Is Torie a "bad" child? Is he "unteachable"? He certainly started his school career without some of the basic skills for learning, highlighted in his disregard for authority, the difficulty he has managing his emotions, and his inability to draw on a language for negotiation and conciliation. His lack of development in these and other areas is at odds with the school's expectations - he is simply not prepared for success in this system. Nor is he motivated to achieve goals set by a system that he does not consider relevant. There are, to a greater or lesser extent, children like Torie in every class in every school throughout Australia. But it is at schools located in communities where social and economic stress is most entrenched that the mismatch is most obvious and where the wellbeing of the whole school is most affected. How are these problems to be addressed systematically and what will guide our search for practical and effective solutions?

The purpose of this paper is to discuss the ways in which the Pathways to Prevention project has been attempting to develop a model for integrated support for vulnerable children and their families. The Circles of Care program was set up within the Pathways project to focus on the ways in which families, schools and community agencies work together. Although better outcomes for vulnerable children are the bottom line, the real "client" for the Circles program is a poorly functioning developmental system. By "developmental system", we mean in this context not just the key developmental settings of family, school, local community and community agencies, but more importantly the various relations or connections between these settings that promote (or more likely inhibit) the ability of the disparate elements to operate in a harmonious and mutually supportive fashion.

In simpler terms, the fundamental goal of the Circles program is to learn more about how to forge relationships in order to support children whose development may be at risk. The aim is to gain greater insight into how to reverse the commonly experienced service provision scenario where independently operating professionals (including teachers and principals) make separate decisions (often also independent of parents) about what a child needs. By contrast, Circles of Care seeks to create conditions that actively engage the child, their parents and wider family, and a key group of professionals in a joint process of appraisal and decision-making. These conditions for genuine participation and collaboration do not automatically exist within the current organisational structures of many child- and family-serving systems (Keast, Brown, & Mandell, 2007). Our experiences in Circles of Care and the wider Pathways to Prevention project demonstrate that even the capacity for cooperative practice, let alone the holistic system integration we envisage, should not be taken for granted.

The need for a fresh approach

Social scientists and epidemiologists have observed that while the latter half of the 20th century ushered in changes that increased the general quality of life for many in our society, these same changes also contributed to what Bronfenbrenner (2005) referred to as a kind of disarray and chaos that threatens the social and ethical development of our children, particularly those growing up outside the mainstream in communities that have been marginalised by social change. This is seen in growing rates of problems such as disruptive behaviour, delinquency, depression, suicide, substance use, lack of connection and poor school performance, all of which have been reported among the youth of many modern societies (Keating & Hertzman, 1999), including Australia (Stanley, Richardson & Prior, 2005). While not all child indicators have portrayed a uniformly bleak picture (e.g., Smart & Sanson, 2008), a recent comprehensive report card by the Australian Research Alliance for Children and Youth (ARACY) has indicated that Australia performs weakly across a wide range of domains, including health and safety, family/peer relationships, and Aboriginal wellbeing (ARACY, 2008; Emerson, in press).

Ecological theory (Bronfenbrenner, 1979) and the more comprehensive developmental systems theory (e.g., Lerner & Overton, 2008) can be applied to the study of children's development to explain these negative trends in developmental outcomes. In this context, social change can be seen to have created a range of conditions (like chronic stress, over-stimulation and frenetic lifestyle, family breakdown and disharmony, time poverty and reduced parental participation in children's care) that interferes with so-called "proximal processes", which are critical to positive development. These proximal processes include such conditions as responsive parenting, nurturing environments, and strong supportive relationships both within and between the key settings in which children live: home, school, neighbourhood and cultural community.

In other words, living and working within stressed, disorganised environments limits the degree to which parents and other carers (such as teachers) are able to provide the conditions necessary to support the various dimensions of children's development, including learning and achievement. For example:

• When families are alienated by hardship and overwhelmed by the strain of conditions - such as poverty, work pressures, relationship problems, mental illness, domestic violence, or substance abuse - they may not have the skill, will, support or access to the kind of external resources that enable them to provide for their children's basic physical and emotional needs or to keep them safe from harm, let alone to provide the kind of experiences that will foster the physical, cognitive, linguistic, social and emotional skills their children need in order to succeed in relationships with others, at school and in adult life.
• Schools are often unprepared for the challenge of supporting the learning of children like Torie, who arrive at school with under- or inappropriately developed skills and are largely unprepared for the challenges of the education system. In the Pathways to Prevention project we have observed that many teachers feel that they are at the front line of a battle with the consequences of social disarray and disadvantage. They are faced with students who are too tired or hungry to concentrate; are confused and distressed by the emotional burden of living in insecure, disorganised or even abusive conditions; are unmotivated through a lack of environmental supports for learning; have undeveloped communication skills; lack competence in peaceful interpersonal problem-solving but are skilled in the use of manipulative and aggressive survival techniques, with little apparent regard for the feelings of others; and are distrustful or suspicious of authority figures. Such challenges to teaching and classroom management can create the feeling of being besieged. Surveys of 150 staff at Pathways schools between 2006 and 2008 reveal that student behaviour is a major cause of job-related stress, with approximately 70% of teachers reporting moderate to high levels of stress (of these, 12.5% said they experienced "more stress than I can cope with"). This often sets up an adversarial situation that may include blaming families for student failure. This is nearly always counterproductive because it erodes links between the critical developmental contexts of home and school.

When schools are so distracted by the by-products of "toxic" developmental settings that they expend the bulk of their energy locked in a continuing cycle of reaction to developmental difficulties, teachers can become too exhausted by stress to work proactively to create conditions that support children's learning and promote developmental competence. This can then help break the next link in the chain, as the likelihood of educational failure increases and in turn jeopardises long-term prospects and future quality of life (Comer, 2004; Knudsen, Heckman, Cameron, & Shonkoff, 2006). How then can we ensure that all children are set up for success, now and in the future?

Using developmental systems theory as a guide to intervention

The developmental systems approach to intervention requires that in order to address problems in a child's development we must first understand the context. The value of an analysis of context is that it emphasises the importance of interventions designed to enhance the conditions, settings, interactions, and especially relationships and connections between settings that shape the course of development. Therefore, when considering how to respond to developmental problems like those experienced by Torie, we need to reframe the question and ask: How is the developmental system failing to provide the basic requisites for the positive development of children from low socio-economic population groups, especially those living in the most socially disadvantaged communities?

Thus, within a developmental framework, interventions must be aimed at the level of the processes and connections between organisations and settings that influence developmental outcomes. It is not enough to simply try to modify or "fix" some specific element of a child's behaviour, out of context. Rather, intervention must work to increase the capacity of invested adults to provide attentive, caring and stable environments and to restructure the systems that support children's development so that disparate elements work together and reinforce each other, instead of independently or in opposition to each other.

This is clearly not a narrowly focused approach and certainly not a quick or easy one. One popular strategy is to focus on the service system and to promote better partnerships between agencies. The value of integrated multi-agency approaches is widely acknowledged and is, to some extent, supported by empirical research, including a systematic review of early childhood interventions by Wise, da Silva, Webster, and Sanson (2005) and a meta-analysis by Manning, Homel, and Smith (2010). However, not all the evidence has been positive, with difficulties in collaboration between agencies often emerging as a key theme (e.g., Department for Children, Schools and Families & Department of Health, 2009).

In fact, the ways in which agencies work together can vary widely. Keast et al. (2007) outlined a continuum of community partnership approaches that distinguishes between cooperation, coordination and collaboration. Cooperation is considered the starting point and involves very little effort or loss of autonomy on the part of organisations, while coordination moves to the development of practices that enable organisations to work together while not losing autonomy. By contrast, collaboration requires more intense relationships and practices and "a blurring of the boundaries between agencies" (Keast et al., 2007, p. 19). Coordination lies at the heart of such successful programs as Team Around the Child (TAC), which coordinates support for families of children with disabilities (Limbrick, 2007), and Communities for Children (Edwards et al., 2009), which operates at the community level to facilitate coordination across agencies involved in the provision of local programs that promote children's development.

However, this growing body of evidence about the value of multi-agency partnerships does not fully address the challenges of system change, by which we mean the ways in which the activities and products of these kinds of human service strategies can be harmonised with the cultures and routine practices of enduring, universal, developmentally relevant institutions such as families, schools, and religious organisations. Such system change includes but extends well beyond agency collaboration, and there is little empirical research to guide reformers.

So how might a structure for a more effective developmental system be set up? And what might be involved in the organisation, management and maintenance of a continuum of care that spans key developmental institutions? The Circles of Care model is one attempt among a range of initiatives to address these questions.

The Circles of Care model within the Pathways to Prevention project

The Pathways to Prevention project operates as a comprehensive family support service and prevention project within a cluster of socially disadvantaged suburbs in Brisbane, by way of a long-term partnership that exists between Mission Australia, Education Queensland and local schools, and Griffith University. The project aims to create a pathway to wellbeing for all local children as they transit through successive life phases, from early childhood to adolescence (Homel et al., 2006). The Pathways model is organised around the concept of a developmental pathway, which refers to the way in which sequences of events, experiences and opportunities over time contribute to changes both within and around the child. The work is grounded in the understanding that just as children influence and are influenced by a range of contexts like family and school, so too are these immediate contexts influenced by the broader contexts within which they are nested (like cultural and geographic communities). In their turn, these broader contexts are nested within political and economic systems.

This "Russian doll" view of the multi-layered influence of social ecology is incomplete and to some extent misleading, but is productive in as far as it highlights the need to work across multiple contexts, and enhance all the environments that influence development, in order to create the possibility of better outcomes for children. However, we have learned that it is critical to ensure that work across contexts does not inadvertently foster a fragmented approach by artificially compartmentalising intervention efforts into separate spheres. Circles of Care was designed therefore as a "connecting mechanism" that would create a process for drawing programs and people together in a way that enables a harmonious, interdependent support system to be mobilised around children like Torie who are experiencing difficulties.

Circles of Care is based on a team approach where a small network of caring adults is created for each participating child. Although envisaged as a process that could just as easily be applied universally to support whole classes of children, our initial trial of the Circles of Care model has operated in a targeted way to build a continuum of support around selected children. In the work that has been undertaken to date, most children have been referred to Circles of Care by schools. A smaller number have been referred by government agencies and, in a few cases, through parents' requests. Most of the current work in Circles of Care supports children with medium to high needs, with Torie presenting only one example of the array of the issues that is typically dealt with. Some of these issues have included overly aggressive behaviour, withdrawn behaviour, medical difficulties and isolation linked to language or cultural differences. Although standard practice has been for one child to be the focus of the referral, it is usual for all children within the family - and often parents or carers as well - to receive support from the Circle, which approaches each child as a part of an often complicated family system.

The Circles of Care program is overseen by a full-time coordinator (a social worker from the Pathways family support program) and facilitated through the work of three part-time community workers from the Pathways team. This small group works with seven participating schools. To date, Circles of Care have been created for approximately 50 children and their families. Each Circle is a team of educators, family support service staff, family and community members who commit to work together in support of the child at the centre of the Circle. The core inner Circle may be supplemented by people from a wider Circle who can be called upon to offer specialist support. The idea is to create supportive environments by purposefully focusing effort on the task of building trust, relationships and connections across the systems and contexts that are critical to children: most notably their families, schools and cultural communities. The key task of the Circle is for members to work together to identify goals for the child's positive development and then to plan and set in motion a set of strategies for achieving those goals.

One essential element of the operation of Circles of Care is the way in which Circle members come together on a regular basis (at least twice a year) for joint meetings. Although these meetings are informal, they follow a structured plan and are facilitated by the Circles of Care coordinator or another member of staff. It is during these meetings that plans for supporting the child are collectively developed, reviewed and adjusted. The meetings provide an opportunity for all members to come together on an equal footing to share their unique information and insights about aspects of the child's development, which helps to create a more complete understanding of the situation. After Circle members have expressed their hopes and concerns for the child, a number of goals are identified and a broad plan for jointly working towards those goals is set out. Circle members agree to complete tasks that will advance the plan of action prior to the next meeting.

Ups and downs: The challenge of system change

Comments by Circle members highlight the way in which Circle meetings have in some cases promoted greater understanding between the school and home, and have influenced both professional practice and how members behave towards each other and the child. For example:

Well it gives you background knowledge - and we need background knowledge of those children to help them. If we don't know they have a problem at home, well, we cannot understand ... what's happening at the home life and what problems they are bringing to school. But if we can understand what problems they are having at home, we can then help them deal with social interactions at school, social wellbeing. It is not to do with the learning as such, but how they get on with friends and why they lash out; because while they are lashing out they are not learning, so it is good. (Teacher A)

The Circles of Care model is illustrated by the way in which the Pathways project staff worked to support Torie. As mentioned earlier, Torie's family was overwhelmed by a range of complex issues. The Circle that was established for Torie helped draw together his family, the school, and a range of services to mount a multi-layered response. The Circle coordinator initiated this process by bringing together the children (Torie and his brothers), their parents, their teachers, and a member of the Pathways cultural support team (an elder within the local Indigenous community). This pivotal group began by sharing information from their different perspectives in order to create a much fuller appreciation and shared understanding of the situation. This allowed them to begin setting short-term goals and to work together to take some immediate steps towards achieving them. This included:

• facilitating a referral to a paediatrician, with an ensuing diagnosis of attention deficit hyperactivity disorder (ADHD) and prescription of medication for one of the boys (which calmed some of the more extreme behaviours among the trio);
• negotiating innovative ways for the family and children to become more involved with learning and the school community (e.g., inviting the parents to class to see the children's work; using school activities like boomerang painting in art to help the children explore their cultural heritage; sharing their cultural heritage with other children to develop a sense of pride and an opportunity to restore relationships and connect in a more positive way with others; encouraging the children's sporting abilities and acknowledging the contribution they made to the school via their representation on school and district football teams - one teacher even began supporting the children's local football team and attending weekend fixtures with them);
• having the Pathways Indigenous family support worker provide intensive family assistance in household management (e.g., developing strategies to organise budgets, routines and positive behaviour management practices; helping family members through traumatic grieving processes; advocating with legal and social services; arranging appointments and transport to access basic services; facilitating entry to wider social networks); and
• having the Pathways Indigenous community worker accompany the children at school to troubleshoot and help maintain some sense of equilibrium during periods that had been identified as flashpoints for difficulties (such as returning to school after holidays and visits to family members in prison).

As the operation of Torie's Circle proceeded, the core group began developing strategic relationships beyond the nucleus to pull in an extended range of specialist services and supports as part of the comprehensive response. This included:

• engagement of a local Indigenous counselling service to provide individual therapy for the children;
• linking in to anger management programs and other local community health programs for young people;
• gaining access to youth programs that focused on developing young people's self-reliance, problem-solving and leadership through recreational and outdoor activities; and
• developing relationships with a professional sporting club to co-opt positive role-models and mentors for the children.

As these services were enlisted, the Circle coordinator organised opportunities for representatives from the associated agencies to come together with other Circle members to consider how they could contribute in a coordinated way to the children's welfare and become part of an ongoing cycle of resetting goals and refining action plans as progress was made.

In effect, the Circle of Care was a structure for coordinating a complex array of interventions for a number of family members in order to deal with a range of issues over an extended period of time. It was both a catalyst and facilitator of cross-system synchronisation. However, the task of establishing processes for effective collaboration was daunting, not least because of institutional barriers faced by Circle members, who then had to find ways to overcome the natural resistance to change, and because of the way different perspectives among Circle members initially made it hard for some members to work together and see how each person's work complemented that of others.

One example of the difficulties experienced by the Circles of Care coordinator in bringing the various players together was an initial tendency at schools for staff to view the program as an "add-on" or external resource rather than as something that schools did themselves. An early comment by one principal at a participating school summed this up nicely:

It is an add-on at the moment, so if it became more of the culture and happened more often it would become something that people saw uses for and become one of the repertoire of things that [teachers] use. At the moment it is a one-off for one or two teachers and children in the school. (Principal A)

This comment underlines the difficulty of introducing new practices into established systems, especially when the initiative asks individuals to reshape the way in which they operate and to develop new processes in order to make that happen. The principal's comment that Circles needed to become part of the school's culture was insightful and has helped the program to move forward. Nevertheless, Mission Australia staff and teachers have had to work hard to build strong relationships and guide participants and the wider school community through the Circles process. The full extent to which Circles of Care might become entrenched in enduring school routines remains to be seen.

The inherent difficulties in establishing genuine collaborative partnerships across systems has sometimes created procedural barriers and frustrations, as is evident in the comment made by a participating school principal when speaking about some of the challenges she faced in the teething stages of the implementation:

The concept was great - the process was a disaster ... to have seven [meetings] cancelled at the last minute and no shows - I was over it! ... [Now] communication is fantastic - between Circles and me, Circles and the family - it is proactive and happening. (Principal B)

As the Circles team worked through some of these initial problems, the vital role of working to enhance clear lines of communication between Circle members and, critically, between participating agencies was apparent. The following comment by another school principal highlights the way in which Circles eventually helped bring a school and a family into harmony by facilitating communication:

There is one family where I can think of where we were having difficulties with communicating and getting messages across to the family and they were suspicious and untrusting of the school ... The process enabled trust to be developed so that we could get them to follow through with the things we needed to do with that student at home, and to work with the school rather than working against the school ... He has improved out of sight now and is now integrated into another school and isn't flagged as one of their bigger issues. (Principal A)

In some ways, this comment also points to the inherent difficulties faced by the Circles team when it came to coordinating Circle meetings. Many families (such as Torie's) are wary of attending school, teachers and principals are extremely busy, and community workers are happy to maintain the status quo and continue their normal routine of working with families directly without going out of their way to involve schools. It is natural for individuals to stick to what they know best (Lewin, 1952; Piderit, 2000), however, through persistence and cooperative effort, the Circles team and participating teachers have made steady inroads and are seeing shifts towards more engagement with the process among families, schools and community workers. It must be acknowledged, nevertheless, that the journey is long and detours are not uncommon. Moreover, even when Circles such as Torie's start to work effectively, there are many downs as well as ups in the lives of the children, with many unresolved problems. Nonetheless, while rigorous quantitative evaluation is not yet complete, evidence to hand suggests that the approach overall has value.¹ Examples of some of the improved efficiencies and outcomes are presented below, using Torie's Circle as a case study, but illustrating more general points:

• Circle members found ways to work with people's competing agendas and timetables to make it possible to undertake essential joint planning and reduce duplication;
• the children's behaviour and performance at school improved (despite regular setbacks);
• the children's parents began to approach the school independently (indicating not only that the quality of relationships between home and school had improved, but that they had become empowered to take greater control of their lives); and
• although the children ultimately changed schools, the school developed strategies to deal more effectively when trouble erupted (e.g., a special "dreaming space" was set up as a cultural precinct for Indigenous children to use at their own discretion, as a haven to distance themselves from emotionally charged situations and regain their composure and equilibrium).

Conclusion

The complex nature of problems experienced by some of the children who have participated in the Circles of Care program had represented an almost insurmountable challenge to the various child-serving systems with which they had been involved (e.g., schools, community health, child protection, and family services). In some of these cases, Circles of Care has been able to provide a bridge that has helped to link systems that had previously been working in independent silos. However, creating the conditions to facilitate such cross-system harmonisation has been a formidable challenge.

Our experience has been that, no matter how much good will exists, genuine collaboration across organisations, sectors and disciplines (such as psychology, social work, teaching) is unlikely to occur in a systematic and sustainable way without a coherent plan for getting there. Effort and resources are required to develop the capacity to work together within a framework of collaborative practice and to become part of a fully integrated network of support for children's development. One virtue of the Circles model is that it does provide a clear road map, even if many obstacles are encountered on the journey.

We are also persuaded that reform of the way in which supportive systems are constructed for children and families will not occur without high-level top-down support for the local-level bottom-up effort. Bottom-up initiatives, no matter how well-founded in good theory and proven practice, will always encounter barriers in organisations with which collaboration is sought if their purposes and practices do not fit with those of the organisation. This is particularly evident in schools, where programs like Circles struggle to be seen as more than an add-on. Under these circumstances, participation by principals or teachers depends on their professional commitment alone, having no obvious or direct pay-off in terms of performance assessment or promotion.

A further lesson from our experience so far is that while each member of a Circle of Care may take responsibility for ensuring that specific activities are conducted or supports set in place, it is essential that the group as a whole works together to ensure each element of the coordinated plan operates synergistically as part of the integrated process for promoting the child's welfare. We have found that it takes a lot of time, effort and ingenuity to build and nurture such partnerships and to develop the shared vision, clarity of purpose, trust, commitment and willingness to work together and share responsibility. Relationships are central at every level within Circles, and the process of building relationships between organisations and across systems parallels the process undertaken by social workers as they build trust and relationships with vulnerable families. At all levels, the art of building trust and relationships is a crucial, delicate and complex business.

One of the guiding principles of the Circles of Care model is the empowerment of all participants, most notably parents and children. It has been argued that without changing the way people live, the many programs and public services that are provided to the most disadvantaged families in the most deprived communities have little likelihood of achieving real social change and resolving so-called "wicked" social problems (Schorr, 1998). When they work well, empowerment strategies facilitate cooperation and shared responsibility, where everyone (including families) participates on an equal footing to identify problems and the barriers to positive development, to set goals, plan a way forward, and then set those plans in motion. Families are empowered to choose their own solutions. However, this determined focus on building genuine partnership is probably one of the main reasons why Circles of Care has been experienced as such a challenge. That is, we have learned that setting up the Circles model is difficult because it involves a paradigm shift within participating organisations. This critical shift involves a movement away from operating within a service orientation that is characterised by what Leadbeater (2009) refers to as doing things to or even for people, towards an orientation characterised by a resolve and increased capacity to work with families and schools.

The Circles of Care program is helping us to appreciate the way key developmental contexts like schools can be used as an opportunity to create an integrated community of care for children's development. Such communities of care do not come cheap, but cost comparison analyses carried out as part of the evaluation of the first phase of the Pathways to Prevention project demonstrate that even the most expensive universal or selected interventions are cheaper per "unit of benefit for children" than remedial programs (Manning et al., 2006). Circles of Care is at the more expensive end of the intervention continuum, since it has elements of both selected and indicated programs; that is, it works both with children at risk in some way who are not yet in a remedial program, as well as with some children for whom conventional remedial measures have failed. Nevertheless, despite its intensity, it may be cost efficient in comparison to other approaches and relative to its benefits: the approximate total cost of Circles per participant is $3,000, which is of the same order as the cost of behavioural management programs in schools (Homel et al., 2006). However, preliminary analyses suggest that the Circles of Care program may be more effective.²

This has important implications for government programs like Communities for Children, as service providers are funded to extend their focus from preschool children and their families to those in the primary school years.³ There are also implications for other federal government programs being developed as part of the work of the Council of Australian Governments (COAG), such as the National Partnership Agreement on Low Socio-Economic School Communities (COAG, 2009). Success will depend, as we have noted, on a creative combination of well-funded, top-down strategies that build on comprehensive, well-grounded, evidence-based programs that strengthen relations between core elements of the developmental system, created from the bottom up. The National Partnership Agreement in particular has real promise in providing the framework and the incentives for school principals to become actively engaged in community partnerships in furtherance of their mandated learning improvements.

Within the Pathways project, and across the whole country, we need to work together more effectively to foster the positive development of the vulnerable children that the "system" does not know what to do with. We need to work towards a better understanding of how to create supportive environments by building relationships across organisations and contexts; how to develop organisational practices that facilitate collaborative efforts; and how to evaluate changes in system functioning.