Understanding social capital within community/ government policy networks.

Abstract

There is an increasing need for joined-up government and joint working between agencies and across sectors in order to address "wicked" social problems and improve policy outcomes. Consequently, it has become imperative to understand the nature of policy network settings, and also the endogenous and exogenous factors that positively and negatively impact on interagency and intra-agency collaboration. From senior managers to front-line workers needing to work collaboratively to achieve outcomes, knowing the types of relationships people have with other agencies may assist in identifying where greater time could be spent improving existing relationships or making new ones. While it is easy to assume that more coordination and cooperation will automatically lead to better and more efficient policy outcomes, collaboration can often be resource intensive, slow the whole process down and actually inhibit the achievement of policy outcomes. Therefore, if agencies are going to invest time and money in collaborative efforts, they need some assurance that such investment will pay dividends. The policy implications for using network analysis within a social capital framework to study policy networks and partnerships are discussed in this article. This paper concludes by describing how a policy network/social capital framework approach could be used to examine the Strengthening Families interagency case management process used in New Zealand.

INTRODUCTION

Governments around the world are realising that approaches they have typically employed to address difficult or essentially "wicked" (2) social problems have been too narrow and compartmentalised. Wicked problems present challenges that cannot be handled by dividing them up into simple pieces in near isolation from one another. In addition, there is now more recognition by governments of the linkages between social, human, environmental, cultural and economic capital in terms of outcomes. Consequently, there is now greater understanding that effective policy responses to many social problems such as inequality, unemployment and social exclusion require a cross-sectoral approach, with the government working with the third sector (3) and the community. In other words, what is needed is a "joined-up" or "whole-of-government" approach.

Networks are often the only governance form that is able to deal with today's complex problems that are not mirrored in the established formal (hierarchic) structure of state bureaucracies and geographical boundaries (Milward and Raab 2002:6). Indeed, an interagency, cross-sectoral case-management approach is often the most effective and efficient way to respond to those families faced with multiple, interrelated problems. As Colin James (1998:7) puts it, "a minister of strengthening families does not so much need a Social Welfare Department as selections of analysts and implementers from the courts and police across to the education and health portfolios, beside 'welfare' specialists".

Neo-Liberal Administrative Reforms in New Zealand

New Zealand has its fair share of wicked and complex problems, and what is needed here is more of a joined-up or cross-sectoral approach to policy. However, a legacy of the neo-liberal state sector reforms of the 1980s and 1990s is a public management system that seems ill-equipped to adopt whole-of-government methods of policy formulation, implementation and evaluation. The problems resulting from the neoliberal public management reforms in New Zealand have been well documented in various state sector reviews and publications by academics (Kelsey 1997, Boston 1999, Boston 2000, Cheyne et al. 2000, Brown 2000, Gregory 2003b, Ministry of Social Development 2003).

The State Services Commission's 2001 publication Review of the Centre identifies the issues of fragmentation, the proliferation of agencies and ministerial portfolios, and the consequential loss of focus on the big picture as major problems for coordinated service delivery. As pointed out in this publication, fragmentation means "ministers need to build relationships with multiple agencies, and at times reconcile conflicting agency positions at an excessively detailed level" (State Services Commission 2001:4). In addition to interdepartmental issues, there are, as identified in the recent work of the Community and Voluntary Working Party (2001:9), issues that need to be addressed to improve relationships between government departments and third sector agencies delivering contracted social services.

Nowhere have these ongoing problems of coordination and cooperation been more salient than in the area of social services, often with tragic consequences. The review carried out by the Office of the Commissioner for Children into the death of five-year-old James Whakaruru in 1999 from physical assaults perpetrated by his mother's partner is a case in point. The Commissioner's investigation found that:

   Poor interagency communication characterised the professional work
   with James and his family. Agencies worked without reference to each
   other, and ended their involvement assuming that other parts of the
   system would protect James. (Office of the Commissioner for Children
   2000:1)

Another investigation carried out by the Office of the Commissioner for Children a few years later into the deaths of sisters Saliel and Olympia Aplin (aged 12 and 11 years respectively) in 2001 at the hands of their stepfather also found that lack of interagency collaboration contributed to the tragic outcome. The Commissioner's investigation identified:

   Poor practice similar to that found in the June 2000 investigation
   ... Policies and procedures were in place to protect these children
   but poor practice within and between agencies contributed to
   increased risks to the girls' safety. Many opportunities for
   appropriate interventions were lost because no single agency had
   the whole picture or a complete understanding of the risks present
   in their lives ... Agencies did not meet to discuss their concerns
   and only dealt with the issues confronting their own agency at the
   time. (Office of the Commissioner for Children 2003:1)

In both the Whakaruru and Aplin cases, neither a Strengthening Families meeting nor a Family Group Conference had been called. Had such an interagency meeting taken place with wider whanau members, then all the information held by the respective agencies and whanau members could have been shared and acted upon in a planned and collaborative way.

A solution proposed by the State Services Commission (2001:5) to the problems of fragmentation in its seminal Review of the Centre is to establish "networks of related agencies" and "gradual structural consolidation" of existing agencies. In Review of the Centre, there is considerable emphasis on a whole-of-government approach, coordinated responses involving multiple agencies, and "circuit breaker teams" to develop solutions to particular, seemingly intractable, problems.

Social service agencies in New Zealand, both government and non-government, are, however, working under increasing pressure with stretched resources. Right across the public sector, many officials face heavy workloads that leave little time to devote to collaborative processes. While these may well be more effective in the long term, collaborative responses require an upfront investment of time and resources. The State Services Commission has already identified signs of "coordination-meeting fatigue" and warns that a general push for more collaboration and coordination may further dissipate resources through ill-defined and unfocused activities. (4)

Networking or relationship building between agencies can be critical to their work, and yet is rarely singled out as a separate output or accorded the significance it deserves. In addition, the intrinsic benefits of networking may be difficult to quantify or not obvious in the short term. This brings us back to the question of whether dedicating resources for collaboration and coordination is really worth it. This is a critical issue, particularly for networks and partnerships in the third and public sectors, where funders, policy makers and front-line workers may all have a strong interest in knowing whether the resources invested will pay dividends (Milward and Provan 1998:389).

A Growing Need for Administrative Analysis

If the reason that policies do not achieve their intended outcomes lies in systemic failure, then refining policy instruments may not be the solution, and we need to turn our attention more towards administrative systems. As Milward and Proven (1998:388) point out, "the reason network analysis is so important to public management lies in the observation that we seem to have an overdeveloped capacity for policy analysis and an underdeveloped capacity for administrative analysis". This rings true for the situation in New Zealand, where we have seen a proliferation of fads and trends in policy-analysis techniques that have so often failed to deliver significant improvements in policy outcomes (Gregory 2003a).

Policies often fail due to systemic failure and/or failure on the part of people within those systems to undertake their roles effectively. A typical and widely used response on the part of governments to systemic failure is to re-organise or restructure those systems deemed to be at fault, thus disestablishing existing departments, creating new ones, or merging one or more existing departments. Indeed, in the quest for improved effectiveness and efficiency, we have seen a constant restructuring of government departments in New Zealand since the mid-1980s, seemingly without a pause for reflection of any resultant gains and/or losses, or any comprehensive review of the administrative structures themselves. The constant restructuring of the public sector has in many cases only exacerbated the problems associated with fragmentation. If we accept that fragmentation across the public sector is a contributing factor to policy failure, we need to determine the breadth and depth of this fragmentation and, even more importantly, the reasons for it. It is only by addressing both endemic and specific structural and agency barriers to collaboration and coordination that we can hope to give practical effect to ideals of joined-up government.

Currently in New Zealand, we are in the midst of what State Services Minister Trevor Mallard (2003) has described as a second wave of reform characterised by a concern for cohesiveness, consistency, constructive relationships and shared values across all parts of government. In other words, what we are seeing here in New Zealand is a government trying to forge a new role for itself, away from the market-orientated and statist mechanisms of government towards new forms of social governance. Le Gales (2001:172) defines governance in the political sense as a "process of coordinating actors, social groups and institutions to attain clear goals that are discussed and defined collectively in uncertain environments". Governance in this sense implies a move towards a catalytic role where government is actively engaging societal actors, from both the public and private sectors, in network relationships through which they strive to steer the policy process towards the realisation of shared goals (Wallis and Dollery 2002:1).

The current emphasis on government and community partnerships has partly grown out of the recognition that the government is not able to accomplish socially desired levels of wellbeing on its own. Brown (2000) makes this point in his review of the significant resource problems in the Department of Child, Youth and Family Services. Brown says that "the combined and varied skills within communities need to be enlisted in a well coordinated and consultative manner, if Child Youth and Family is to succeed in reducing its unallocated cases and delivering a safe and satisfactory service to children, young people and their families" (Brown 2000:88).

Partnership has now penetrated the lexicon of state policy as part of governmental reconfigurations towards new forms of local social governance that aim to be citizen-focused, relationship-based and collaborative. Partnerships are virtually a requisite for community-based initiatives that seek holistic and multi-sectoral solutions. These points are illustrated in a speech by the Minister for Social Development and Employment, Steve Maharey (2003), where he said, "productive working relationships across all agencies are required to ensure all ships in the fleet are sailing the same course".

Promotion of a social development model is a core feature in the current Government's arsenal for achieving its social objectives. Maharey (2003) describes the social development model as "working in partnership with people, their families and communities, to help break the cycles of dependence and build the bridge to participation--socially and economically". The social development model is, in essence, based on establishing an effective framework for cross-sectoral social policy (Ministry of Social Policy 2001).

THE STRENGTHENING FAMILIES STRATEGY

The Review of the Centre identified a need to enhance "regional coordination of State sector agencies, including their interaction with local government and community organisations, by adapting and building on existing successful models of local coordination" (State Services Commission 2001:7). I argue here that the Strengthening Families model of interagency case management adopted throughout New Zealand would be a useful model to build on.

The Strengthening Families Strategy (5) introduced by the National Government in 1997 involves the coordination and cooperation of a range of both government departments and third sector organisations. It was based on a model of interagency cooperation known as Effective Practice, developed by the Waitakere City Council. The Strengthening Families Strategy was initiated in response to concern about gaps in services at the local level and the recognition by the health, education and welfare sectors that there were many similar areas of concern, for example, mutual clients in common areas of New Zealand.

The Strengthening Families initiative was an attempt to use a collaborate approach to improve outcomes for young people and families experiencing multiple disadvantage, employing a case management approach (Ministry of Social Policy 1999:12). These "high-risk families" (6) can have up to 23 social service agencies working with them at any one time. Local coordination groups have been established throughout the country and include front-line workers from the health, education, welfare, justice, housing and employment sectors, and other government and third sector agencies, as well as iwi-based social-service providers.

The goal of local-level coordination or interagency case management is to get all the agencies (both government and third sector) together with the family to collaboratively develop an action plan to enable the family to access a range of social services to address their needs. Without a high level of coordination and collaboration between agencies, this proliferation of support has the potential to add to the stress these families are experiencing as the family can end up having to spend a lot of time dealing independently with each agency.

Evaluations carried out to date (Bennett 2002, Parsons 2002, Visser 2000, Christchurch City Council 1999, Nuthall and Richardson undated) indicate that, in those areas where the Strengthening Families interagency case management process is well established, there has been enhanced communication and coordination between agencies, which in turn helped to identify gaps and eliminate overlaps in services, clarify roles and strategies, and provide a more holistic approach to meeting families' needs.

Research at the local case-management level found that interagency coordination also meant that agencies Could get the whole picture of the family and could then develop appropriate interventions. This coordinated approach also resulted in families receiving a consistent message from agencies and their staff. Networking with employees from other agencies, honest and open sharing of information and resources, and organisational commitment were identified by research participants as the main elements that facilitated interagency collaboration. In addition, the improved relationships developed through Strengthening Families also facilitated coordination and collaboration in other areas of joint work. Front-line staff and managers identified that a long-term outcome of the case-management model was the establishment of a norm or culture of collaboration between and within agencies involved in Strengthening Families.

POLICY NETWORK ANALYSIS

If we accept that one way of increasing the likelihood of effective policy outcomes is by paying more attention to administrative systems and structures, then how should we go about this in such a complex policy world crowded with so many players? The answer is to look beyond single organisational settings and focus on networks. The network approach has already made an important contribution to the policy sciences in general and to interorganisational research in particular (Carlsson 2000, Marsh and Smith 2000). Yet, as Carlsson (2000) argues, a major advance in this field would be to analyse networks as independent rather than dependent variables, which is not only to describe the features and structures of networks, but also to demonstrate to what extent they have some explanatory power.

Within any given sector, such as welfare, there is what O'Toole (1997) refers to as dense policy spaces. This is where any one department has a number of initiatives or programmes on the go, which are likely to have different objectives, funding sources and priorities, mandated criteria and targeted stakeholders. In addition, departments may also be involved to varying extents in a number of initiatives led by other departments. Achieving something meaningful in any one programme or initiative often means taking into account what is happening in other policy areas. In other words, public administration increasingly takes place in settings of networked actors who necessarily rely on one another and yet cannot compel compliance on the part of the rest (O'Toole 1997). While contracting can provide some leverage for enforcing compliance, this is limited to the parties to the contract, and to the areas specified within the contract. Agencies undertaking joint initiatives do not always formalise their working arrangements and expectations through contracts.

Networks can, accordingly, be seen as structures of interdependence involving multiple organisations or parts thereof, where one unit is not merely the formal subordinate of the others in some larger hierarchical arrangement. Consequently, a policy network (7) is described by its actors, their linkages and its boundary. The linkages between the actors serve as channels for communication and for the exchange of information, expertise, trust and other policy resources (Monge and Contractor 2003). The boundary of a given network is determined from a process of mutual recognition dependent on functional relevance and structural embeddedness.

Policy network analysis is, therefore, a useful methodological tool for describing contemporary settings in which policy making and implementation often take place. However, if policy network analysis is to move beyond the descriptive, there is a need to marry it with theory. Doing so can improve our understanding of the outcomes of network activities (i.e., network performance) and thus ideally contribute to the improvement of actual policy processes.

The question that then follows is, which theory is most appropriate? Although some progress has been made, there continues to be, as O'Toole (1997) argued, a paucity of scholarship in the field of networks in public management.

Policy Network Theory Building

Organisational theories would appear to have some utility for understanding networks. However, the problem with linking network analysis with organisational theories is that the latter tend to focus on relations between whole organisations, whereas policy or programme networks often consist of pieces of many different organisations. Some research by implementation theorists and others have focused on game theory for understanding network settings (Stoker 1991, Lynn 1993). However, while game theory is useful for looking at interdependencies, it does not easily adapt to settings that combine hierarchical structures and horizontal and vertical relationships, which have significant implications for functionally induced interdependence and bargaining and power differentials within a network (O'Toole 1997).

While network analysis combined with some of the above theories can help shed light on the effects of network settings on outcomes, as Milward and Provan (1998:388) pointed out, no social network research has attempted to map the set of programmatic relationships among sets of organisations in the public and non-profit sector that jointly produce publicly funded services. This is what Milward and Provan set out to do through their research on mental health services networks in four cities in America. As O'Toole (1997) comments, Milward and Provan's work marks one of the first efforts to assess network performance, and to link performance with structural features of the networks themselves.

Borgatti and Foster (2003:993) suggest that probably the biggest growth area in organisational network research is social capital. Burt (2000:346) considers, if attention were focused beneath the social capital metaphor on the specific network mechanisms responsible for social capital, the current research on social capital could be usefully extended. Burt (2000:346) goes on to say that "... as it is developing, social capital is at its core two things: a potential technology and a critical issue. The technology is network analysis. The issue is performance".

When combined with policy network analysis, a social capital framework holds much promise for examining cross-sectoral policy involving both government and third sector agencies in network settings such as Strengthening Families. By using this approach, the interrelated impacts of both structure and agency on processes are taken into account. Consequently, using this framework to examine Strengthening Families networks can provide important insights into determinants of interagency coordination and collaboration. Using a combination of qualitative and quantitative research methods in this approach can help to identify the ways in which individuals involved in Strengthening Families networks work together, and the processes and mechanisms of policy mediation adopted by individual members within institutional contexts. Analysing individual discourses will allow theoretical inferences as to the mechanisms of mediation of policy in multi-agency settings to be drawn, and give voice to the different interpretations that individuals at the local level provide for policy texts. (8)

Researching Policy Networks

Using policy network analysis within a social capital framework approach involves two principal stages. The first is to map networks to identify who is involved and the strengths of links between agencies. The second stage is to determine the factors that affect the nature of those links. There are a number of software programs developed specifically for analysing social network data that can be adapted to examine and map networks of agencies involved in any particular programme or policy. One such program, UCINET, (9) has the facility to measure network characteristics such as size, density, connections, power and reciprocity. Borgatti et al. (1998) have devised a range of network measures that can assess variables that relate specifically to social capital, for example, standard ego-network measures such as composition, (10) and centrality measures such as closeness. (11)

Network analysis can be used to show the structure of connections within which the actor is embedded. Using network analysis, actors are described by their relations (or nodes and ties etc.), not just their attributes (Hanneman 2001:3). For example, interval measures of relations can be used to show frequency or intensity of ties or connections between actors using data on number of email, phone, postal or face-to-face communications between actors. Full-rank ordinal measures of relations could be used to assess which actors within the network other actors consider most trustworthy.

Using software to analyse data and produce matrices, graphs and sociograms of multiplex relations and links between actors will enable policy makers to have a clearer picture not only of which agencies are actually involved in a policy arena, but also their level of involvement, the strengths and weaknesses in the network links, and who the key players are. This information could be useful for determining and allocating resources within the network, as well as identifying actual or potential gaps and overlaps in service provision.

UCINET can also analyse network data to show substructure in terms of groupings or cliques. In this sense, a clique is a subset of a network in which the actors are more closely and intensely tied to one another than they are to other members of the network. A map of the whole network can be built up by examining and noting the sizes and overlaps of the various cliques (Hanneman 2001:79). The number, size and connections among the subgroupings in a network can tell us a lot about the likely behaviour of the network as a whole. It can give us indications of how resources such as information might flow to actors in the network. It can also show the extent to which subgroups and social structures overlap one another. In addition, it may provide useful information for identifying certain individuals or agencies that might act as "bridges" among groups. Network mapping techniques can, therefore, provide vital information on the extent and nature of agency fragmentation in the area of social services.

While mapping techniques such as those mentioned above can be useful for developing a picture of organisations involved in a network and their links, differences in the structures of networks can be hard to explain. Qualitative research methods such as interviewing key stakeholders can be useful, here, to make sense out of the quantitative data and help explain why network structures differ in importance (Lazega 1997). Qualitative research would be needed to supplement quantitative data to determine levels of social capital in terms of trust and shared norms and values within the network.

SOCIAL CAPITAL

Social capital (12) is a contested term, with differing views on the extent to which it can be measured, let alone built effectively though public policy interventions (Balloch and Taylor 2001:43). While there is no single definition of the term "social capital", Spellerberg's (2001:9) definition seems particularly relevant for the policy context. She defines it as "relationships among actors (individuals, groups and/or organisations) that create a capacity to act for mutual benefit or a common purpose". Partnership as a concept sits alongside the emerging social capital paradigm in stressing that the relationships that connect people and organisations, rather than the unique qualities of different forms of organisation and structure, are important determinants of outcomes (Robinson 1999). We can see, therefore, that any assessment of social capital within networks would need to focus on the "quality" of relationships among network members. Social capital can be used to benefit both individuals and collectives. However, there can also be negative externalities to the use of social capital. Many groups achieve internal cohesion at the expense of outsiders, who can be treated with suspicion, hostility or outright hatred. Both the Ku Klux Klan and the Mafia achieve cooperative ends on the basis of shared norms and therefore have social capital, but they also produce abundant negative externalities for the larger society in which they are embedded (Fukuyama 1999).

For several years now, social capital has been of interest in New Zealand. In particular, central and local governments have recognised that an understanding of social capital may contribute to a broader analysis of policy options and issues. Relevant policy areas are varied--from education, health and justice, to industrial development, productivity and economic growth. There are now an increasing number of departmental strategy documents making reference to the advantages of addressing issues of social capital in policy as a means of improving social outcomes and promoting community cohesiveness and development. Conversely, promoting aspects of social capital such as networks, trust, and shared values and norms, which research shows are so important for effective collaboration, can be one way of addressing fragmentation. Trust, for example, can lower the cost of cooperation and make people more willing to share information, which can be critical in areas such as identifying and acting on cases of child abuse. Developing good working relationships with other agencies can not only have a positive effect on building trust, but can also allow low-cost knowledge or information sharing as a by-product of contacts occurring for other reasons.

Social Capital and Policy Networks

Using James Coleman's (1990) concept of "appropriable social organisation", (13) we can see how dense Strengthening Families interagency networks, with high levels of social capital, have the potential to become a forum for addressing a range of complex social issues and problems that are facing families in New Zealand and that would benefit from a cross-sectoral approach.

In relation to Strengthening Families networks, applying the concepts of bonding and bridging social capital (14) may help to understand how different networks of organisations such as faith-based agencies, voluntary agencies and government agencies come together and share information and resources. Linking social capital (which refers to relations between different social strata in a hierarchy where various groups access power, social status and wealth) is also a useful concept to examine networks of agencies involved in Strengthening Families. Woolcock (2000) relates linking social capital to the capacity of individuals and communities to extract resources, ideas and information from formal institutions beyond the immediate community radius. In terms of Strengthening Families, using a holistic approach to respond to families with multiple needs necessitates accessing resources and services from a range of agencies in the public, private and third sectors. Such agencies may, in turn, be locally, regionally or centrally driven, with varying resources available at their disposal and with different degrees of autonomy.

A word of caution is needed here. While social capital has the potential to positively affect the efficiency of the public sector, high concentrations of social capital among the political and bureaucratic elite can also threaten the effectiveness and legitimacy of the state by breeding corruption (Milward and Raab 2002). Within a community setting, the same strong ties that help to bind together members of a group paradoxically often enable it to exclude those deemed "outsiders". Such exclusion can be explicit or implicit, overt or covert, conscious or unconscious. Whatever the reason or basis for the exclusion, the effect is a denial of access to potentially useful networks and resources, which has significant policy implications for agencies or communities that may already be marginalised.

The Case Against Coordination

Questions inevitably arise about whether multi-agency initiatives actually work and whether the investment in planning pays off for all partners. Issues concerning power, control and access to resources are particularly important for networks or partnerships. It may be crucial to know which agencies invest most time and energy in multi-agency initiatives, and which interests are really served. Another critical question that often arises in relation to networks and partnerships is the extent to which one or more key agencies dominate in terms of decision making, resource allocation and so on.

While it is easy to assume that more coordination and cooperation will automatically lead to better and more efficient policy outcomes, we must also bear in mind that collaboration can often be resource intensive, slow the whole process down, and even bring it to a standstill if agreement between parties cannot be reached (Miller and Ahmad 2000). Moreover, the probability of reaching a stalemate increases dramatically when there are a number of involved actors holding veto positions. Some studies further suggest that the openness of networks can deteriorate with ongoing stabilisation and institutionalisation, and that instead of creating innovative solutions, the old structures are reproduced over and over again (Milward and Raab 2002).

The prevailing view has generally been that interdependent groups of two or more organisations that consciously collaborate and cooperate with one another are more effective than when the same organisations go their own way (Alter and Hage 1993). The logic behind this belief is powerful and builds on concepts from game theory that cooperation will produce outcomes that are more favourable to both parties than when they compete (Axelrod 1984). Nevertheless, despite their promise, without sufficient research and evidence it may be premature to categorically conclude that networks are the most effective mechanisms for addressing complex policy problems.

Local networks can have their own political economy, that is, inequalities and irreconcilable interests within and between networks. For example, even the flattest of networks will have some sort of hierarchy of power that distributes tasks, competencies and rewards unevenly between members. Networks are not necessarily democratic institutions, nor do they necessarily see it as their mission to put collective social goals before individual or even agency goals.

In addition, networks can become a vehicle in which the powerful, however defined, maintain or extend their sphere of influence. There will usually be some individuals within the networks that have veto power. Indeed, we cannot ignore the facade of "trust" and the rhetoric of "collaboration" used to promote vested interests through the manipulation of and capitulation by weaker partners. It may be that the powerful players in a network end up dominating rather then empowering the local community.

CONCLUSION

The traditional concerns of public administration, such as efficiency, effectiveness, equity, responsiveness and responsibility, apply just as much to network settings as they do to traditional hierarchical and bureaucratic settings (O'Toole 1997). Ultimately, the most important policy implications concerning social capital within policy networks for public administration are the pragmatic ones--the how, why and what of coordination and collaboration. This paper suggests that using network analysis to research what works in terms of intra-agency and interagency collaboration in New Zealand can potentially lead to improved policy outcomes through better understanding of the agency and structural factors that inhibit or facilitate collaboration.

Maintaining relations with many different agencies is important, as agencies depend on one another to jointly produce services funded by taxpayers. By understanding the relationships between agencies that need to work together, staff can target where they need to spend their time building new relationships, or improving existing ones. In addition, by identifying those agencies that are considered crucial for the success of a cross-sectoral policy initiative, but which are not well linked to the network, resources can be directed towards building up relationships with those agencies.

(2) A "wicked problem" (a term coined by Horst Rittel and Melvin Webber) is one for which each attempt to create a solution changes the understanding of the problem. Wicked problems cannot be solved in a traditional linear fashion, because the problem definition evolves as new possible solutions are considered and/or implemented. In addition, the challenge of solving wicked problems is exacerbated by social complexity, that is, the number and diversity of stakeholders who are part of the problemsolving process (Conklin 2001, Rittel and Webber 1973).

(3) The term "third sector" in this article refers to non-government organisations including voluntary, community, faith-based and not-for-profit organisations providing social services.

(4) "Progress Report on Integrated Service Delivery Programme", Joint State Services Commission/Ministry of Social Development report to the Minister of State Services, and Minister of Social Services and Employment. Released by the Minister of State Services, 2 April 2003, http://www.ssc.govt.nz/display/document.asp?docid=3056& PageType=toc&displaytype=pf [accessed 2 May 2003].

(5) For more information on the Strengthening Families Strategy and related programmes, see the Strengthening Families website at http://www.strengtheningfamilies.govt.nz

(6) "High-risk families" are defined as experiencing more than one of the risk factors of poor parenting, poor education attainment, poor housing, poor neighbourhoods, low income, long-term unemployment, single parenthood and high residential mobility.

(7) The term "policy network" is used in this paper to refer to multiple agencies that have some ongoing involvement in relation to a particular policy, programme or strategy at any one or more stages of the policy cycle. Stages of the policy cycle are not seen as distinctly separate or linear; consequently, the term "policy" in this paper may refer to any stage of the policy process. Indeed, using Michael Lipsky's (1980) typology of "street-level bureaucrats", front-line workers are deemed to be effectively making policy through their exercise of discretion and relative autonomy from organisational authority.

(8) The term "policy text" refers to the conceptualisation of policy not as a fixed entity, but as evolving through a process of contestation, thus emphasising how arrangements for implementation are integral to what policy becomes.

(9) For more information on the software UCINET, see the software developers Analytic Technology's website at http://www.analytictech.com.

(10) Compositional measures identify the number of alters (person that the ego is directly connected to) with high levels of "needed characteristics" such as wealth, power or expertise that have "currency" as social capital (Borgatti et al. 1998:4)

(11) "Closeness" refers to the total graph theoretic distance from ego to all others in a network. In terms of social capital, the greater the distance to other nodes, the less the chance of receiving information in a timely way (Borgatti et al. 1998:5).

(12) The three theorists most frequently cited for their contributions to the development and use of the concept of social capital are Pierre Bourdieu (1983), James Coleman (1988) and Robert Putnam (1993, 2000).

(13) Coleman (1990) describes how organisations brought into being for one set of purposes can also assist others. This form of social capital particularly relates to voluntary organisations that are created to further one purpose, but that stay in existence with new objectives after the original objective is achieved.

(14) Putnam (2000:22-23) uses the terms "bonding" and "bridging" social capital in his book Bowling Alone. He describes how "bonding capital is good for undergirding specific reciprocity and mobilising solidarity ... Bridging networks, by contrast, are better for linkage to external assets and for information diffusion".

REFERENCES

Alter, C. and J. Hage (1993) Organisations Working Together, Sage Publications, Newbury Park.

Axelrod, R. (1984) The Evolution of Cooperation, Basic Books, New York.

Balloch, S. and M. Taylor (eds.) (2001) Partnership Working--Policy and Practice, The Policy Press, Bristol.

Bennett, T. (2002) Evaluation Report--Agency Workers Participating in the Strengthening Families Programme: Tauranga Moana, Ministry of Social Development, Wellington.

Borgatti, S. and P. Foster (2003) "The network paradigm in organisational research: a review and typology" Journal of Management, 29(6):991-1013.

Borgatti, S., C. Jones and M. Everett (1998) "Network measures of social capital" Connections, 21(2):1-10.

Boston, J. (2000) "The problems of policy coordination: the New Zealand experience" Governance, 5(1):88-103.

Boston, J., P. Dalziel and S. St John (1999) Redesigning the Welfare State in New Zealand: Problems, Policies, Prospects, Oxford University Press, Auckland.

Bourdieu, P. (1983) "Forms of capital" in J.C. Richards (ed.) Handbook of Theory and Research for the Sociology of Education, Greenwood Press, New York.

Brown, M. (2000) Care and Protection is about Adult Behaviour--the Ministerial Review of the Department of Child, Youth and Family Services, Department of Child, Youth and Family Services, Wellington.

Burt, R. (2000) "The network structure of social capital" in B. Staw and R. Sutton (eds.) Research in Organizational Behaviour--An Annual Series of Analytical Essays and Critical Reviews, 22:345-423.

Carlsson, L. (2000) "Policy networks as collective action" Policy Studies Journal, 28(3):502-520.

Cheyne, C., M. O'Brien and M. Belgrave (2000) Social Policy in Aotearoa New Zealand--A Critical Introduction, Oxford University Press, Auckland.

Christchurch City Council (1999) Evaluation of the Canterbury Strengthening Families Project conducted for the Canterbury Strengthening Families Management Group, Christchurch City Council, Christchurch.

Coleman, J. (1988) "Social capital in the creation of human capital" American Journal of Sociology, 94:95-120.

Coleman, J. (1990) Foundations of Social Theory, Harvard University Press, Cambridge.

Community and Voluntary Sector Working Party (2001) Communities and Government--Potential for Partnership, Ministry of Social Policy, Wellington.

Conklin, J. (2001) Wicked Problems and Fragmentation, CogNexus Institute, Maryland, www.cognexus.org/id29.htm [accessed 12 May 2003].

Fukuyama, Francis (1999) "Social Capital and Civil Society", IMF Conference on Second Generation Reforms, IMF Headquarters, Washington, DC.

Gregory, R. (2003a) "Political life is larger than intervention logic" Public Sector, 26(1):4-6.

Gregory, R. (2003b) "All the King's horses and all the King's men: Putting New Zealand's public sector together again" International Public Management Review, 4(2):41-58. Available at www.ipmr.net

Hanneman, R. (2001) Introduction to Social Network Methods, University of California, Riverside.

James, C. (1998) "The state ten years on from the reforms", State Services Commission Occasional Paper No. 6, State Services Commission, Wellington.

Kelsey, J. (1997) The New Zealand Experiment--a World Model for Structural Adjustment?, Auckland University Press--Bridget Williams Books, Auckland.

Lazega, E. (1997) "Network analysis and qualitative research: a method of contextualisation" in G. Miller and R. Dingwall (eds.) Context and Method in Qualitative Research, Sage Publications, London.

Le Gales, P. (2001) "Urban governance and policy networks" Public Administration 79(1):167-184.

Lipsky, M. (1980) Street-Level Bureaucracy: Dilemmas of the Individual in Public Services Russell Sage Foundation, New York.

Lynn, L. Jr (1993) "Policy achievement as a collective good: a strategic perspective on managing social programs" in Barry, Bozeman (ed.) Public Management: The State of the Art, Jossey-Bass, San Francisco.

Maharey, S. (2003) Speech Notes--Connecting Policy, Research and Practice--29 April 2003.

Mallard, T. (2003) Speech to the Chilean Congress on New Zealand's State Sector Reform--17 April 2003.

Marsh, D. and M. Smith (2000) "Understanding policy networks: towards a dialectical approach" Political Studies, 48:4-21.

Miller, C. and Y. Ahmad (2000) "Collaboration and partnership: an effective response to complexity and fragmentation or solution built on sand?" International Journal of Sociology and Social Policy, 20(5/6):1-38.

Milward, H. and J. Raab (2002) "Dark networks: the structure, operation and performance of international drug, terror, and arms trafficking networks", paper presented at the International Conference on the Empirical Study of Governance, Management and Performance, 4-5 October, Barcelona.

Milward, H. and K. Provan (1998) "Measuring network structure" Public Administration, 76(2):387-407.

Ministry of Social Development (2003) Care and Protection Blueprint 2003: Families, Communities, and Government Working Together for the Safety and Wellbeing of Children, Young People and Their Families, Ministry of Social Development, Wellington.

Ministry of Social Policy (1999) Welfare to Well-being and Strengthening Families Initiatives--Implementation, Ministry of Social Policy, Wellington.

Ministry of Social Policy (2001) The Social Development Approach, Ministry of Social Policy, Wellington.

Monge, P. and N. Contractor (2003) Theories of Communication Networks, Oxford University Press, New York.

Nuthall, J. and M. Richardson (undated) Strengthening Families: Evaluation of Agency Practice, La Trobe University, Victoria, www.latrobe.edu.au/aipc/ HTML%20abstracts/Nuthall,%20J-166.html [accessed 14 May 2003].

O'Toole, L. Jr (1997) "Treating networks seriously: practical and research-based agendas in public administration" Public Administration Review, 57(1):45-52.

Office of the Commissioner for Children (2000) Final Report on the Investigation into the Death of Riri-o-te-Rangi (James) Whakaruru, Office of the Commissioner for Children, Wellington.

Office of the Commissioner for Children (2003) Report of the Investigation into the Deaths of Saliel Jalessa Aplin and Olympia Marisa Aplin, Office of the Commissioner for Children, Wellington.

Parsons, J. (2002) Evaluation Report--Agencies involved in the Palmerston North, Manawatu and Horowhenua Strengthening Families Management Groups, Ministry of Social Development, Wellington.

Putnam, R. (1993) Making Democracy Work--Civic Traditions in Modern Italy, Princeton University Press, Princeton, New Jersey.

Putnam, R. (2000) Bowling Alone: The Collapse and Revival of American Community, Simon and Schuster, New York.

Rittel, H. and M. Webber (1973) "Dilemmas in a general theory of planning" Policy Sciences, 4(2):155-169.

Robinson, D. (1999) "Partnerships--practice and theory", Institute of Policy Studies Policy Paper--Partnerships from Practice to Theory, 2:1-3.

Spellerberg, A. (2001) Framework for the Measurement of Social Capital in New Zealand, Statistics New Zealand, Wellington.

State Services Commission (2001) Report of the Advisory Group on the Review of the Centre, State Services Commission, Wellington.

Stoker, R. (1991) Reluctant Partners, University of Pittsburgh Press, Pittsburgh.

Visser, V. (2000) Strengthening Families--Views on Inter-agency Collaboration and Collaborative Case Management: a Report on the Results of a Survey, Ministry of Education, Wellington.

Wallis, J. and B. Dollery (2002) "Policy leadership styles and the process of paradigmatic policy change: three propositions", Economics Discussion Papers No. 0203, University of Otago, Dunedin.

Woolcock, M. (2000) Using Social Capital: Getting the Social Relations Right in the Theory and Practice of Economic Development, Princeton University Press, Princeton, New Jersey.

Ann Walker (1)

School of Government

Victoria University of Wellington 

(1) Acknowledgement

I would like to thank Associate Professor Bob Gregory, Dr Karen Baehler and Patrick Nolan from the School of Government, Victoria University of Wellington, and the two anonymous referees for their helpful comments on earlier drafts of this paper. All errors are my own. I am using the Strengthening Families Strategy as a case study for researching cross-sectoral policy implementation as part of my doctoral research. My doctoral research was supported by funding from the Social Policy Evaluation and Research Committee (SPEAR).