Abstract
Youth justice settings should provide safe, therapeutic environments, tailored to young people's needs. Current custodial models rarely meet these aims, mainly because a focus on security tends to outweigh an emphasis on care, diminishing rather than encouraging young people's positive development. This article reports on a three-year evaluation of youth justice reforms in the Netherlands, including small-scale, community-embedded facilities with an emphasis on relational security. We outline key operational elements and conditions for implementation of these facilities to provide guidance for youth justice professionals, and for managers and policymakers seeking to promote political and financial investments in effective youth justice strategies.
Introduction
The principle that children and young people should only ever be remanded or sentenced to a custodial setting as a last resort is firmly enshrined in the United Nations’ Convention on the Rights of the Child (UNCRC, 1989). Last-resort custodial settings should be places where young people's developmental needs are met, strengths and protective factors are built upon, and positive community connections are forged (as we detail below). While differences between jurisdictions are evident, conventional detention facilities are generally inconsistent with the need to respond to children and young people in the justice system through a therapeutic lens that promotes positive youth development. Instead, young people are frequently and routinely detained in punitive, large-scale institutions, far from their home and communities. The time for reform is long overdue. Failing to provide children and young people with genuine opportunities to get their lives back on track has detrimental effects, not only for the child and their family, but also for society at large.
There is a lack of evidence, however, about the operational elements and implementation of alternative custodial models. To address this gap, in 2015, the Dutch Ministry of Justice examined the feasibility and potential efficacy of implementing small-scale, community-embedded youth justice facilities. In 2016, three pilot sites were opened, including an eight-bed remand facility in Amsterdam (hereafter ‘the Amsterdam facility’). The general conclusion following these pilots was that such small-scale, community-embedded facilities could support justice-involved young people to get their lives ‘back on track.’ The facilities provided tailored security and care, close to young people's social environment and support, so that offence-related factors and young people's developmental needs could be more effectively addressed. In mid-2019, the Amsterdam facility was accorded permanent status (House of Representatives the Netherlands, 2019). This article is based on a three-year evaluation of the Amsterdam facility (from the end of 2016 to the end of 2019), using participatory action research. We reflect on lessons learned through the development of this ‘Dutch model’, and we consider the implications for replicating such models in different places, communities, and jurisdictional settings.
The Harms of Youth Incarceration
The potential harms of incarceration are well-documented. Research shows that characteristics of custodial settings – isolation and stigmatization, limited educational and job opportunities, and little contact with prosocial peers – are particularly harmful for children and young people (Cox, 2018; Mendel, 2011; Nowak, 2019). Children themselves report experiences of “fear, isolation, trauma and harm in addition to discrimination, stigma and disempowerment” (Nowak, 2019, p. 8). These factors, especially when combined with other adverse childhood experiences, can impair children's physical and mental health, and neurological, cognitive, and social development (Desmund et al., 2020; Nowak, 2019). For all these reasons, a predominantly punitive, risk and security-focused approach tends to foster, rather than curb, youth offending (Zoettl, 2021). Thus, youth incarceration can amplify violence, both committed by and against young people. We conceive violence in this context as complex and multiple, comprising violence used by as well as experienced by young people. We thus blur the line between young people as either ‘victims’ or ‘offenders’, instead seeing these categories as overlapping and inextricably entwined.
The developmental vulnerability of justice-involved young people is well-established, particularly those who come into repeated contact with the justice system. For instance, evidence consistently shows high rates of neuro-developmental impairment (e.g., Bower et al., 2018) and previous experiences of violence, abuse and neglect among imprisoned children and young people (Cox, 2018). Any subsequent failure to meet young people's health and developmental needs constitutes an additional form of neglect. Furthermore, youth justice system responses to children that are predominantly punitive, risk- and security-focused can be experienced as yet another – albeit legally legitimized – form of violence against young people. Young people who offend (e.g., by using violence), therefore, may very often have experienced violence at the hands of adults. Thus, the victim/offender binary becomes blurred. From this perspective, youth justice is not simply a matter of responding to harm caused by young people, but of addressing and working to counteract harms to young people. This behoves a therapeutic, strengths-based, developmentally-informed approach.
Strengths-Based and Developmentally Appropriate Approaches
Research highlights the importance of recognizing young people's developmental needs and competencies, while promoting strengths and protective factors. These include individual factors (i.e., self-esteem, autonomy, independence) and social environmental factors (i.e., stable housing, community ties, positive peer relationships, and employment opportunities) (Lerner et al., 2015; Ttofi et al., 2016). Strengths-based approaches that emphasize relational work are crucial, especially for supporting young people's attainment of education and employment goals (Johns et al., 2017). This means working with young people to strengthen and build family, peer and community relationships, rather than focusing narrowly on offence-related factors, in isolation from their social context (Johns et al., 2017).
Young people in youth justice facilities constitute a heterogeneous population with wide-ranging cognitive, psychological, and social needs (Hillege et al., 2017). An effective youth custodial system, tailored to diverse care and security needs, therefore requires a broad spectrum of interventions and environments. The last resort principle suggests that young people should not be held under higher levels of security than necessary to reasonably ensure public and institutional safety. Following this logic, a focus on security should not override the focus on building and strengthening protective factors and positive community relationships wherever possible (Austin et al., 2005; Souverein et al., 2022).
Besides the Netherlands, other jurisdictions have been implementing similar reforms, replacing ‘one-size-fits-all’ prison-like settings with alternative custodial models. Such models provide more developmentally-appropriate responses and work to maintain young people's community connections (McCarthy et al., 2016; Mendel, 2011). New York City, for example, in 2003, sought to address the problem of young people's arrest, prosecution and “placement” in a violent and chaotic youth prison, often hundreds of miles from home” (Schiraldi, 2022). The City introduced various diversionary and non-custodial programs that aimed to balance young people's security and care needs and, by 2011,the number of young people in New York's youth justice facilities had been reduced by half (Szanyi & Soler, 2018).
Under the 2012 Close to Home Act, the State of New York recognized that the wellbeing of everybody concerned “would be best served by minimizing the dislocation of youth from their families and building on positive connections between young people and their communities” (New York State Office of Children & Family Services, 2014, p. 2). This new law shifted responsibility for custodial placements from the State to the city. Instead of placement in geographically isolated large-scale prison-like State facilities, all justice-involved young people were placed in “small homes” (6–20 beds) (Schiraldi, 2022) within the city, closer to their family and local community. Striking results have emerged: between 2012 and 2021, New York City has seen an 86% decline in youth arrests (Schiraldi, 2022). Program outcomes are equally encouraging: 91% of young people passed their academic classes; 82% successfully transitioned home to their family; 91% of young people who transitioned home enrolled in community-based programs; only 7.6% of young people were recalled to custody for violating their terms of release (Schiraldi, 2022; Weissman et al., 2019). The Close to Home initiative points to the need for deeper understanding of how “better treatment of young people and improved public safety [can indeed] go hand in hand” (Schiraldi, 2022).
The Aim of this Article
Trends towards small-scale, community-embedded youth facilities highlight the need to systematically examine these models’ implementation and operation, to understand how positive results might be replicated in other places (Szanyi & Soler, 2018). Such evidence is important, both for practitioners working with young people, and for managers and policymakers seeking to promote effective youth justice strategies. In this article, by shedding light on recent reforms in the Netherlands, we aim to build knowledge about implementing viable custodial alternatives for young people. This article is based on a comprehensive three-year evaluation 1 of one of three pilot facilities: a small-scale, community-embedded facility in Amsterdam. Through action research, the evaluation provided a unique opportunity to examine the development and implementation of the Amsterdam facility, including the perspectives of practitioners, policymakers, young people and their families. Our objectives in this paper are to: 1) identify the key operational elements of a small-scale, community-embedded facility, and the lessons learned in developing them ‘from paper to practice’; and 2) identify the facilitators and barriers to implementation.
Method
Setting
This study was conducted as part of a three-year evaluation project initiated by the Dutch Ministry of Justice to examine the feasibility and efficacy of alternative custodial models for youth justice. From the end of 2016 to the end of 2019, a research team of the Academic Collaborative Centre for Forensic Youth 2 (AcademischeWerkplaatsRisicojeugd; AWRJ) carefully monitored three pilot facilities: Nijmegen (November 2016 – December 2017), Groningen (December 2016 – December 2017), and Amsterdam (September 2016 – December 2019). These facilities each accommodated eight young people, pre- and post-trial, following the same vision and practical framework. This article focuses on the Amsterdam facility, largely because the Amsterdam pilot ran the longest and therefore yielded the greatest amount of data.
Action Research
The evaluation involved action research, engaging all stakeholders in a cyclic process of action, research and critical reflection. This approach facilitates stakeholder ‘buy-in’ and understanding of complex processes in practice, and thereby enhances the applicability of study outcomes (Abma et al., 2017; Nyström et al., 2018). The action research cycle comprised qualitative, semi-structured interviews (to gather the perspectives of key stakeholders), accompanied by an iterative validation, feedback and reflection loop.
Semi-Structured Interviews
In determining our sample, source triangulation increased validity of the results. Our sample comprised three groups of stakeholders: professionals, young people in the facility, and their parents/caregivers. A total of 67 interviews were conducted in four rounds during the evaluation period, September 2016 to December 2019.
We conducted 44 semi-structured interviews with professionals. Purposive sampling was used to ensure representation across all relevant youth justice agencies and different roles within these organizations. These included staff working at the pilot facility (15 interviews) and other professionals (29 interviews). The staff included the behavioral expert (responsible for the assessment and planning of young people's trajectories), the project manager (responsible for management of staff and facility at large), and unit staff including security staff (n = 5) and social workers (n = 7). The other professionals included representatives from the child care and protection board, probation services, youth lawyers, the public prosecution office, youth court, custodial institutions agency, the municipality, and a local residents committee in the neighborhood where the facility was located.
During the evaluation period, 156 young people were remanded to the Amsterdam facility (the total sample). Of these, 18 young people were approached to participate in the evaluation, which resulted in 15 interviews. The interview sample was selected through a combination of availability sampling (guided by which young people were placed in the facility at the time of the interview rounds), convenience sampling (guided by the process of data collection and analysis), and purposive sampling (to ensure a heterogeneous representative sample). We ensured our sample represented the diversity of young people in the facility, regarding age and offence severity. The participating young people were aged between 14 and 18 years (ages varied between 13 and 18 years in the total sample). They were remanded for either a violent or property crime, or a violent property offence, which reflected the most prevalent offences among the total sample (18% violent, 22% property, 55% violent property offence). Parent/caregivers of the 15 young people interviewed were approached to participate in the research (convenience sampling), which resulted in eight interviews, in most cases with the biological mother.
A trained and supervised research team, including authors FS and SO, conducted the interviews. A member of the research team explained the nature and objective of the study to each participant, who received a plain language statement and provided their informed consent prior to participation. Each interview lasted approximately 1 hour. New themes that were not anticipated with the initial topic-list were followed up on in subsequent interviews.
All interviews were tape recorded, transcribed verbatim, and uploaded into MAXQDA. Verbatim transcripts were coded employing a method of ‘thick analysis’ (Evers, 2015) that included thematic and open coding. Axial analysis was then used to build categories by sorting, integrating, and grouping the thematic and open codes. Iterative analyses involved constant comparison of themes emerging within and between participant groups (Boeije, 2014). In the last round of interviews, data saturation occurred (Boeije, 2014), in that no new themes emerged from participant narratives.
Validation, Feedback and Reflection
Validation, feedback and reflection took place through 1) project team meetings, 2) onsite observations, and 3) advisory board meetings. These activities improved the validity and reliability of the results and their applicability in practice. First, throughout the duration of the study, the results were continuously discussed and refined during reflective project team meetings (including FS, SO, EM, LvD, AP). In these meetings, the results were used as a basis to formulate and evaluate practical recommendations. In general, there was inter-researcher agreement about the results, interpretation, and recommendations. Second, about three-weekly onsite observations involved the researchers’ spending a day at the facility interacting with young people and staff. This allowed the research team to develop a good understanding of the research setting and daily practices. These observational data were used to contextualize and supplement the interview data. Third, as part of the pilot, local and national government representatives and managers and practitioners from stakeholder organizations met every six weeks to discuss the pilot's progress and formulate actions. Observations during these the advisory board meetings – about developments in practice and interactions between stakeholders – also served to contextualize and supplement the interview data. Further member validation (Birt et al., 2016) was undertaken through these advisory board meetings after each cycle of data-collection and analysis. Apart from some linguistic modifications, no major changes were suggested by advisory board members, indicating validity and practical applicability of the outcomes of this study.
Results
The interview data indicated wide support for the Amsterdam facility, not only amongst professional stakeholders, but also young people and their families. Interviewees generally agreed that the facility promotes positive youth development, and that reduced recidivism would be expected. In terms of operational findings, the following section details the main elements of the Amsterdam pilot, after which we highlight key lessons learned through implementation.
Key Operational Elements
From our results, seven key operational elements were identified: 1) assessment and indication; 2) small-scale and tailored; 3) relational security; 4) community-embedded; 5) integrated and multidisciplinary assessment, planning and service delivery; 6) active collaboration with young people and their social network; and 7) selection, guidance and supervision of staff.
Assessment and Indication for Placement
In setting up the working processes, it was recognized that strategically matching justice-involved young people with the appropriate setting requires structured assessment and indication for placement of each young person. The professionals stressed that this may also buffer against net-widening (referring to the risk of alternative sanctions, intended to divert people, actually increasing custodial placements). At the Amsterdam facility there were no indications of increases in remand placements after the start of the pilot. This facility focused on young people who previously have been placed in a large-scale, high-security facility, but who were deemed appropriate for the lower level of security and able to benefit from the opportunities offered by a community-embedded facility (i.e., continuation or initiation of school, work and care in the community). Throughout the pilot, stakeholders developed a set of criteria for placement through consensus-building and continuous learning processes.
Assessment – emphasizing strengths – involved collaboration between multiple youth justice agencies. In line with this approach, professionals, young people and parents/caregivers highlighted two criteria: 1) motivation, and 2) protective factors. First, to be placed in the community-embedded facility, young people had to demonstrate motivation to comply with facility regulations and responsibilities and be willing to accept staff support and guidance. Young people also had to express at least some desire for ‘a wake-up call’ to get their lives back on track. Second, sufficient protective factors (e.g., school/work engagement, a positive social network, professional care) – or potential to develop these – needed to be present. These factors would buffer against risk factors and help facilitate positive outcomes for them. Other criteria identified – offence severity and circumstances, offence history, support network, mental health and intellectual abilities – were weighted for each individual assessment according to a young person's security and care needs.
Several lessons were learned through developing these criteria and working processes. Initially, for instance, professionals agreed young people offending for a firstor second time would be most suitable for placement in the facility. During the pilot, however, this shifted: the professional stakeholders realized that, in practice, offence history is not necessarily related to a successful placement, as it does not equate to a young person's current motivation to get their life back on track. Similarly, it was originally not possible for young people to be placed in the facility multiple times. During the pilot, the perspective of professionals shifted towards allowing young people to learn by trial and error. This meant every potential placement in the facility was regarded as a ‘new situation’, with consideration being given to whether the facility was the most suitable place for the young person at that time. As the pilot progressed, less emphasis was put on the severity of the alleged offence. According to interviewees, offence severity should not be a discriminating factor in determining whether young people should be placed in the facility. In general, the offence circumstances (motive, role, provocations) were considered more indicative of whether a young person was suitable. Likewise, problematic substance abuse and an IQ lower than 70 (as a proxy for intellectual disability) were initially considered contra-indications for placement. However, in practice, it turned out that it was more appropriate to weigh these factors against the other criteria. Finally, practice showed that it was important that young people and their network were actively involved in the process of assessment and indication for placement. A young person's current volitional state of motivation, attitudes, and feelings could not simply be determined from a case file.
Small-Scale and Tailored
The design of the facility was small scale (one unit with eight beds). This allowed staff to identify and understand each young person's risks, needs, skills and strengths, and to tailor individual trajectories accordingly. The facility offered differentiated arrangements of security and care. This meant level of access to phones or spaces, for instance, could be scaled up or down depending on young people's individual risks and needs. This tailored approach also meant that there were no standard sanctions. Individual arrangements were made with each young person following a rule violation. Staff initiated informal activities and organized a communal dinner every day, but there was no structured group-work, nor was there a fixed program with set times. Instead, each young person had a weekly schedule customized to their individual needs and activities.
Developing this way of working in practice, professionals learned it was important to find a good balance between a fixed structure and tailored trajectories; an individual approach should not be confused with a lack of structure. Customization, especially in an environment with less physical and procedural security, required a clear structure and strong underpinning principles. In developing this way of working, relational security became the basis of this framework (see next subheading). The results suggested that to provide customized care and security, it was important to gather a complete picture of the young person and their context, to understand how to reduce risk factors and strengthen protective factors. For this, professional and policymakers developed an integrated and multidisciplinary style of working (see subheading further below). Finally the results indicate that the small-scaled and tailored way of working reduced the risk of undesirable group formation or the explicit teaching of deviant behavior by group members (deviancy training). In interviews, both young people and staff expressed that, because each trajectory was carefully tailored to individual risks and needs without a strong group-based approach, young people were more focused on themselves and their future, rather than on the other young people and their position within the group.
Relational Security
While some physical security (e.g., locks and cameras) and procedural measures (e.g., regarding external movements and possessions) were in place, the facility was set up with relational security as the core of the security framework. Relational security relies on the relationships between staff and young people to ensure security (Hilhorst, 2016). A more concrete conceptualization of relational security, however, was not defined on paper. Being a pilot facility, one of the aims was to develop a good understanding of the concept and practice of relational security, and how it ensures safety. Also, professionals initially wondered if relational security could be the basis of the security framework, or whether they would need to rely predominantly on physical and procedural security.
The experiences of professionals, young people, and their parents/caregivers highlighted that relational security was grounded in three distinct, but interrelated, elements. The first concerned the manner in which professionals were present and shaped interactions with young people. When they worked from a ‘basic attitude’, this contributed to safety. This basic attitude comprised the connection with and attunement to each individual young person and their context. Professionals actively invested in getting to know each young person, and were able to see things from their perspective. This involved the professional accepting the young person as they are, giving the young person “space” to be themselves and showing genuine interest. By finding a connection, the professionals were able to understand the young person and use this understanding to promote safety.
The second component of relational security was a constructive equal collaboration between the professional and young person. The professional stood next to and walked alongside each young person as a coach. As co-owner of their case plan, the young person was involved in decision-making (see active collaboration subheading, below). The young person was given autonomy and space to take responsibility and to learn by trial and error. Following any rule violation, the young person was included in the settlement and determination of an appropriate consequence, with the emphasis on restoration and the underlying causes of the behaviour instead of punishment. Clear boundaries were set, as required, but the professional's attitude was advisory and motivational, rather than repressive and authoritarian.
The third component of relational security related to professionals’ being physically present at the unit. This guaranteed safety in several ways: the preventive effect of the adult's presence; their observing, signaling, intervening early, and de-escalating if necessary; and professionals’ being available for young people to seek emotional support. Furthermore, the many opportunities for informal contact contributed to relationship-building between staff and the young people, which helped generate a positive institutional climate.
The evaluation results showed that, when these three components are in place, relational security can provide an effective core of the security framework. First, professionals relied primarily on relational security strategies to ensure safety, and only sought secondary support through physical and procedural security measures. Second, these latter measures were deployed in line with the principles of relational security, that is, always putting the relationship with the young person first and providing tailored reponses. This way of working allowed staff to establish a safe and therapeutic environment, which promoted important developmental competencies (e.g., autonomy). In interviews, staff, young people, and their parents/caregivers all reported experiencing high levels of safety, with very few reports of violent incidents.
Because young people were given a degree of freedom and autonomy, they felt responsible for promoting a safe environment and not breaking the rules, and were motivated to take on that responsibility. Young people indicated that, though there were opportunities to abscond, they consciously did not do so, because they viewed the facility as a valuable opportunity. Also because they knew what was at stake if they did try to run away - transfer to a large-scale facility. Staff and young people reported good, constructive alliances characterized by mutual respect. Also, the risk of undesirable group formation or the explicit teaching of deviant behavior by group members (deviancy training) appeared to be reduced by relational security. Staff reported a safe working climate and high job satisfaction.
To effectively establish relational security, the physical infrastructure was important as it must allow for physical security measures to be scaled down as much as possible. The Amsterdam facility was renovated to fit this purpose: it had a homely look with no fences, no window bars, no steel doors, and ‘normal’ furniture. According to the professionals, this automatically strengthened relational security. Instead of falling back on physical security measures – because these options were not available – professionals were more inclined to adopt de-escalating behavior and invest in constructive relationships with the young people.
Community-Embeddedness
One of the central principles of the facility was that it was embedded in the community. This meant, in the first instance, its location in a neighborhood close to each young person's home environment (maximum 1.5-h travel time). Second, the facility utilized resources within the community. For example, it did not offer any internal school, work, therapy or medical care, but collaborated with local authorities to continue or initiate these within the community.
The broader evaluation showed that in almost all cases (97% for care and 98% for school and work-related activities), protective factors were successfully continued or initiated within the community. All participants (professionals, young people, and parents/caregivers) emphasized that promoting protective factors in this way enhanced the young people's positive development, which they would expect to contribute to reduced recidivism. Further, due to the community-embedded approach, even while incarcerated, young people did not feel excluded from their family, community, or society at large. The fact that the facility mostly relied on relational security (see relational security, above), and the physical infrastructure and location of the building, promoted this connection with the community.
Integrated and Multidisciplinary Assessment, Planning and Service Delivery
In line with the principle of community-embeddedness, the facility took an integrated, multidisciplinary approach to assessment, treatment planning and service delivery. From the moment of placement in the facility, case management was directed towards the period after release. Continuity was considered key: case management was provided by external professional stakeholders (e.g., youth probation service) who remained involved after the young person's release. The underlying principle was that the period of incarceration, especially when it is relatively brief, is considered part of a continuous life-course trajectory towards desistance; the facility is only one link in the chain. In practice, this meant the facility was easily accessible for local stakeholders, and regular case meetings with stakeholders (including youth probation services, child protective services, mental health care professionals) were held on the premises. These meetings involved information-sharing about the case and collaborative treatment planning, incorporating different intake and diagnostic instruments used at different times by various partners along the youth justice chain.
These collaborative working processes promoted integrated service delivery. However, it remained a challenge to produce one integrated case plan. The idea was that following the stakeholder meetings, instead of each stakeholder writing a separate case plan, one individual action plan was formulated as a ‘co-production’ between all key stakeholders. This integrated case plan comprised short-term and long-term goals geared towards building prospects and opportunities. This aim was not achieved as it was hindered by legal, practical and technological barriers (see barriers to implementation, below). Information was shared verbally, and services were generally delivered in an integrated manner, but this was not substantiated in structural or formal data-sharing processes or a sustainable and holistic ‘inter-agency’ action plan on paper.
Despite this hindrance, the results showed that this integrated way of working increased the probability of ‘getting things done’ and arranging the appropriate post-release support. Young people stated that this community-embedded approach allowed for their smooth transition back home. At the same time, for some young people, the contrast between the structured setting of the facility and their home environment still posed a challenge for their adjustment following release. This highlighted the importance of staff from the facility having opportunities to stay involved post-release to provide guidance to the young person and their family.
Active Collaboration with Young People and Their Support Network
In line with the concept of relational security – recognizing young people's agency and autonomy – the facility was based on the principle that young people and their support network should be actively engaged in intervention planning and service delivery, through active collaboration. This stemmed from an understanding that a tailored approach must be informed by young people's and parents/caregivers’ perspectives on what is needed to reduce risks and strengthen protective factors. This was mainly organized by allowing young people and their network a seat at the table during case management meetings with professional stakeholders. Further, the focus on relational security and the physical infrastructure of the building made the facility easily accessible for parents and young people's broader support network (e.g., no fixed visiting hours, no visitation, facility was within reasonable distance to family, the building had a welcoming appearance). Staff members kept parents and family members frequently updated and family were able to call staff directly if they had any questions or concerns. Staff actively engaged with parents/caregivers and helped them take responsibility in their caregiving role, according to their ability, such as by providing transport to court.
Actively involving partners/caregivers provided valuable insights into family dynamics, which informed case management. Staff, young people, and parents/caregivers said this allowed for a smoother transition back home after release. Young people and their network experienced some sense of autonomy and self-determination, by having a seat at the table, but only if their input was reflected in concrete actions. However, this was not sufficient: young people and their network stressed the need for active involvement in planning and decision-making throughout their confinement and having their input reflected in their case plan and activities.
Selection, Guidance and Supervision of Staff
Selection, training, and managerial support of staff is the final operational element. Staff were responsible for establishing a safe and therapeutic environment. Their relationships with young people were especially important given the emphasis on relational security. In developing this way of working from paper to practice, several lessons were learned. In terms of staff selection, rather than professional qualifications, staff skills were often the main deciding factor in determining suitability for the job. Most important was the ability to genuinely connect with young people, which was thought to be a natural, innate ability rather than a learned skill. It was deemed equally important that staff values and work ethics were in line with the underlying values of the model (that is a strengths-based, rather than a punitive risk-averse approach, emphasizing young people's perspectives within a tailored and collaborative way of working and promoting young people's autonomy through relational security).
Given the complexity of working in a youth custodial setting, structures to guide staff and allowing them space and time to learn on the job were paramount. Staff members were encouraged to continuously reflect on their own and their colleagues’ behaviors. Staff and management engaged in weekly, intensive supervision sessions. Combined with transformational leadership (using inspiration and innovation), this allowed for any repressive attitudes or feelings of unsafety among staff members to be addressed/counteracted. For these reasons, staff interviewed reported high work satisfaction and a positive work climate.
Implementation Lessons: Facilitators and Barriers
Our analysis revealed a range of factors facilitating or hindering implementation of the small-scale, community-embedded facility in Amsterdam. These included broader political conditions that were conducive to policy and practice reform and innovation; aspects of the pilot project and its local context that ensured its gradual development and garnering of wide support; and elements of monitoring and evaluation that promoted confidence in the reforms. We outline facilitators and barriers to implementation, and conducive conditions, in Table 1, below.
Overview of Conducive Conditions for Implementation: Facilitators and Barriers.
Discussion and Conclusion
In this article, based on three years of action research, we reflect on lessons learned through the development and evaluation of a small-scale, community-embedded youth justice remand facility in the Netherlands. Suited for a diverse population (Souverein et al., 2022), this approach contributes to a graduated justice system, oriented towards meeting young people's developmental needs, while maintaining a balance between security and care. By building on existing community resources, the facility allowed for protective factors (such as connections to school, work, sports, or professional care) to be continued or initiated during the period of confinement (Souverein et al., 2020). These results were long-lasting, as indicated by a follow-up study that showed for a large part these protective factors remained a year after release (Souverein et al., 2020).
Aspects of the small-scale, community-embedded model that appeared to be conducive to positive outcomes for young people were: 1) assessment and indication, 2) small-scale and tailored, 3) relational security, 4) community-embedded, 5) integrated and multidisciplinary assessment, planning and service delivery, 6) active collaboration with young people and their social network, and 7) selection, guidance and supervision of staff.
These key operational elements find their support in practice and theory. The model addresses both criminogenic needs as well as protective factors (as stressed by Andrews & Bonta, 2010). In line with literature on positive youth development (e.g., Johns et al., 2017; Lerner et al., 2015), these key operational elements promote important developmental competencies, both individual (i.e., self-esteem, autonomy, independence) and within the social environment (community ties, education and employment opportunities, and positive connections with their family and support network). These key operational elements promote a safe and positive environment for both young people and staff. They seem to buffer against risks and antisocial behavior being exacerbated within the group, or deviancy training, a common challenge in youth justice settings. This small-scale, community-embedded model – eschewing a strongly punitive, risk-averse approach – provides young people with an opportunity to get their lives back on track and promotes their motivation to do so. The success of the Amsterdam facility resulted in wide support from policymakers, professionals, young people and their families.
This article highlights crucial factors in the successful implementation of the model, which facilitate its key operational elements, and identifies barriers that may hinder implementation (see Table 1 for an overview). These factors align with other literature on youth justice reforms (Case & Hampson, 2019). For instance, this study reflects similar facilitators and barriers as those encountered with the earlier New York City reforms (Szanyi & Soler, 2018; Weissman et al., 2019). Most critical to the reforms in New York was, and remains, “achieving consensus among all stakeholders on the core vision and basic principles of the initiative and maintaining their support throughout the implementation process” (Szanyi & Soler, 2018, p. 16). The same can be said about the Dutch model.
It should be noted that even if the conditions for implementation are all in place, the developments in Amsterdam show that it remains a challenge to align the different roles and responsibilities of all the professional stakeholders. Relations between local partners and the national custodial institutions agency is a particular sticking point, as elaborated below. Setting up integrated data-sharing processes and co-producing one integrated case file and action plan proved problematic due to structural and policy constraints. These challenges persist to this day.
Development and implementation of the key operational elements should be regarded as a continuous process that requires ongoing investment of time and resources. Youth justice authorities currently struggle to effectively utilize the existing small-scale facilities. In the Netherlands, there are currently five small-scale facilities that are fully operating, but are vacant for the most part. With the exception of the Amsterdam facility, it appears that for the other four facilities, not all of the conditions for successful implementation have been met. This requires allocation of the appropriate time, attention, and resources. Recent attention in the media (Hindriks & Karawazi, 2022) and politics (House of Representatives the Netherlands, 2022) has suggested concern for the future of these facilities if they do not reach full capacity.
One of the bottlenecks is a disagreement between local partners and the custodial institutions agency of the national government. This disagreement centers on the question of who should have authority to decide if a young person is placed in a small-scale or large-scale facility: the national government (currently the case as decided by the government) or a judge (the general conception of local partners). The local partners’ argument is that the judge is in direct contact with the young person, parents, and local professionals and therefore has all the information needed to indicate placement in a local, small-scale facility. In contrast, Ministry of Justice officials need to make this judgment from a paper case file. This issue underlines a more fundamental question about whether a judicial power or government officials should decide the appropriate response to a young person involved in the justice system. This reaffirms the need for consensus among stakeholders about the underlying vision and carefully considered working processes.
A final reflection on the ‘Dutch model’ is that is it currently available for only a sub-group of young people in custody. As we have described, there is a thorough screening and indication process prior to placement. Further, during their stay, young people can be transferred to a large-scale facility with higher levels of physical and procedural security in certain situations (e.g., use of severe violence or absconding). This happens in about 14% of all placements (Souverein et al., 2020). This raises the questions of whether and how this model can be applied to the wider, diverse youth justice population.
In addition to the practical considerations of the Dutch model, the evaluation brought several methodological and conceptual issues to light. First, while multiple procedures were followed to increase the validity and reliability of the results (method and source triangulation, peer debriefing, member validation), one may question whether all results will transfer from this specific context to another youth justice setting. At the same time, while this paper only reports on the results of the Amsterdam facility, in the first year the evaluation also focused on two other pilot sites generating similar results (Souverein et al., 2018). Second, even though maximum diversity in the sample of young people and parent/caregiver participants was sought, the study relied on their willingness to participate. Young people and parents/caregivers who were not willing to participate could have had a different opinion on the topics addressed. Finally, data systems in practice are not necessarily fit for monitoring. To collect information about young people during and after their period of incarceration as reliably as possible, information from different systems of organizations was compared and combined for each young person individually. Besides being very labor-intensive, it turned out that, even with this detailed way of doing research, mapping the trajectories during and after placement was a challenge. The way information is now recorded in practice makes this type of research, particularly collecting follow-up data, almost impossible (information is missing, inconsistent with other sources or not accessible) and/or very labor intensive. It is not possible to make substantiated statements about the effect of judicial measures without thorough and adequate data collection along the youth justice chain. A unified, national information management and tracking system, with the young person as the starting point, will not only help practitioners work together, but will also promote sound scientific research.
Despite the practical and methodological considerations outlined above, this study contributes to filling a gap in the current literature about key elements of effective custodial models and what is needed for cross-jurisdictional implementation. We realize that each of the aspects described in this paper holds enough depth to justify several separate papers. Souverein et al. (2022), for example, provide a detailed qualitative description of the domains that guided screening/assessment for placement. Meijer et al. (2023) focus on the suitability of the Amsterdam facility for young people with a mild intellectual disability. The current paper was written to provide an overview of the Dutch reforms and the Dutch model, to provide guidance for practitioners working with young people, and for managers and policymakers seeking to promote political and financial investment in effective youth justice strategies.
At the same time, individual jurisdictions and their histories, systems, cultures, and politics, “can be just as quirky and esoteric as individual human beings” (Cavadino & Dignan, 2006, p. 452). Jurisdictions are distinguished by the interaction of various social, cultural, political, and economic factors. These interactions are unpredictable, either facilitating or undermining attempted reforms, and thereby pose distinct challenges for policy transfer (McFarlane & Canton, 2014). Any attempts to implement policy and/or practice from elsewhere should always involve considered analysis of the local context – identifying specific facilitators and barriers that could arise – and potential adaptation to meet local needs. Crucially, the principles and guidelines underlying any place-based reforms should be discussed and developed in collaboration with the local community, so that the needs and strengths of their young people are accommodated. Considering all of the above, we can (and should) redesign and reconfigure custodial models and promote developmentally-appropriate responses to justice-involved children and young people, everywhere. Crucially, we must act to ensure that – even in local, small-scale facilities – depriving a child of their liberty should only ever be the last resort. All necessary efforts should be made to prevent young people from reaching that point.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors received financial support from the Dutch Ministery of Justice for the research, authorship, and/or publication of this article.