Mental health, welfare or justice: An introductory global overview of differences between countries in the scale and approach to secure placements of children and young people

Abstract

Background

Estimates suggest that over a million children per year are deprived of their liberty across the world. Little is known about the types, ethos or distribution of secure beds in which they are detained.

Aim

This study aims to provide quantitative data with background information, to explore similarities and differences across jurisdictions, and to inform critical inquiry into key concepts and practices.

Methods

Data was obtained using an opportunistic sample of affluent countries, derived from an emerging academic/practice network of senior professionals. Depending on jurisdiction, data was already in the public domain or specifically requested. Data requests were related to the nature and size of health, welfare and criminal justice elements of secure beds and recent occupancy. Key professionals working in child secure settings, within jurisdictions, provided commentary on local approaches.

Results

Data was incomplete but allowed for comparisons between 10 jurisdictions. The proportions of the populations of children and young people detained varied by jurisdiction as did their distribution across variations of secure settings. Not all jurisdictions had all three kinds of secure settings. Definitions of secure beds varied depending on the use of relational, procedural or physical security.

Conclusion

Findings are tentative but suggestion solely considering numerical descriptions of children's detention is misleading; our study highlights ways in which comparative studies may be improved. Within reported jurisdictions, the framework of health, welfare and justice was meaningful but this may not hold true with a wider international application of this method. Open interrogation of this data would be enhanced by the inclusion of children's perspectives.

1. INTRODUCTION

A minimum of between 1.3 and 1.5 million children are deprived of liberty per year worldwide (Nowak, 2019). Of those, the largest number is deprived of liberty in welfare or care institutions (430,000–680,000), followed by those in justice facilities (410,000), migration‐related detention centres (330,000), in armed conflict situations (35,000) and for national security reasons (1,500). Additionally, approximately 19,000 children are living with their primary caregivers in prisons. These numbers are very conservative as they rely only on judicial or administrative decisions. Also global studies often do not include secure mental health beds. Literature providing a nuanced and detailed understanding of international similarities and differences is scarce (Cox & Abrams, 2021; Human Rights Watch, 2016; Souverein et al., 2019). A data‐based, international discussion about the philosophy and organisation of secure placement of children across multiple sectors is overdue. As outlined by Nowak (2019), we acknowledge that there are many different settings in which children are deprived of liberty, which increases the complexity of understanding crude numbers. For this paper, as a first step, we focused on three of the larger settings where children and young people are deprived of liberty by judicial or administrative authority: criminal justice, child welfare and mental health settings.

While across the globe differences between settings and jurisdictions are clearly evident, some common characteristics (e.g. isolation from society and limited social, leisure and/or educational opportunities) of secure settings may harm children. Additional risks of adverse experiences in secure settings include neglect, violence and sexual abuse (e.g., Medway Safeguarding Children Board, 2019; Spielman, 2021; van Ijzendoorn et al., 2020). Secure settings, particularly when combined with other adverse childhood experiences, can impair children's physical and mental health and their neurological, cognitive and social development (Desmund et al., 2020). Nevertheless, the UN study (Nowak, 2019) recognised the potential benefits of such settings and also noted that, in many jurisdictions, evidence exist of attempts by governments and practitioners both to reduce the number of children deprived of liberty and to improve conditions in the secure facilities. Acknowledging this nuance, when we consider the potential risks of these settings and the developmental vulnerability of children in secure facilities, especially those who come into repeated contact with the system, there is a clear need for a granular and sophisticated analysis of the scale, approach and impact of deprivation of liberty.

The aim of this paper is to enhance knowledge of the scale, ethos and approach across health, welfare and criminal justice sectors of secure placement, within countries represented in an emerging stakeholder network.

2. METHODS

We obtained and analysed reported rates of confinement of children aged under 18 years via mental health, child welfare and criminal justice legislation. This includes secure services: (1) for children and young people with mental illness/disorder (e.g., locked psychiatric hospital wards), in the sector of ‘mental health’; (2) to protect children and young people from abuse and neglect and risk to themselves (e.g., secure children's home), in the sector of ‘child welfare’; and (3) secure services for children and young people suspected of or convicted of criminal behaviour (e.g., youth justice institutions), in the sector of youth justice, including pre‐trial remand but excluding temporary police custody.

An international collaboration, largely of senior professional stakeholders, grew organically and in the past year has met monthly in online focus groups, using English as a working language. Not all network members were able to obtain the minimum data set for publication. We include data from 10 of the 13 jurisdictions represented in the network: Australia, Denmark, England, New Zealand, Northern Ireland, Norway, Scotland, Sweden, The Netherlands and Wales.

Each member searched their country's data sources for published information on available numbers of secure beds in each sector (health, welfare, justice), and the population under 18 (to enable calculation of rate of confinement) at a specified point in time since 2015. This approach was based upon the terminology and data templates of a successful 2016 study of English children and young people detained in health, welfare and mental health secure services (Hales et al., 2018). Data were given, where available, for number of secure beds available in each sector and a point‐documented census number of children or young people detained, or average number of occupied beds. Raw numbers were reworked to produce frequency data (of total available beds in each sector per young person and number of children and young people detained in each sector per 100,000 population of young people at a set point in time) for comparison by jurisdiction.

Initial data returned from multiple members revealed a diversity of terms and usages. Terminology was refined using a detailed presentation of definitions and approach, followed by group discussion exploring similarities and differences across sectors and jurisdictions. This facilitated categorisation and enabled basic international comparison within the context of country‐specific practices and definitions.

Not all numerical and linguistic ambiguities were resolvable. Therefore, in addition to quantitative data reporting the number and distribution of children and young people in secure settings, short jurisdiction‐specific narratives accompany the numerical results (see Appendix). These were written by a network member familiar with that jurisdiction. In these structured narratives, we discuss the definition of “secure” within the three sectors: the kinds of secure settings included, types of institutions (e.g., the scale), age cut‐offs and, if available, information about the demographics of children in each sector.

3. RESULTS

Within the 10 jurisdictions examined, the balance between secure provision of mental health, welfare and youth justice facilities varied as well as total number of children or young people detained. Not all the countries surveyed had designated secure health, justice and mental health facilities for children and young people. Scotland and Sweden have no specific secure mental health provision. Norway has no specific secure welfare provision. More detailed interrogation of the numbers revealed a variety of practices affecting day‐to‐day life for detained children, meaning that the definition of ‘secure setting’ is not ubiquitous. For each jurisdiction a more detailed description to support the numbers per setting is therefore included in Appendix.

Notably, not all countries were able to provide data on children and young people in all types of secure facility. The greatest difficulty reported was finding data to represent the number of children in secure hospital placements. Norway and The Netherlands have no disaggregated data on children and young people in secure hospital placements. For Australia it was not possible to obtain national data on the number of beds in secure mental health, welfare or justice, but rates of detention in justice were available. A table of raw data and references for all numbers is available in Appendix (Table S1, Appendix). This lack of robust data begs for a nuanced understanding of the numbers, particularly the numbers for the countries mentioned above: Norway, The Netherlands and Australia.

The number of available secure beds per 100,000 young people is shown in Figure 1 and distribution of the secure beds across mental health, welfare and justice in Figure 2. The lowest number of secure beds per 100,000 young people was in Norway and the highest in the Netherlands. However, for both countries this might be an underestimation as data on secure mental health beds is lacking. The distribution of secure beds within each country (see Figure 2) also varied. Justice provision accounted for more than 50% of identified provision in five countries (England, New Zealand, Northern Ireland, Norway, Wales) whilst welfare provision accounted for more than 50% of provision in four countries (Denmark, Scotland, Sweden, The Netherlands). Mental health provision in England and in Wales was greater than welfare provision and in Denmark was greater than the justice provision.

FIGURE 1.

Number of available secure beds within country across health, welfare and justice/youth population per 100,000 young people (N = 9)

FIGURE 2.

Distribution of available secure beds within country across health, welfare and justice (N = 9)

Looking at the descriptions of these numbers (see Appendix) there are considerable differences within and between sectors (mental health, welfare, justice) and between jurisdictions in the type of secure institutions in terms of design, scale and running authority. For example, in England the largest proportion of secure placements fall under youth justice legislations. However, exploring the description to the number, with regards to the type of institutions there is overlap between welfare and justice beds as children under these different legislations can be placed in mixed groups within the same secure institution. Or when looking at Norway, official numbers count no secure beds within child welfare. A more detailed understanding of these numbers however shows that children and young people are still deprived of liberty in seemingly ‘non‐secure’ facilities where children can still be prevented from leaving the institution and a range of other restrictive measures can be applied.

Where comparative numerical data was available (see Figure 3) rates of detention in designated secure mental health services ranged from 0 to 2.65 per 100,000 (highest in England), in secure welfare settings from 0 to 30.4 per 100,000 (highest in Sweden) and in youth justice settings from 0.53 to 12.7 per 100,000 (highest in Australia, see Table S1).

FIGURE 3.

Within country rates of detention per sector, rate per 100,000 young people (N = 10)

Overall rates of detention were calculated for seven countries (see Figure 4). As secure mental health rates of detention were not available for Australia and the Netherlands, overall rates of detention were not calculated. Sweden had the highest overall rate of detention and Wales the lowest. Sweden particularly stands out with high numbers driven by high numbers in welfare. To bring some nuance to these numbers, this may be partly due to differences in the history and organisation of the welfare sector.  Also these numbers most likely include secure settings that in other jurisdictions are labelled ‘non‐secure’ and therefore are not included in their numbers. Considering the rates of detention in Northern Ireland and Wales, it is notable that they are lower than the rates in England and the patterns across the sectors are different. The laws for England and Wales are the same and therefore differences between these jurisdictions are likely to be related to the interpretation of the law, cultural and contextual differences, or need for secure services; with children and young people in Wales either not needing or not thought to need placement in a secure setting as much as young people from England. In Northern Ireland the secure services have been developed recently and care has been taken to provide joined up services and devolution means that the laws are different from England.

FIGURE 4.

Overall rates of detention per 100,000 young people (N = 7)

Finally, age cut‐offs (see Appendix) varied between countries and within countries across sectors. For example, the minimum age of criminal responsibility (MACR) varies across the jurisdictions in our study, with the highest (15 years) in Norway, Sweden and Denmark and the lowest (10 years) in Australia, England, Wales, Northern‐Ireland and New Zealand. The Netherlands and Scotland have a MACR of 12 years. With regards to other characteristics of children and young people – where data was available – we see similar patterns across jurisdictions. For example, Australia, England, Wales and the Netherlands report an overrepresentation of young people from certain ethnic/cultural backgrounds in the justice sector. Also, moving from mental health to welfare to justice, the ratio of boys to girls shifts with boys being strongly overrepresented in the justice sector. For example, in Sweden in the mental health sector girls comprise 70% of placements, while in the welfare sector boys are the majority (70%) and in the justice sector the population in the facilities are almost entirely boys. Similar patterns can be observed in other jurisdictions like England and The Netherlands.

4. DISCUSSION

We have found significant variations in approach to the detention of young people within an opportunistic sample of western countries across the world, highlighting different ideological views on how to care for young people needing secure care. Differences emerged in overall and specific bed numbers and associated rates of detention. There was no consistency in the pattern or proportions of specific provision (mental health, welfare, justice) across jurisdictions. The majority of secure beds in Denmark, Scotland, Sweden and The Netherlands were based in the welfare sector whereas in England, New Zealand, Northern Ireland, Norway and Wales, the secure beds were mainly within the justice sector. England, Denmark and Wales were the only countries with over 10% of the secure beds in the mental health sector. Our findings are modest but they do support the utility of conceptual and data categorisation based on mental health, welfare and youth justice which, with caveats, could be more widely applied.

More detailed description by jurisdiction did allow for some nuanced understanding of ethos and practice. However, even with the supportive descriptions provided by representatives from each jurisdiction, our understanding of differences in approaches underlying the numbers remains incomplete. We therefore used this discussion to consider challenges (i.e. limitations) encountered so far; these challenges reflect the lessons we learnt while discussing the numbers within our group and thus outline the way in which comparative studies in the future can be improved.

Firstly, it was hard to source reliable, comprehensive, up‐to‐date and disaggregated data from all jurisdictions surveyed. Administrative records seemed to be particularly limited in the mental health sector. The lack of robust data was also noted by the UN global study (Nowak, 2019). Accurate accounting of the scale of secure placements, and measurement of outcome, are vital to understand the scope and quality of services, and progress made as a result of policy changes. Nowak (2019) stresses that failure to inadequately register and monitor secure placements raises the risk of human rights violations for the children and young people involved and this applies regardless of the nature of the secure setting. Our research highlighted common gaps in information across jurisdictions and illuminates areas where further analysis is required.

Secondly, even where data is available, crude numbers can also be significantly misleading; comparative detention rates must be understood within the historical, economic, political and cultural context of each jurisdiction, and requires multi‐professional expertise to bring meaningful depth to their interpretation. We warn against solely quantitative studies and recommend scrutiny by mixed international groups as an aid to policy and practice.

During our discussions we found significant differences in definitions of ‘secure’ and ‘non‐secure’, types of institutions, architectural design and related physical security. This made comparing data across jurisdictions and sectors difficult. Importantly, procedural and relational security practices which can dictate the tone of a setting are harder to map. Secure institutions may allow some degree of ingress and egress, and apparently open institutions may still isolate and confine children and young people. In addition, the multiple perspectives available on these more subtle distinctions, notably the perspectives of the detained child or young person, are not adequately represented in this paper (or other comparative studies). It is also important to remember that a setting that by official legislation and the professional perspective may be labelled as ‘non‐secure’ can be experienced as severe deprivation of liberty by a child or young person. Future research should focus on a more in‐depth understanding of the concept of ‘deprivation of liberty’ and ‘restrictive practices’, incorporating different perspectives and particularly the perspective of children and young people and their families.

Thirdly, age cut‐offs varied between countries and within countries across sectors. The variations in the MACR was not as large as the international variation (from age 6 in North Carolina to age 18, e.g., in Brazil) but there were some notable differences. The issue becomes more complicated when one considers the details of underlying legislation, for example, acquitting defences such as doli incapax (a legal defence whereby the child can be found not guilty as a result of not knowing that what they were doing was seriously wrong) which operates in some countries (e.g., Australia and New Zealand) but not in others (e.g., England), as well as stratification by offence type which operates in some countries (e.g., New Zealand) but not others (e.g., England). The United Nations General Comment Number 24 (2019) stated that ‘parties are encouraged to increase their minimum age to at least 14 years of age’. All countries in our study aside from Denmark, Norway and Sweden would be deemed to have a minimum age of criminal responsibility which is too low. The age of majority, specifically the age at which children are managed in secure systems in exactly the same way as adults, also varies. Emanating from neuroscientific research, there is growing awareness that emerging adults (aged 18–25) are a developmentally distinct group who may benefit from services targeted specifically for young adults (Matthews et al., 2018). We call for more detailed analysis of the historical and political reasons for variation in age cut‐offs, the economic implications of one system versus another and the direct and indirect impacts upon children and young people and their development.

Finally, besides age it is important to look at other demographics underlying crude numbers. The information provided by different jurisdictions shows similar patterns of disproportionate representation of certain groups—in terms of gender and ethnic/cultural background—across the sectors of health, welfare and justice. These patterns of demographics may reflect inequality of service provision and biases in processes of decision‐making about placement.

5. CONCLUSION

The work thus far highlights the areas in which more research is needed and may galvanise a critical discussion around the organisation and ethos of secure placement of children and young people. We consider this paper a starting point. Within these jurisdictions the distinctions between mental health, welfare and youth justice appeared to be a common framework. No jurisdiction reported a significant separate provision of secure educational establishments, for instance, despite these being under development in the United Kingdom where they have previously been a mainstay of reform provision. However, these categories still exclude a large number of other settings in which children are deprived of liberty, such as migration‐related secure facilities and police custody or those deprived of liberty de facto (i.e., when an individual is, in theory, free to leave an establishment but in practice could not do so). The paper further reports from a westernised perspective. The extent to which these fundamental categories of analysis apply meaningfully elsewhere in the world is unclear. Global collaborations and comparative studies require the participation of a much larger number of jurisdictions and potential conceptual shifts which may affect data collection and analysis. It is also important that children and young people can exercise their human right to participate in discussions and decision‐making that affects them. Children and young people are experts in their own lives and they have valuable insights into their experiences; that, as we discussed, may be a different perspective on what is ‘deprivation of liberty’ as compared to a professional or legal perspective. We therefore aim to amplify the voices of young people globally, particularly those who have been in secure settings, so that they can be part of our network to understand and advocate for improved care and outcomes for all children and young people.

CONFLICT OF INTEREST

The authors declare no competing interests.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Supporting information

Supporting Information S1

TABLE S1

Souverein, F. , Hales, H. , Anderson, P. , Argent, S. E. , Bartlett, A. , Blower, A. , Delmage, E. , Enell, S. , Eske Henriksen, A.‐K. , Koomen, K. , & Oostermeijer, S. (2022). Mental health, welfare or justice: An introductory global overview of differences between countries in the scale and approach to secure placements of children and young people. Criminal Behaviour & Mental Health, 32(3), 238–247. 10.1002/cbm.2234

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author; though almost all data was derived from publicly available sources (as listed in our references).