to move away from risk-based approaches . Workers and operational managers discussed striking a balance between risk assessment and management, and welfare-based approaches such as child first: The risk stuff, I'm not with the school of thought that says risk has no place in a child-centred system, because I think it does. Even if you look at places like Norway where it's a welfarebased rather than criminal justice system, they are still assessing risk. (YOT Team Manager 1) Others discussed the importance of risk assessments remaining a core feature of youth justice work from the perspective of defensive practice and decision making: If you have a risky case that comes out of custody and does a serious offence, you can have people who are going to be looking at your work and your assessments from a risk management perspective. And basically, from a hindsight knowledge perspective, where they're going to be thinking, well we know this happened. And this was terrible. And what on earth was done about it? So that never goes away, and that's always at the back of your mind. And I guess management would say, well, we've got all these risk procedures in place that will take care of risks. (YOT Worker 1) There was a perception among staff that, should a serious incident occur, the risk assessment and management systems provided a degree of protection and clarity about why key decisions were made, offering a standardised scoring system to potentially explain and justify their actions. Although, this, to a certain degree may be true, risk-based approaches can also be used as evidence of 'poor' decision-making and be used to responsibilise individual practitioners when a serious incident occurs. It is apparent that the 'risk culture' within youth justice practice goes beyond interactions with children, and has created a 'risk averse culture' and workforce that is fearful of 'getting it wrong' and being subject to scrutiny and sanctions: There are so many specialist risk assessments, that the fear is that you haven't got time to do them all, so you leave yourself exposed. (YOT Officer 3) Despite the apparent fear of 'getting it wrong', some practitioners and managers felt that, although child first should form a core part of their work, risk should also be a central feature.
There was also concern about how children deemed to be 'very high risk of harm' or 'dangerous' would be safely managed in the community if risk management controls were removed: And I think the bit for me that's really important about it is the difference between internal and external controls in risk management, because the reality is that some of our young people, whilst it may be symptomatic of lived experience, the reality is in the here and now they are dangerous, they do present a significant risk of harm to other people. The important bit for me is actually those external controls in terms of what you put around them to contain and support them whilst they develop those internal controls that keep them safe as much as other people safe. (YOT Team Manager 3) One worker described the value of an assessment process in being that it allows a worker to use their professional 'gut' instinct, but the assessment allows one to make sense of that instinct and ensure that nothing is 'missed': I'm not disputing the professional . . . A lot of professionals have that ability, but that process of working it out, I think, and transferring it onto paper and working it through is still a useful one, because I know that when I do that with . . . When I pick up cases, my thought process can change, my hypothesis at the start can change. I will miss things, I will start seeing connections that I didn't perhaps see initially, and that's where I think the value of the assessment process is. (YOT Team Manager 1) The value of comprehensive assessments of a child is not in question here. Rather, it is the 'risk lens' through which the assessment is conducted that has been widely critiqued. A fear of 'getting it wrong' emerged strongly from the interviews -whether it is the fear of getting an assessment wrong and 'missing something' and leaving oneself open to professional scrutiny; or a fear how the public can be protected from children labelled as 'dangerous' or a 'high risk of harm to others' if the RFPP is removed. It is apparent that fear is a significant feature of the 'risk culture' in youth justice, and, for some workers, the very suggestion that this should be replaced with child first, exposes these fears and leads to reticence and resistance.
A number of workers felt that there was value in retaining a focus on offending behaviour work, a core part of the RFPP, even though it was referred to as the 'negative stuff' that they 'got out of the way': I definitely think offending behaviour needs to be addressed to start with. And I think it's quite good to get it done early on, so then all the negative stuff's done and out of the way. They can focus on that and then move on to the more positive things. It's definitely something that needs doing and I think it makes them (the child) think a little bit. (YOT Education Worker) Here, there appears to be a recognised value in completing offending behaviour work with a child, but for this practitioner, it is one of the less favourable aspects of their time together, and it is difficult to ascertain what the identified value is.
Some workers recognised that the introduction of AssetPlus in 2014 had led to an increased focus on desistance and, in particular, a child's strengths in sentence plans and work completed with children. However, supporting Hampson's (2018) findings, it was acknowledged that it takes more than new assessment frameworks to 'change people's minds': I think it's probably achieved a bit of a shift, but maybe not as much as it was intended to. I think the reason for that is because you can't change people's minds with a piece of paper. (YOT Worker 5) It is apparent that there is a perception among front-line practitioners that tinkering with the youth justice system through the creation of new assessment systems, rather than seeking wholesale reform, has had limited impact. This can also be explained, in part, by the risk averse culture that was evident with some youth justice practitioners.
A number of staff discussed their efforts to bring about the cultural shift away from risk-based approaches to child first and some of the resistance that they had met along the way: And I said, exactly this thing, we should be future-focused and that's how we can manage the risk. And the person responding it, it was like their mic drop moment, if you don't want to manage risk, you should not be in youth justice. And it totally floored me, because it's like we'd had this half hour discussion and basically they'd just chucked out everything by just saying that it's all about risk. (YOT Senior Practitioner) Evidence of a resistance and reticence to completely abandoning the risk factor prevention paradigm emerged strongly from the data. There appear to be several reasons for this including a fear of 'getting it wrong' and a perception that the RFPP provides a degree of protection; how to manage children who are deemed to be 'dangerous' or a 'high risk of harm'; or a perception that an inherent part of working in the youth justice system involves the assessment and management of risk. It is possible that emerging practice on the ground supports Goodman et al.'s (2017) thesis that the development of criminal justice practice should be understood as a complex blend of many competing approaches, rather than a 'pendulum swing' between risk and child first.

Contradiction and Bifurcated Practice
Apparent contradictory messages from HMIP (Her Majesty's Inspectorate of Probation) and the YJB about whether the focus of youth justice practice should be on risk-based or child-first approaches was causing a number of difficulties and challenges on the ground. A recent inspection of a youth justice team that had adopted a child first model had been subject to criticism of its risk assessment and management processes, leading to a negative outcome. This had a huge impact on practice in that youth justice team, and had also been felt in neighbouring local authority areas: I know with *****, and it was a shock to everybody what their inspection result was. And I'm sure that anybody who knows ***** YOT would know that they're not that YOT that got rated like they did. (YOT Team Manager 2) But just off the back of the inspection, we've gone back down, we've totally changed our management risk processes. They're much more labour intensive for case workers now and that's only off the back of the inspection. So, for me, I see we've gone down the wrong route because we've taken workers away from being with young people. Because the process now it's much more time-consuming. That time has to come from somewhere. And the time comes from the time they could spend with the young people. (YOT Senior Practitioner) The lengthy bureaucratic procedures and paperwork involved in the increased focus on risk assessment and management of children was also a concern for workers, as they felt it reduced the time that they could be spending with the children. The increased focus on risk also changed the nature of the relationship with children, potentially undermining elements of a child-first relationship such as co-production, future focus and individuality. It was highlighted that the increased levels of bureaucracy and paperwork following the inspection of a neighbouring area had gone 'too far' for staff: I think that there is a preoccupation with risk. I think there has to be a preoccupation with risk, but I get the sense that there's a feeling amongst my colleagues who are doing all those risk assessments, that it's gone a little bit too far in terms of the paperwork right now. But I don't want to speak for them. That's just the sense that I get. (YOT Specialist Worker) It is apparent that the contradictory messages from HMIP and the YJB appeared to be increasing workloads for staff, as they were trying to meet, what they felt, were the competing demands of two differing approaches to working with children. It was widely acknowledged that this meant increased levels of bureaucracy, less time spent with the children, a return to deficit-focused offending behaviour work, and a form of bifurcated practice. Some examples of this included that a number of YOTs sought to amend their practice to focus on risk, and meet the demands of HMIP, while also seeking to work with a child in a way that is child first. Some youth offending teams had created atheir own 'child friendly' plan: The bifurcated practice was evident in that youth justice teams had developed their own child first plan that they used as the basis for all their work with children, but they continued to complete the AssetPlus plan, using risk-based terminology to satisfy the requirements of HMIP: Yes, because the work involved in assessments is significant. Maybe we need to be better at explaining something. So, if we're putting in a plan we're going to help you join a club, but we put in our bit of the plan and that's offending behaviour work in a way, because it means that. It's just difficult, isn't it? (YOT Officer 3) But what it means to me and in terms of thinking about practice, it's thinking about how do we move away from backward facing and negative formulated plans and work that we're going to do with young people to be more future orientated and positive. And know that that's recognised, that the research that you've been part of tells us that that's the way to do it and people like HMIP need to catch up with that. (YOT Officer 5) It was apparent in interviews that the implementation of a child-first or desistancebased initiative had created a tension for both front-line staff and managers: they have a tool, AssetPlus, which, although seeking to introduce elements of desistance, still requires them to assess risk and consider historical, deficit-based factors. However, the Constructive Resettlement approach requires a move towards resettlement and sentence planning that is positive, future oriented and based on a child's identity shift, further exacerbating the bifurcated practice: When I'm countersigning assessments and looking at the plans I do think one of the hardest things to write is a plan in the positive. It's always easier to write we're going to work on this and we're going to work on that, but it's about trying to think about what do we want the end result to be . . . A lot of the time you might say I'm going to work on substance misuse. Well, that doesn't mean anything, does it? What does that measure? Whereas, actually, we want a young person to live without using substance misuse. I suppose it's around language, the way that we say stuff and it is that identity shift stuff and about young people not feeling that they are a label, really. (YOT Team Manager 2) Who wants to talk about something bad that you've done repeatedly? It's thinking about how do we move away from backward facing and negative formulated plans and work that we're going to do with young people to be more future orientated and positive. (YOT Team Manager 1) Ultimately, for some practitioners who were embracing child-first approaches, they felt that there needed to be changes to the inspection processes and criteria to reflect the values and approaches of individual youth justice teams. As highlighted by Smith and Gray (2019), there are multiple models of youth justice practice across England and Wales. If we have an inspection framework that is not flexible and adaptable to reflect the diversity of practice, it is likely that youth justice teams who do not have a model of practice based on RFPP will be at a disadvantage and may be likely to receive a lower inspection outcome: Until we have the discussion with the inspectorate and the inspection process fits more in with our values and principles as it is now, I think we're always going to struggle with this. Because, ultimately, nobody wants to 'need improvement'. We want to be 'good' or 'outstanding'. So, whilst you've got that pull, it were almost like those dodgy salesmen back in the day, where our commission comes from getting good and outstanding. So, you can spend very little time with a young person but do all the processes and look absolutely fantastic. But with no better outcomes for young people. (YOT Senior Practitioner) The contradictory messages, bifurcated practice, and increased paperwork were all taking front-line practitioners away from spending time with children in conflict with the law. It was perceived that the competing and contradictory demands of HMIP and the YJB were creating a sense of frustration in workers that pointed to the urgent need for a clear central narrative to emerge. The inherent warning that the demands were going 'too far' highlights the findings from Weston and Mythen (2021) and Bovarnick (2010) that contradictory messages and competing demands of different approaches can have cause frustration and stress among workers.

Confusion
There was confusion among practitioners about the meaning of key terms, and how they relate to their own practice. As stated, the interview data were taken from a wider evaluation of 'Constructive Resettlement' -an approach to resettlement developed by the YJB. As such, all staff were asked about their understanding of 'Constructive Resettlement'. A number of different perspectives were offered, and many were based on a practitioner's individual background and training. For example, a staff member from a Secure Children's Home felt that Constructive Resettlement aligned itself with the Secure Stairs Programme, a psychologically informed model of care that focuses on comprehensive, co-produced assessments of children that are individually tailored to meet all their needs: What I will say is that it fits in very well with the Secure Stairs Programme. Because the two things do absolutely go hand in hand, because in order to tackle some of the barriers, you've got to get to the bottom of some of the causes. The root causes for some of the behaviours. Because otherwise it just gets dressed up as kids that don't behave very well. (Education Worker, Secure Children's Home) Staff with experience of working with children who displayed sexually harmful behaviour drew comparisons with the Good Lives Model, a strengths based, future-oriented therapeutic programme designed to address sexually harmful behaviours: Okay yes, we're looking at, we introduced the Good Lives Training about five years ago, so that's pretty much the same thing. So, that's why it didn't feel like a revolution really. (Youth Offending Team Worker 6) One Youth Offending Team Manager compared Constructive Resettlement with Trauma Informed and Desistance approaches of working with children. She stated that the trauma informed approach had underpinned all their interventions and assessments with children for above 2 years: This was my feedback from the training that ***** did recently, is that two years ago, that training probably would have been quite interesting and informative for us, but two years on, there was nothing new for us in it, there was nothing ground-breaking or surprising, it was basically what we do. (Youth Offending Team Manager) For staff with a background in person-centred counselling, they could see parallels between counselling and Constructive Resettlement: I think there's an acknowledgement that, if we work with people where they are, instead of where we want them to be, that they do better. That they're less likely to reoffend, that they're more likely to be able to move on successfully and positively. And I'm liking the idea that we're building on strengths rather than focusing in on negatives, because I think . . . I've got a bit of a counselling background as well. And all of that is based in understanding people's strengths, so that they can build on that and go on, go forward rather than reminding people that they've made a mistake. (Youth Offending Team Worker 5) The understanding of Constructive Resettlement and the extent to which it was viewed as a change in direction of practice for staff therefore varied considerably and depended on their own individual backgrounds and experiences. There was a general pattern emerging that, for staff who had worked with other therapeutic and welfare-based models of intervention, that they viewed the Constructive Resettlement approach as a useful refresher that reinforced their own individual knowledge and practice. The comparisons to other models were rather general and seemed to focus on co-production, and welfare, and suggested that there may be some confusion specifically about what child first and desistance approaches are.
Confusion about how to work in child first and desistance-based approaches with children was also evident. Interestingly, staff wanted more practical guidance, training and a toolkit of resources that they could use with children: Obviously, I've got the slides, but it would've just been nice to have had a bit of guidance or some work that we could use with our young people. (Youth Offending Team Worker 3) However, evidence-based practice, intervention guidance and toolkits tend to be quite prescriptive and are considered part of the risk paradigm. Again, this suggests that staff have a degree of reticence about 'how' and 'what' to do with children without a prescriptive guide, suggesting that even where staff are trying to move away from risk and embrace child-first approaches, they are still dependent on risk-based methods to mobilise this.
Finally, confusion emerged about how to strike the 'hard balance' between risk assessment and management, and child-first/desistance-based approaches: So, it's about relationship building really, isn't it? And it's about having the time and the space to do that and potentially the paperwork and the policies and procedures around managing risk can undermine that. Well, that's a contentious thing, isn't it? People don't bring in these things because they want to undermine risk management. But there's a balance, isn't there? You know, it's a tricky one. It's a hard balance to find. (YOT Worker 2) Yes, we've got to protect the public and we've got to protect the young person. But like I say, a change of terminology for a start off, that's got to change. Capture the same meaning, but just change it somehow. And if someone is scoring at high risk, then they've got to manage that. But we're telling him that he's risky, and if we're telling him that he's going to be risky, then he's going to act to be risky. So we need to change that somehow, but I don't know how. (YOT Resettlement Specialist) Some sought to reconcile the 'hard balance' by focusing entirely on a child's identity shift and their future direction: I think they could come out and just focus straight on the future. I think while they're in custody, they get enough of, you're an offender, you've done this, and overcoming that offending behaviour. Coming out, I'd like to think that they come out into a community where they're treated exactly the same as everybody else and they should have the same opportunities. (YOT Education Worker) Others had moved away from offending behaviour work completely towards trauma informed approaches, which they felt complimented the implementation of Constructive Resettlement: And we just had a rethink and scrapped the lot. We don't do any offending behaviour programmes in ****** at all, we don't do any offence-specific interventions. We deal with the underlying causes, basically, so we look at the [inaudible] and we look at attachment, we look at trauma, we look at relationships, we look at support networks . . ..
A final comment outlined how a YOT Senior Practitioner felt a child's 'risk' could be managed by completely embracing a child-first approach, and abandoning the risk paradigm: Probably one of our biggest challenges, is risk and how we define, how we manage risk. So, at the moment we've got a young person who carries a knife or something like that. We say we put him on the Knife Crime Programme, yes, calling it 'lives matter', and then we complete a safety plan with him. And that is how we say we're managing his risk. Now, on the flip side of that, being future orientated, we've got this kid, who may have been carrying a knife in the past. Well, what does he want to do? So, if we can get him involved in something within his community, whether it be education, some positive activity, if he gets involved in that and that's how he sees himself and if that where he sees his future going, I would say, he would stop carrying a knife, so you've managed his risk. But I'm not sure how effective, I don't know any young person who walks around with their safety plan. But you know what, if I'm going to play football with my mates, I don't need to carry a knife. If I'm going to walk the street and I'm 15, 16, and I'm doing certain illegal activities, then actually, I'll carry a knife to protect myself. And you can write as many safety plans as you want, but that piece of paper isn't going to stop me from getting stabbed. (YOT Senior Practitioner) In the above example, a child's risk is 'managed' anyway by using child-first approaches, but allows one to move away from bifurcated practice by removing the language, tools and processes of RFPP from front-line practice.

Concluding thoughts
To effectively implement child-first approaches, a number of key challenges on the ground must first be addressed. One of the strongest themes from the interviews was that the 'balance' between risk assessment, management and offence focused work and desistance or child-first approaches was difficult to negotiate. Some staff could see the continuing role and importance of having some form of risk assessment and management as part of a youth justice practitioner's role; but the extent to which that directly impacted on the language used with children, and the work completed with them varied hugely. This difficulty appears to have been exacerbated by a growing awareness among front-line staff that the HMIP appears to prioritise risk assessment, risk management and the completion of offence focused work over child-first approaches in their inspections. This suggests a potential tension centrally between the YJB and HMIP about where the focus on youth justice work should lie. Until this is resolved, the 'difficult balance' and lack of clarity for front-line staff and managers will persist. A central and clear policy narrative from the centre is crucial, as evidence emerged in this study that the contradictory messages are impacting on practitioners' abilities to spend time and build relationships with children, and much to the frustration of front-line practitioners and managers. It is crucial that, to resolve the divergent messages from the centre about whose approach takes precedence, both HMIP and the YJB involve front-line practitioners in this dialogue, given the difficulties it is causing.
Equally, confusion about key terms such as desistance and child first, and how they relate to other 'welfare based' approaches such as trauma informed practice, The Good Lives Model, Secure Stairs and person-centred counselling were evident. Further training and resources should be made available to front-line practitioners that clarify this confusion.
Finally, the reasons for the reticence among some staff to move away from RFPP must be directly addressed. Evidence emerged in this study that RFPP has created a culture of fear among staff of 'getting it wrong', and that the risk assessment and management processes provided a degree of protection. Consideration therefore needs to be given to how staff can feel protected from wider scrutiny and criticism, while also being given the space to creatively work with children in ways that meet their individual needs and embraces child first approaches.
This article has highlighted the importance of engaging in a meaningful dialogue with front-line youth justice practitioners about how and what impact changes in central policy narratives are having at the 'coal face' (Case and Hampson, 2019: 33). A number of obstacles and challenges are evident on the ground to implementing the YJB's (2021) vision to become a 'child first youth justice system' (p. 9). The confusion and concern caused by the competing narratives of the YJB and HMIP must be urgently addressed and resolved centrally by engaging in a dialogue with practitioners locally. The mobilisation of child first is completely dependent on how youth justice practitioners and professionals interpret and understand this approach. However, until academics, policy makers and senior management structures meaningfully engage with practitioners about the challenges on the ground, there is a danger that the risk culture cloud will loom heavily over the emerging child first world.