Executive Summary

Introduction
The Learning in and for Interagency Working (LIW) research project is one of twelve studies that comprise Phase III of the Economic and Social Research Council’s Teaching and Learning Research Programme.  The LIW project was designed in the policy climate that produced the Every Child Matters Green Paper (DfES, 2003) and 2004’s Children Bill.  Current policy initiatives such as these, which address the needs of children, young people and families identified as being at risk of social exclusion, call for ‘joined up’ responses from professionals.  These responses need to be flexible and require practitioners to be able work together to support clients.  In this context the LIW project is concerned with the learning of professionals who are engaged in the creation of new forms of practice that to meet complex and diverse client needs.  The research team will study professional learning in services that aim to promote social inclusion through interagency working.  The aim of LIW is to develop a model of work-based professional learning that will transform interagency collaboration among practitioners working across education, health, mental health, social services and criminal justice. 

Aims of the review
This literature review comprises a review of research on interagency and cross-professional collaboration aimed at enhancing the capabilities of clients.  It pays particular attention to analyses of interagency working that are informed by activity theory and which offer object-orientated analyses of complex, radically distributed work settings.  The importance of activity theory to our research, therefore, is that it offers a conceptual framework for analysing forms of interagency working in which children and families work with frequently changing combinations of professionals from diverse services over extended periods of time.  These professionals may be unused to learning to collaborate with workers from different services.  In addition, to negotiating new professional practices with each other, professionals engaged in interagency working may also find themselves working in settings in which client participation is of key importance and children and families are expected to collaborate in the development of service patterns.

Scope of the review
In general the reviewed literature reports descriptive, single (or comparative) case studies.  These include small-scale, local studies that employed generic evaluation methods but also a series of intervention studies informed by activity theory and employing developmental work research methodology.  The reviewed literature covers four conceptual categories: literature drawing directly upon activity theory; literature informed by other theoretical approaches (particularly organisational/ bureaucratic theory); narrative or evaluative papers which are largely atheoretical; strategic or policy documents which propose models of ‘good practice’ in interagency working.

Interagency working: the context
Current UK government policy has given priority to tackling social exclusion: that is, the loss of access to life chances that connect individuals to the mainstream of social participation.  Social exclusion can occur when individuals or communities suffer from combinations of linked problems such as unemployment, poor skills, low incomes, poor housing, high crime, bad health and family breakdown (Social Exclusion Unit, 2000).  Government guidance since 1997 has exhorted traditionally separate agencies to work together in order to counter social exclusion and to develop public services that are organised to meet the needs of citizens, rather than the convenience of providers.  ‘Joined-up’ welfare services have, therefore, been characterised as the driver of social inclusion.
 
Present policy enthusiasm for developing ‘joined-up solutions to joined up problems’ has generated a plethora of terminology to describe the collaborative approaches required: ‘interagency’, ‘multiagency’, ‘inter-professional’, ‘inter-sectoral’, and ‘partnership’ being prevalent (Lloyd et al, 2001).  Moreover, portmanteau terms such as ‘interagency’ and ‘multiagency’ may be used to imply a range of structures, approaches and rationales.  The literature reviewed herein is derived from studies of diverse models of ‘interagency’ or ‘multiagency’ working.  For this reason, the review is not concerned with prescribing an exhaustive definition of the term ‘interagency working’.  However, Lloyd et al (2001) offer useful, albeit tentative, definitions that loosely encompass most of the structures and practices described in current literature.  These working definitions include:
Interagency working: more than one agency working together in a planned and formal way, rather than simply through informal networking (although the latter may support and develop the former).  This can be at strategic or operational level.
Multiagency working:  more than one agency working with a client but not necessarily jointly.  Multiagency working may be prompted by joint planning or simply be a form of replication, resulting from a lack of proper interagency co-ordination.  As with interagency operation, it may be concurrent or sequential.  In actuality, the terms ‘interagency’ and ‘multiagency’ (in its planned sense) are often used interchangeably.
Joined-up working, policy or thinking refers to deliberately conceptualised and co-ordinated planning, which takes account of multiple policies and varying agency practices.  This has become a totem in current UK social policy.

Interagency working as co-configuration
The development of coherent models of interagency working is dependent upon systematic analysis of new forms of professional practice, framed by understanding of the historically changing character of organisational work and user engagement.  The LIW project’s analysis of interagency working will draw directly upon current developments in activity theory, which focus specifically upon the transitions and reorganisations within work settings that draw together multiple agencies (e.g. Engeström, 1999, 2004; Puonti, 2004).  The form of work currently emerging in complex, multi-professional settings might be characterised as co-configuration: a form of work orientated towards the production of intelligent, adaptive services, wherein ongoing customisation of services is achieved through dynamic, reciprocal relationships between providers and clients (cf. Victor and Boynton, 1998).  The definition of co-configuration is comparable with emerging forms of social provision in which a range of agencies and otherwise loosely connected professionals are required to collaborate with young people and their families to develop forms of support over extended periods of time.

Importantly, co-configuration is a participatory model, in which ‘interagency’ relationships include clients as well as professionals.  Co-configuration is also characterised by distributed expertise and by shifts away from compact teams or professional networks.  In short, professionals working with particular families may not share a common professional background or values, or share a common physical location and may meet quite fleetingly in a variety of configurations.  This distributed form of work has encouraged a shift away from team working to knotworking: a rapidly changing, partially improvised collaborations of performance between otherwise loosely connected professionals.  The reviewed literature suggests that, within UK social provision, many agencies are operating on the cusp between the new co-configuration and longer established work forms.  This is apparent in tensions between strategic and operational practice, in ambivalent attitudes towards distributed expertise and in anxieties over non-consensual practices.

Object-orientated analyses of interagency working
Activity theory literature emphasises the importance of focusing on the object of the activity system in collaborative, distributed work settings.  In other words, its principal concern is with identifying what professionals are working on and their perceptions of the ends that are to be achieved. The object serves as a centring and integrating device in complex, multi-voiced settings; it becomes a way of conceptually framing diffuse professional groups, individual agents and complex practices and services.  However, specific tools for collaborative, interagency practice are lacking at an operational level.  Current developments in activity theory are concerned with producing conceptual tools to enable understanding of dialogue, multiple perspectives and networks of interacting activity systems.  Central to activity theory’s analysis of learning in practice is the notion of expansive learning among both professionals and service users.  Expansive learning can be defined as the capacity to re-interpret and expand the definition of the object of activity.  By rethinking their goals and activities and their relationships with other service providers and clients, professionals may begin to respond in enriched ways, thus producing new patterns of activity, which expand understanding and change practice.

Thus object-orientated analyses of interagency working are ‘post-bureaucratic’, in that they move beyond simply offering systemic prescriptions for managing collaboration and also avoid focusing exclusively on actors and their discursive interactions at the expense of focusing on object formation.  This approach is pertinent to the radically distributed forms of ‘joined up’ working intended to counter social exclusion, wherein clients may encounter multiple agencies and individual practitioners over extended periods.  In much of the reviewed literature current shifts towards radically distributed work and expertise are under-acknowledged.  The increasing tendency for professionals to work in loose, constantly shifting configurations is often depicted as a ‘barrier’ to effective interagency working, rather than a shift to a new, expansive form of work.  It is still often implied that the conflicts generated by interagency working must be denied and that the ideal work form is conventional team working, wherein professional expertise coalesces into tight, consensual communities of practice.

The emphasis placed upon consensual models of working in strategic and good practice may place constraints on expansive learning in practice and, in particular, may tend to under-acknowledge the importance of the internal tensions generated by activity systems as mechanisms for transforming practice.  Consequently, Engeström et al (1997) stress the importance of developing tools ‘for disagreement’: ways of working that allow practitioners to capitalise upon inter-professional tensions and tensions between providers and clients.  Existing practices, designed for single service provision may not suffice.  Engeström emphasises the special importance of ‘future-orientated tools’: practices and instruments that do not merely address immediate working needs but which suggest means by which to expand learning and practice, so as to encourage continual innovation.

Bureaucratic analyses of interagency working
Analyses of interagency working that are rooted in activity theory define organisational learning as extending beyond the formation of organisational forms, rules, procedures, conventions and strategies).  However, outside of the activity theory derived literature, organisational routines and forms remain the key research focus and there is little explicit emphasis upon tool creation or upon object-orientated analyses.  In the reviewed literature, conceptions of interagency working are often truncated because ‘joined up’ working tends to be equated with systemic reconfiguration and ‘partnership’ processes.  Moreover, strategic and good practice literature tends implicitly to propose interagency collaboration as a progressively linear ‘solution’ to social exclusion.  By contrast, the position of the LIW project that interagency working is a learning process marked by tensions and contradictions, rather than an ‘ideal’ model of service delivery.

Knotworking and boundary-crossing
The demands of interagency working exceed current conceptualisation of work-related learning, in that standard concepts of learning in practice still often rely upon conventional notions of partnerships, teams, networks and communities of practice.  In interagency/ co-configuration settings the emergent form of work is characterised by knotworking, which is intensely collaborative activity but relies upon constantly changing combinations of people coalescing to undertake tasks of relatively brief duration.  By utilising developmental work research methods, Engeström et al (1999) have explored the extent to which it might be possible to facilitate knotworking at a more formal level, by introducing rules and tools explicitly designed to structure knotworking interactions.  In short, his work raises the question of whether professionals can be trained to knotwork.

The notion of boundary-crossing offers a means of conceptualising the ways in which collaboration between workers from different professional backgrounds might generate new professional practices.  Standard notions of professional expertise imply a vertical model, in which practitioners develop in competence over time as they acquire new levels of professional knowledge, graduating ‘upwards’ level by level in their own specialism.  By contrast, boundary-crossing suggests that expertise is also developed when practitioners collaborate horizontally across sectors.  Where practitioners from diverse professional cultures, such as education, mental health or youth offending teams, are engaged in shared activities, their professional learning is expanded as they negotiate working practices that cross traditional professional boundaries.  In short, the working practices required to support ‘at risk’ young people and families are not the discrete province of any one profession but require planned configurations of complementary expertise drawn from across education, health and social services.  Whereas standard professional role theories tend to focus on anxieties over professional barriers, Engeström’s (1999, 2001a,b) notions of boundary crossing suggest that new developments in learning for interagency working should focus upon the potential spaces for renegotiation of professional practice that are opened up when workers from traditionally separate sectors begin collaborating.  A related debate within the reviewed literature is whether moves towards interagency working will encourage professionals from diverse sectors to become adept at operating within the discursive practices of colleagues from other backgrounds or whether more fundamental reconfigurations of professional practice might lead to the emergence of hybrid professional types.

In activity theory ‘boundary objects’ are the focal points for analysing and understanding boundary-crossing practices.  Boundary objects may take the form of physical objects or, alternatively, pieces of information, conversations, goals or rules.  These become ‘boundary’ objects when they are worked upon simultaneously by diverse sets of actors.  For example, a child’s care plan may be negotiated by a nexus comprising teachers, social workers, health workers and educational psychologists.  In such a situation the care plan assumes particular importance in the learning of these diverse professionals because it sits at the intersection between different professional practices or cultures.  It can be used differently by the corresponding communities, providing a means to think and talk about an idea in multi-voiced fashion, without the necessity of any one community completely adopting the perspective of the other.  A boundary object provides a mechanism for meanings to be shared and constructed across professional boundaries (and across boundaries between professionals and clients).   Thus boundary objects provide key moments of meaning-creation, renewing learning through collaboration.

 

Conclusion
The reviewed literature suggests that conceptualisation of interagency working to counter social exclusion is under-developed, given the complex demands placed upon providers and clients in the post-Green paper context.  In particular, both the learning processes that take place within interagency settings and the learning processes that might form a prerequisite to effective interagency collaboration remain under-explored.   In the current policy context the prevalence of policy and strategic literature that emphasises good practice models is unsurprising but tends to perpetuate the notion of interagency working as a virtuous solution to ‘joined up’ social problems and to under-acknowledge interagency working as a site of tensions and contradictions, rather than an ideal model of service delivery.  In addition standard analyses of interagency practice too often equate interagency developments with ‘partnership’ tools and with systemic analyses of collaboration.

Strategic literature and good practice models offer little in the way of conceptual tools to enable understanding of dialogue, multiple perspectives and networks of interacting activity systems.  Outside of the activity theory derived literature, organisational routines and forms remain the key research focus and there is little explicit emphasis upon tool creation or upon object-orientated analyses.  The development of coherent models of interagency working is dependent upon systematic analysis of new forms of professional practice, framed by understanding of the historically changing character of organisational work and user engagement.  With regard to emerging practices around interagency working to counter social exclusion, there is a pressing need to identify and conceptualise the key features of learning and practice in work settings in which a range of agencies and otherwise loosely connected professionals are required to collaborate with young people and their families to innovate and develop forms of provision over extended periods of time.