Defining severe and multiple disadvantage from the inside: Perspectives of young people and of their support workers

The aim of this study was to explore the way young people facing severe and multiple disadvantage and their support workers perceive the circumstances of the young person. The leaders of 11 U.K. and five U.S. support ‐ providing organisations identified 30 young people aged 16 to 25 who were in great need and 35 workers for this study. Thematic analysis of the semi ‐ structured interviews found that shame was a primary emotional reaction to severe and multiple disadvantage. A history of poor relationships led them to develop an acute assessment of others and a tendency to back away from help. Over time, they perceived themselves to be unworthy and unable. Findings underscored the young people's role in shaping their trajectories and shone light on the role of emotions in explaining disconnection from support. supports. The experience of being let down by past relationships creates mistrust and acute assessment of support. The combination of risk, shame, a deep distrust of help and distancing from society leads young people to feel unworthy and unable.


| INTRODUCTION
There have been many attempts to define severe and multiple disadvantage, reflecting public system inclusion criteria and epidemiology. The diversity of definitions is evident in the professional and academic language, with terms such as multiple disadvantage, dual diagnosis, multiple and complex needs, high support needs, multiple disabilities, multi-problem, poly-victimisation, and adverse life experiences widely used (for a review of definitions, see Burnside, 2012;Finkelhor et al., 2007;Keene, 2001;Porche et al., 2017;Rosengard et al., 2007). The words tend to capture the mixture of the risks people face, the condition as defined by professionals, and the legal or administrative sanctions for the provision of money or services.
However, despite the increased recognition of "users" voices in relation to welfare state services (Beresford, 2002), there have been few opportunities for people facing severe and multiple disadvantage to contribute to and therefore shape the scholarly definitions and classifications concerning their condition (Rose & Thornicroft, 2010). In this article, the author briefly reviews the research literature on the two main approaches to understanding this population. The author then discusses studies that offer a different view of people facing extremely difficult circumstances before describing a study addressing this issue. After presenting the findings, future theory, research, and policy implications are considered.

| A SOCIAL POLICY APPROACH TO UNDERSTANDING SEVERE AND MULTIPLE DISADVANTAGE
Classification is intrinsic to social policy (Pinchbeck & Hewitt, 1969). Definitions change to reflect governmental concerns at the time legislation or other policy is framed. Axford (2017) showed how different lenses focused on impaired development, restriction of human rights, loss of quality of life, or social exclusion ruled needy children in or out of eligibility for state support.
A comprehensive literature review of multiple and complex needs (Rosengard et al., 2007) identified at least nine sets of labels, including (1) mental health, including severe and lasting disorders; (2) medical dual diagnosis; (3) disadvantage due to age or transition from one stage of life to another; (4) geographical or temporal disadvantage, for example refugees fleeing abuse and violence; (5) cultural and circumstantial exclusion; (6) marginal, hard-to-reach, and social exclusion; (7) risk to self, service staff, or the community; (8) structural disadvantage such as poverty; and (9) severe or long-term impairment or disability. Bramley et al. (2015) linked three administrative data sets on people using services in England between 2006 and 2012. They estimated that 15 in every 1000 adults of working age experienced some combination of homelessness, substance misuse, and offending behaviours. The study revealed that gender, age, poverty, childhood trauma, and poor educational experience played some role in the disadvantage. The most common profile was a male, aged 25-44, with a long history of poverty, difficult family relationships, and poor educational background. Famutimi and Thompson (2018), drawing on data from the Substance Abuse and Mental Health Services Administration (SAMHSA), found that a third of the U.S. homeless adults admitted for substance use treatment between 2005 and 2015 had a co-occurring mental disorder. Rankin and Regan (2004) synthesised definitions by giving priority to two types of vulnerability: breadth and depth of need. Breadth of need captured interconnected risks that demanded the resources of several publicly funded agencies.
Depth of need related to the impact of risks on health and development and the amount of state support required (Cunningham & McCollam, 2001).
These social policy definitions have utility for government and public systems, as evidenced in both the U.K. government's "troubled families" initiative (Day et al., 2016) for those out of work or school and antisocial over an extended period of time, and the U.K. Big Lottery Fund's programme for people with multiple and complex needs (homelessness, reoffending, problematic substance misuse, and mental ill-health) (Moreton et al., 2016). SANDU | 1471 3 | AN EPIDEMIOLOGICAL APPROACH TO DEFINING SEVERE AND MULTIPLE DISADVANTAGE Epidemiology focuses more on the risks to health and development and less on system responses. Hobbs et al.
(2019) surveyed approximately 100,000 children and young people from 23 English and Scottish local authorities to measure key developmental outcomes and a wide range of risks for these outcomes. They found that 25% of their sample had significant needs. Porche et al. (2017) drew on data from three U.S. national surveys on the health and well-being of children and adolescents and concluded that 10% of U.S. youth experienced three or more adverse life experiences in adolescence; these included: economic hardship, domestic or neighbourhood violence, parental incarceration, parent death, family member mental illness, or family member substance abuse. Other studies report that 10%-48% (conceptual and methodological differences explain most of the variation) of adolescents have multiple adverse childhood experiences (McLaughlin et al., 2013;Saunders & Adams, 2014). Finkelhor et al. (2007) referred to this group as polyvictims.

| LIMITATIONS OF THE CURRENT APPROACHES TO DEFINITION
The proportion of the population exposed to multiple risks far exceeds the proportion getting specialist government-sponsored help. Hobbs et al. (2019) found only a quarter of young people in greatest need were supported by social care, mental health, or youth justice systems. This is an empirical representation of Hart's (1971) law that high disadvantage is inversely related to public systems response. Hobbs et al. (2019) also measured civil society support, finding that most young in great need were supported by friends, family, and community activists. A small proportion of young people (5%) experienced significantly impaired development and were also disconnected from both public systems and civil society support.
Disconnection from support is an overlooked dimension of severe and multiple disadvantage. People in great need who do not seek or get support tend to fall into two categories. One focuses on resilience, showing that many people exposed to multiple risks and some whose health or development is impaired live relatively uneventful lives (Fergusson & Lynskey, 1996;Garmezy, 1993;Luthar, 2003;Rutter, 1985). The second deals with the inefficiency of public systems (Rosengard et al., 2007), meaning that a high proportion of people exposed to and succumbing to risk do not get the help they need from public systems.
This literature does not explain why people in great need disconnect from both formal and informal sources of support. Kools (1999) and Samuels and Pryce (2008) showed that young people facing challenging times build effective strategies to deal with their situations and cope with their trauma. These strategies draw more on their own resources than assistance from others. Kools (1999) found that adolescents in the foster care system used four strategies: hiding the foster child status, maintaining a defensive posture, distancing from others, and keeping relationships superficial. Young people carefully assess the ability to relate of people who could help them Little, Sandu, & Truesdale (2015). Understanding in more depth the way young people use these strategies can help improve the response of practitioners.
Adversity undermines young people's sense of self. Kools (1999) found that the veneer of self-reliance masked a negative view of self, leading to social detachment. Emotions have been key in the process of social connection, particularly in the development of close bonds with caregivers who can provide a "safe haven" and a "secure base" and serving as barometers of attachment security (Bowlby 1969(Bowlby /1982(Bowlby , 1973. Depending on attachment style, individuals will differ in how they feel in their close relationships and in their ability to regulate these emotions (Ainsworth et al., 1978). Given the deficiencies in family support characteristic of vulnerable populations (Fergusson et al., 1994;Osgood et al., 2010), young people facing severe and multiple disadvantage present an unexplored opportunity to understand the role of emotions resulted from past relational experiences in the future process of social disconnection. Moreover, self-conscious emotions such as shame, defined as a negative view of self as compared to own and societal standards (Tangney & Dearing, 2003), have been shown to be experienced by disadvantaged populations, for example in the context of poverty (Chase & Walker, 2013), but more research is needed to understand how they influence the trajectories of young people faced with severe and multiple disadvantage.

| ADDITIONAL CONSIDERATIONS
Despite the well-evidenced challenges of offering effective support for this population-partly explained by a shortage of evidence-based programmes to tackle severe and multiple disadvantage and public systems' difficulty managing needs that involve multiple agencies-research is accumulating on the potential of relationships with professional helpers to alter trajectories of young people facing severe and multiple disadvantage (Author, 2019;Parr, 2016;Virat & Dubreil, 2020,). The family-like nature of these relationships (Author, 2019)  The sample included 30 young people and 35 workers from 11 U.K. and five U.S. organisations. The young people were aged between 16 and 25; had experienced a significant risk to health and development, including abuse or neglect, poverty, and substance misuse; had manifested significant impairment to health and development, for example, challenging behaviours, and mental ill-health; had been supported by the organisation in the study for at least 6 months, leading to perceived positive change relative to their situation. The organisational leaders viewed the positive relationship between a worker and the young person as contributing to this change.
The 35 workers were described in diverse ways, such as therapists, case workers, youth workers, and volunteers. Their supervisors and peers considered them to be exemplary in relating to young people, as demonstrated across many cases.

| Procedure
The present research is part of a comprehensive qualitative study into the potential of professional helping relationships to alter trajectories of young people facing significant disadvantage (Author, 2019). A purposive recruitment strategy was used, selecting a sample based on particular characteristics that allow for a rich SANDU | 1473 exploration of particular phenomena. Participants were recruited through the author's professional networks, to identify U.S. and U.K. voluntary, non-profit organisations that supported young people. Leaders of each organisation were asked to nominate young people and workers to consider the consent procedure.
Interviews of between 45 min and 1 hour undertaken in a place where young people felt safe, usually on the premises of the support organisations, were voice recorded. The author conducted all the U.K. interviews and helped conduct the second set of interviews undertaken by the U.S. research team. Participants received gift cards (£20 in the United Kingdom and $50 in the United States). The workers did not accompany the young people in the interviews but were available if needed (Tables 1 and 2 Female 12 Reported housing difficulties 9 Reported mental health disorders 8 Reported emotional regulation difficulties 13 Reported self-harm (three attempts to commit suicide) 5 Reported substance misuse 5 Reported difficulties with education 11 Reported loss of family members (two murders) 5 Reported involvement with social care system 7 Reported involvement with criminal justice system 8

Social network
Social support Perception of availability of someone to rely on Yes 5 No 20 Composition Positive ties Absence 7 Limited (one person) 9 Two or more people 9 Negative ties 14 Note: The characteristics above were mentioned freely and were not elicited by the researcher during the interviews. Given the young people's support from agencies that deal with populations with severe and multiple disadvantage, the actual number of risks they faced is likely to have been higher.
were undertaken, one in each country, to assess the feasibility of the instruments. Minor changes such as the ordering of questions were identified and implemented (see Tables 3 and 4 for final interviewing templates).

| Analysis strategy
Two out of the three main epistemological positions described by Willig (2013) informed a critical realist approach to the data. The strategy sought to understand and interpret the factors that led the young people into disadvantage and to understand their behaviour, thinking and motives. The views of the workers provided new perspectives and tests of young people's views. The phenomenological elements were drawn out by repeatedly listening to participants' experiences.
The data were analysed using Braun and Clarke's (2006) six stage thematic approach.
The coding started with a list of pre-defined codes with new codes added as required. Semantic as well as latent themes were coded. Themes were selected based on frequency and relevance to the research question. The analysis was conducted using the qualitative software NVivo11.
Trust in the findings (Lincoln & Guba, 1985) involved several steps. Interviewers debriefed after each visit to discuss emerging perspectives and potential biases. Triangulation was achieved both at the data source level, young people and workers, and at the analyst level. Agreement between the author and an independent coder was tested on a third of the youth sample: Cohen's κ = 0.76. The author kept a reflexive journal of the coding process, summarising case studies, changes to the analysis strategy, and reoccurrence of and connections between themes.
Three further checks for variation in findings were conducted. First, a deviant case analysis, where a case that could challenge the emergent model was compared with a typical case, was conducted (Madill et al., 2000). This revealed differences attributable to sample selection (e.g., young people who had recently been matched to their workers as opposed to those who had more than 6 months of engagement), problem definition (e.g., despite using the same label of "severe and multiple disadvantage," support organisations differed in its conceptualisation), or F I G U R E 1 Characteristics of young people facing severe and multiple disadvantage as perceived by themselves and the workers who supported them SANDU | 1475 young people's cognitive abilities (e.g., some young people had difficulty expressing themselves during the interviews, e.g., their answers were short, repeated words, and used often "don't know"). Second, the sample was divided into subgroups of those for whom the data indicated reasonable or low-quality relationships before being supported by the worker. The analysis strategy was then repeated, revealing the study themes to have the best fit for young people with a history of major relationship breakdown, evident in the perception of having no one to rely on, challenges related to their main caregivers (e.g., heavy drug use of parents), and sustained engagement with public systems. Reported years of experience of supporting people with multiple and complex needs 1-5 12 5-10 10 Over 10 8 a Not all workers spoke about their professional backgrounds in terms of their professional degrees.
b The diversity in worker labels reflects the fact that their support-providing role is not clearly defined in this context.

| FINDINGS
The primary findings are summarised in a thematic map (see Figure 1), covering the following themes: • Poor relationships, past and current.
• The role of shame in the way young people think and feel about themselves.
• The acute assessment of those offering help.
• Perceptions of self as unworthy and unable.

| Bad relationships, past and current
Whereas relationships are a source for support for most people (Cohen, 2004) Loss of relationships was a recurrent experience for the young people in the sample. Bereavement was frequent and the causes of death-drugs, alcohol, and crime-were atypical. In other instances, young people actively broke personal ties, escaping a violent father for example or giving up on a drug-abusing mother. A few of the young people interviewed were parents themselves and had lost their children to foster care because of their chaotic life styles: T A B L E 4 The structure of the interview protocol for workers

Sections interview Topic
Part I: Introductory chat Workers identify a young person with whom they formed a positive relationship and helped to improve their situation Over half (57%) of the young people had received extended support from public systems, a few for their entire lives. For some, professional supports and substitute carers had been their close ones, and quality of relationships was variable. The young people described being let down by people charged with their care and not being treated as individuals: Y001: I grew up in the system, and I don't like social workers. They tell you some stuff, and then lie.

| The role of shame
Extended exposure to dysfunctional, unreliable, or unresponsive relationships influenced young people's sense of self, especially a sense of shame, of distress produced by understanding past behaviours to have been unworthy.
Words such as useless, worthless, disgusting, not amounting to anything, broken, second-class citizen, not worthy of others, and not good enough were frequent in young people's self-description. The language used applied to the entire self, extending beyond specific actions, such as crime: The shame was deep as well as broad. It was rooted in a rational assessment that someone with similar experience would feel the same. As such, the shame was hard to shake off:

| An acute assessment of others
Disadvantage encourages people and agencies to offer help. Teachers support pupils beyond their formal educational and pastoral responsibilities. A young person kicked out of school is directed to a variety of professional supports. Homeless young people take their chances with a mostly empathetic general public. How did these young people view help?
They were sceptical. The recurrent negative relationships described appeared to lead the young people to doubt the motives and honesty of others. They were suspicious. They knew they could be hurt, let down, and disappointed. They were prepared to protect themselves from people who help. Their primary assessment was "how will this person make me feel?" This acute assessment of others is not a routine part of human interaction. It comes to the fore in contexts that require an assessment of the need for fight-or-flight. For the young people in this sample, however, the vigilance was continual. Accordingly, the assessments were visceral not rational assessments, relying more on nonverbal cues than answers to questions: W001: It was just the connection-the way she [the worker] was talking to me, the tone of her voice, and the way she looked at me in my eyes. It just showed she's really there paying attention to whatever I'm saying and to the reasons why I'm upset. (Young person, United States) Any worker wanting to form a relationship with a young person had to pass the assessment. A good first impression was not sufficient for the young people. It was followed by a rigorous testing of whether the worker cared, understood, and would provide practical help. Workers talked of "being watched, being analysed, and psycho-analysed": X001: I could tell it wasn't really just her trying to do her job. I could tell it was really her caring because anybody else, they could've just been like, just go to jail. Just go back to jail. You feel me?
We don't care about your life right now, but she wouldn't. (Young person, United States) Young people searched for acceptance, for people who would take them on their own terms. "Authentic" was a common descriptor, someone would be who they said they were, even if that fell short of what was required.
"Genuine" was another word frequently used by young people and their helpers: T001: But he seemed, out of all the staff, when I first started coming here…the most genuine-the most that actually like cared about the kids and stuff in the centre, you know. (Young person, United States) H412: I think you have to be yourself, always. I think women or people who've been through stuff they can pick up bullshit very, very quickly, so if you're presenting as something that you're not, then they won't… You're not authentic in the work and how can you support them if you're not authentic in yourself. (Worker, United Kingdom)

| Backing away
In addition to a range of behaviours consistent with their mental ill-health, addiction, and educational and employment disadvantage that often set them apart from mainstream society, their sense of shame and recurrent experience of being let down by others led them to further back away, including from people who could offer them support.
Young people described being overwhelmed by their experiences and disadvantage. They found it difficult to understand and process feelings. To compensate, they developed protective strategies to separate themselves from their emotions. The transcripts contain recurring phrases such as "shield up," "bottling up," "big wall up," or "shut down": When unleashed, young people were frequently unable to regulate their feelings and became verbally and physically aggressive. Official records referred to them "storming out," "having a short fuse," "banging off," "bouncing off the walls," "being jumpy," and being subject to "blind rage":

| Research implications
This study confirms Samuels and Pryce's (2008) finding that the protective instinct developed by young people when they were consistently under threat is counterproductive. In particular, it supports Kools' (1999) research into the relationship between a negative perception of self and social detachment. Moreover, the findings in this study underscored how the young people's emotional reactions to the risks faced are core to an understanding of their life trajectories. The heavy, self-conscious, and deep-seated nature of shame takes forward the conceptualisations of Tangney and Dearing (2003) by showing its role in forging young people's trajectories and the accumulation of disadvantage. The evidence fits with Lewis (1971), who wrote that we feel guilt over specific actions and shame about the global self.
The evidence from this study provides detailed examples of Schafer et al. (2009) work pointing out that negative self-assessments pushing people facing difficult times away from help and society are also the foundation of their building resilience.

| Theory implications
The findings have implications for how attachment theory applies to a context of severe and multiple disadvantage.
The lives of young people in this sample was underpinned by poor relationships. Their mistrustful, protective pattern of connection to others suggest a combination of avoidant and preoccupied attachment tendencies. However, the young people also used their template of relating when describing their relationships with their workers. Whereas typically the process of filtering new data operates outside of conscious awareness (Griffith, 2004), so making it less amenable to change and adaptation, in this study the young people used what they had learned from their previous relationships and analytically engaged with what new people could offer to them.
This study has some implications for how we understand identity development by expanding the narrative of apparent self-reliance proposed by Samuels and Pryce (2008) and Ungar (2001) to focus on self-worth and competence.
Charting processes that reinforce positive or break up negative identities would improve support for this population.

| Limitations of the study
The study has several limitations. First, because the sample comprised young people and workers perceived to enjoy strong relationships, the findings cannot be generalised to all young people facing severe and multiple disadvantage. Second, and in a similar vein, the young people were supported by organisations known to the author's social network and considered to be exemplary in their practice. Nothing can be said about the variation in practice or young people not supported by agencies. Third, because data on young people's backgrounds were collected retrospectively, only limited conclusions can be drawn about their life trajectories. Fourth, the data were collected using a single instrument applied in two countries, limiting the potential to gain deeper insights by adapting protocols in light of emerging findings such as the role of cognition in the way young people define their situations. Fifth, with the benefit of hindsight, more data could have been collected about the organisational context in which the workers operated and where support was provided to young people. Sixth, the model presented in this study adopts a realist approach to causality, as discussed by Maxwell (2004) and defined as real mechanisms and processes that are in principle observable but not necessarily regularities. Although the study adopted several strategies that could afford causal explanations such as long-term engagement with this population, rich data, a detective style for data analysis, and checking for alternative explanations, further validation of the model, possibly using a quantitative data set of a similar population, is recommended to firm up these findings.

| Policy and practice implications
The emotional reaction to risk and its impact on sense of self point to the role of connection in the recovery process. At present, most public policy is focused on interventions to reduce a risk or ameliorate defined SANDU | 1483 conditions. Most of the young people in this study had previously been exposed to a range of treatment interventions which they and their workers discounted as important to recovery.
This leads to several opportunities for practice development. How does the absence of early nurture, love, and safety and later disconnection influence resilience in protecting against multiple disadvantage and in the recovery process? As Maibom (2010) noted, shame requires an audience; another route into this could be to find effective ways of breaking down shame, for example by connecting young people to relationships that unshackle them from negative emotions and help them to recover their abilities and worth to society.
This study put a spotlight on the relational capabilities of workers. Young people closely analysed the intervention techniques and personal qualities of workers looking for what in previous studies (2015) was described as three "H"s, of: head, meaning the ability to focus on shared goals, not organisational objectives; heart, referring to a deep sense of caring; and hands, concerning the practical help provided to the young person. More could be done to identify workers who have these capabilities.
Finally, the findings prompt reflection about the types of outcomes society seeks for young people facing severe and multiple disadvantage. Public policy is focused on the reduction of psychological symptoms, educational disadvantage, and addictions. Without a sense of worth or competence it is hard to predict the success of any intervention in reducing psychological disorders, educational disadvantage, or their outcome.

ACKNOWLEDGEMENTS
This study was part of a doctoral dissertation and was supported by the author's previous employer, Dartington Social Research Unit. The author is grateful to the team at the Centre for the Study of Social Policy for securing access to the U.S. study organisations, data collection, and for participating in early data debriefs. The author also thanks B. Clarke and C. Sambo for technical assistance.

PEER REVIEW
The peer review history for this article is available at https://publons.com/publon/10.1002/jcop.22572

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.