Understanding Collective Impact in Australia: A new approach to interorganizational collaboration

Abstract

Countries around the world spend substantial amounts of money on programmes designed to address social issues such as place-based disadvantage, health and aged care. Despite such huge investments, evidence shows that many of these social problems are complex and remain far from being resolved, and in some situations, they are worsening. To face these challenges, many organizations have turned to interorganizational collaboration as a more effective means of dealing with social issues. This exploratory qualitative study investigates a relatively new framework for tackling complex social challenges – Collective Impact. We show that while the interpretation and application of the Collective Impact framework varies, broad similarities can be found. We conclude that Collective Impact is best conceptualized as a method for network-based collaboration rather than a distinct methodology or philosophy, and emphasize the importance of the relational aspects of interorganizational collaboration.

Keywords

Collective Impact complex social issues interorganizational collaboration network structures not-for-profit sector place-based disadvantage systemic change

1. Introduction

In many countries, a substantial proportion of gross domestic product (GDP) is directed into programmes aimed at addressing social issues. In Australia, for example, approximately AUD240 billion was spent on social programmes in 2013¹, representing around 16.4% of the country’s GDP² for that year. Moreover, it is estimated that in 2050 half of government expenditure in Australia will be spent on health, age pensions and aged care – a doubling of current expenditure (Australian Treasury, 2010). Despite this significant investment, there is evidence that the problems being addressed, such as place-based disadvantage, indigenous health and homelessness, are far from being resolved and in some cases are worsening (Australian Institute of Health and Welfare (AIHW), 2014; Vinson and Rawsthorne, 2015). For example, while there are over 1200 specialist homeless services across Australia, there is a ‘distinctly linear feel to some parts of the homelessness service system’ (Johnson, 2012: 184), in spite of Australian approaches being ‘more strategic’ than those found elsewhere (Minnery and Greenhalgh, 2007: 652). Government and non-government organizations may be working on similar issues, using established mechanisms for cross-organizational communication and reporting, but not with the degree of collaboration needed. Social issues such as these are often referred to as ‘wicked problems’ (Rittel and Webber, 1973) as they are inherently complex and difficult to describe and often cross organizational, state or national boundaries. Furthermore, they evolve within broader systems made of many interdependent parts and have ‘multiple causes, complex interconnections and feedback loops’ (Anderson, 1999: 217).

Organizations concerned with responding to social issues form part of the ‘social purpose’ system – an increasingly mixed sector in which government and for-profit, not-for-profit and philanthropic organizations as well as individuals, work either separately or together towards improved social outcomes (Centre for Social Impact, 2016). This system as a whole can be very different from the sum of its parts, rendering the behaviour of the whole not predictable from the behaviour of constituent parts (Van Beurden et al., 2011). There are often conflicting views of the causes and solutions on the part of stakeholders, and constraints and resources to solve the problem often shift over time, requiring new knowledge and behavioural change (Bovaird, 2008). For this reason, single-organization or project-based approaches often fail to address the problems they seek to resolve (Keast et al., 2004). Indeed, it is typically beyond the purview of individual organizations − regardless of sector or resourcing − to address complex social issues or generate social change at the desired scale. Instead, a systemic approach is required, recognising the inherent interconnectedness among actors in the social purpose system and the need to combine resources and efforts directed towards a collective cause (Keast and Mandell, 2014). For these reasons, many organizations have turned to interorganizational collaboration as a more effective means of dealing with social issues (Mandell and Steelman, 2003). In Australia, there have been some successful collaborative initiatives involving social purpose organizations communicating and working together. These have included the federally funded Communities for Children (Muir et al., 2010) and Better Futures Local Solutions (Department of Human Services, 2012) initiatives and the New South Wales (NSW) Government’s Connected Communities strategy (NSW Department of Education and Communities, 2011). However, initiatives such as these remain subject to the vagaries of political funding processes, and in spite of early successes, social issues remain significant.

In this article, we focus on a relatively new framework for tackling complex and systemic social challenges known as ‘Collective Impact’ (hereafter CI). At the core of this collaborative interorganizational framework is the idea that lasting and meaningful social change requires coordinated efforts from multiple stakeholders working towards a common overarching goal. According to its proponents, CI provides a mechanism for taking a systems view to social change (Kania et al., 2014). CI projects have been successfully completed in North America (Hanleybrown et al., 2012) and some Australian states (O’Neill and Graham, 2013). A review of the published literature indicates, however, that most of the work is descriptive and practitioner based (see, for example, Hanleybrown et al., 2012; Pro Bono Australia, 2014). There have been no theoretically informed studies of CI processes in the organizational studies literature, undoubtedly due to the nascent nature of this concept.

In this exploratory study, we are interested in gaining insights into CI practices as a form of network-based interorganizational collaboration (Mandell and Steelman, 2003). Research into interorganizational collaboration has tended to reflect concerns in strategic management (where collaboration among private sector organizations is the main focus) or public administration (where the focus is mainly on collaborations involving government organizations). Network management (and, relatedly, network governance) frameworks have featured in both economic sociology and public policy literature for some time, but less so in the management literature (Dal Molin and Masella, 2015). Our study hence asks, ‘How is the CI framework being applied and interpreted in Australia?’ Our theoretical contributions are threefold. First, and importantly, we situate the CI framework within its rightful theoretical roots – network-based interorganizational collaboration – and provide a theoretically robust study of CI processes typically missing in the field. Second, we position the CI framework as a ‘method’ (or technique) within which to collaborate rather than a new ‘methodology’ (or underlying philosophy). And third, we emphasize the importance of the relational aspects of interorganizational collaboration in networks.

This article is organized as follows. We start by canvassing the literature on network-based interorganizational collaboration and the CI approach. We then present our methods for our field-based investigation of CI initiatives in Australia and our findings. Finally, we conclude by discussing implications for theory and practice.

2. Literature review

2.1. From relationships, to networks, to collaboration

Organizations operating within the social purpose system have been criticized in the past for their isolated ‘silo’-based approaches that promote fragmented and duplicated efforts, thus limiting their impact and adding to overall expenditure on addressing social issues (Kania and Kramer, 2011). To maximize their impact, Reitan (1998) argues, organizations need to work together, through the development of effective and robust interorganizational relationships.

Interorganizational relationships can differ in terms of the intensity of the commitment by partner organizations, the mechanisms of control and the extent to which the relationships and joint activities are formalized (Keast and Mandell, 2014). There are three main types of relationships. Cooperation, the most informal type of interorganizational relationship, is characterized by shorter time frames and little investment of time or money and partners’ objectives remaining fully independent so that risk is minimized. Coordination is slightly more formalized. It is characterized by medium-term time frames, shared information and funding and more regular contact and exchange of information, although it is often limited to specific programmes or projects. Collaboration is the most formal of interorganizational relationships. Collaboration involves shared ideas and resources, reciprocity and, ideally, a long-term shared mission that transcends individual organizational interests (Keast and Mandell, 2014). Collaboration is hence distinguished by ‘strategically interdependent action’ and ‘broad tasks that reach beyond the simultaneous actions of independently operating organizations’ (Keast et al., 2004: 364).

The literature on interorganizational collaboration is substantial, particularly within the fields of public health, public administration and (strategic) management, although in the latter case, it has largely been confined to business organizations. In the management literature, collaboration between organizations has been found to lead to a number of positive effects (Hardy et al., 2003). From a strategic management perspective, collaboration enables organizations to expand their pool of resources and skills and develop distinctive capabilities. From a learning and innovation perspective, collaboration engenders synergistic knowledge creation through ‘communities of practice’. A political perspective throws light on the political effects of collaboration that interorganizational ties extend beyond specific individual relationships, shifting patterns of influence and power (Hardy et al., 2003). In practice, collaboration allows partner organizations to aim for ‘an objective which no individual organization could have met alone and achieving the objectives of each collaborating organization better than it could alone’ (Huxham and Macdonald, 1992: 51; emphasis in original). When addressing complex social issues, collaboration is hence an ideal approach as it allows a more holistic provision of services and the synchronization of efforts and resources (Bryson et al., 2006).

As Bryson et al. (2006) note, however, interorganizational collaborations that span public, private and not-for-profit sectors are inherently ‘difficult to create and even more difficult to sustain’ (p. 52). Nonetheless, some key elements underpinning successful collaboration can be identified in the literature, as increasing attention is being directed towards this area (Gajda, 2004). At a structural level, key factors include clear roles and responsibilities, commitment to a common mission, a governing structure and evaluation mechanisms (Dal Molin and Masella, 2015). Process characteristics, such as open and frequent communication, conflict management and effective, trust-based informal relationships, are identified as critical elements of working together successfully. Environmental factors – for example, stakeholder support and resources available – can also affect the success of collaboration (Mandell and Keast, 2008).

The types of arrangements through which collaborative relationships are structured are important as they determine the rules, norms and interests governing the relationship (Williams, 2005). In the social purpose system, the most appropriate arrangement for collaboration is the network structure (Dal Molin and Masella, 2015). Podolny and Page (1998) define networks as a ‘collection[s] of actors […] pursu[ing] repeated, enduring exchange relations with one another’ (p. 59). They are characterized by long-term, horizontal relationships between mutually dependent organizations (Keast et al., 2004; Podolny and Page, 1998; Powell, 1990). As collaborative structures, networks offer the flexibility required in complex contexts such as those we find in the social purpose system. Effective network structures in the social sector are characterized by trust, reciprocity, goodwill and commitment – all components of successful collaboration (Podolny and Page, 1998; Powell, 1990; Provan and Lemaire, 2012). This relational orientation arguably provides a better mechanism for solving wicked problems than traditional top-down (hierarchical) or market-based approaches (Weber and Khademian, 2008), and there is evidence that collaborative network structures have been core to initiatives that seek to address, for example, climate change (Pinkse and Kolk, 2012) or community well-being (Keast et al., 2004). Network structures allow the combining of efforts of individual organizations, thus increasing the pool of resources available and providing more effective service delivery by reducing duplication of services and administrative functions (Provan and Milward, 2001). Furthermore, unlike markets or hierarchies, network structures are built on a sense of mutual dependence, which tempers the pursuit of organization-specific gains at the expense of other partners. It is from this context of collaborative network structures that the distinctive form of CI emerges.

2.2. The CI approach to interorganizational collaboration

Introduced in 2011 in North America, CI has been defined as ‘the commitment of a group of important actors from different sectors to a common agenda for solving a specific social problem’ (Kania and Kramer, 2011: 36). As we have seen, collaborative approaches to complex social issues have been in place in Australia and elsewhere for some time. However, without alignment and a uniting, overarching strategy, such efforts, although collaborative to some extent, often have a limited impact (Irby and Boyle, 2014). Proponents of CI argue that it differs from these other approaches because it adopts a systems perspective that promotes large-scale cross-sector collaboration and provides a structured framework through which to enact and implement long-term social change (Kania and Kramer, 2011). The framework consists of five elements: (a) a common agenda, (b) mutually reinforcing activities, (c) continuous communication, (d) shared measurement systems and (e) a ‘backbone’ organization (Kania and Kramer, 2011).

First, a common agenda refers to ‘a shared vision for change’ (Kania and Kramer, 2011: 39), one that requires partner organizations to agree on a common issue and way to address it. It involves defining the problem being tackled and its boundaries, as well as clearly articulating roles and responsibilities of all partners involved (Hanleybrown et al., 2012). Such a process requires time and an explicit commitment to working through the perspective of each organization.

Second, mutually reinforcing activities refers to the need to build on the strength of each partner organization. It is about ‘encouraging each participant to undertake the specific set of activities at which it excels in a way that supports and is coordinated with the action of others’ (Kania and Kramer, 2011: 40). The selection of appropriate (and diverse) collaborators is important in this regard (Boyce, 2013), as is the identification of the most relevant services to provide.

Third, continuous communication enables partner organizations to build trust. Consistent and honest communication is important as it facilitates the development of a common vocabulary and helps maintain and improve collaborative efforts through the sharing of knowledge and the monitoring of results.

Fourth, shared measurement systems refers to a set of metrics shared among partner organizations to assess progress towards the initiative’s goals. Not only do they enable partners to monitor each other’s performance and hold each other accountable but they also provide a platform for continuous improvement by capturing learning from successful and failed experiences.

Fifth, and last, a backbone organization refers to an organizational entity specifically responsible for coordinating the various dimensions of the initiative. The backbone organization serves six main functions: (a) coordinating the development of a common agenda, (b) enabling dialogue among partners, (c) coordinating shared measurement systems and facilitating the collection of data, (d) liaising with external stakeholders, (e) managing community awareness campaigns and (f) fundraising.

It is the specific focus of the framework on shared measurement systems and the role of the backbone organization that sets it apart from other models of network-based collaboration (Easterling, 2013). CI’s remaining elements are common to many existing interorganizational collaborative mechanisms, as we have seen. Kania et al. (2014) contend that CI requires a change of mindset that is ‘fundamentally at odds with traditional approaches to social change’ (p. 2). The anecdotal evidence from case studies of CI initiatives, although limited, suggests that the framework has the potential to help networks successfully tackle complex social issues (Kania and Kramer, 2011). As a result, an increasing number of CI initiatives that either start with the framework in mind or retrospectively apply it to their work have emerged, particularly in North America and in Australia. However, Thompson (2014) argues that despite the interest in the framework and confidence in its potential to achieve unprecedented social change, practitioners are ‘deeply unsure about what exactly [CI] is and what changes it might imply’ (p. 2). In the light of such ambiguity, Edmonson and Hecht (2014) call for further research into CI to better understand the framework and its implementation. Australia offers a unique environment in which to explore CI because of the differences in the landscape of the social purpose sector compared with that of the United States – for example, in terms of government contribution to the sector and the relatively small number and size of philanthropic foundations (Chau, 2014).

In sum, although less formalized forms of interorganizational collaboration have been featured in the public administration literature, the CI framework remains underexplored, particularly in the Australian context. Hence, in this study, we aimed to map out the key features and success factors of CI initiatives currently under way in Australia.

3. Methodology

We approached our research questions using a qualitative methodology, the most appropriate means for considering a phenomenon such as CI within its distinct context (Gephart and Robert, 2004). Given the complexity of the social purpose system, as well as the constellations of organizations and operations collaborating under the CI ‘banner’, it was necessary to use a methodology that allowed ideas, themes and relationships to be explored as they arose (Hardy et al., 2003).

3.1. Data collection

CI is in its infancy in Australia, and up until now, there has been little formal mapping of the CI territory. The first step for our exploratory study, therefore, was to build a register of CI initiatives under way in Australia by searching online clearinghouse portals (e.g. Pro Bono Australia, 2014), state government websites and lists of participating organizations at CI events (i.e. ‘Collective Impact 2014’³ and ‘The Search’⁴). This yielded 20 potential case initiatives. This low number of organizations was in line with our expectations, as the CI practice in Australia is in a phase of early adoption, and we were aiming for analytic (rather than statistical) generalizability (Yin, 2013). Our unit of analysis was a ‘CI initiative’, which typically comprised multiple organizations representing not-for-profit, government (federal, state and/or local), business and philanthropic organizations. The relevant backbone organization for each initiative was identified and then contacted via email with an invitation to participate in the study. We received 10 replies from organizations interested in participating. From this sample, we selected five initiatives, using a diversity sampling approach (Daniel, 2012), aiming for a variety of activities and geographic locations, representing rural and urban communities in four Australian states.

Data were collected primarily through semi-structured interviews, which allowed for the focused inquiry to be balanced with our exploratory orientation, enabling diverse responses, new insights and time to explore the nuances of practice (Travers, 2012). The semi-structured approach also allowed the interviews to flow at a natural pace, which was important to build rapport with interviewees from different backgrounds (Rowley, 2012). To identify and recruit suitable participants, we asked each backbone organization to suggest participants from their partner organizations for us to contact. Further participants were recruited through snowballing techniques based on the recommendations of other interviewees. Participants were assured of confidentiality, and recruitment was undertaken in compliance with our institution’s ethics protocols. Interviews were conducted in person (in most cases) or by telephone, by one or two of the authors, and the average interview length was 55 minutes within a range of 45 minutes to 2 hours. We asked open questions that allowed participants to express their views and enabled the researchers to explore key issues and probe responses (Travers, 2012). Interviews were tape recorded and transcribed, with pseudonyms used to ensure anonymity. A total of 34 people were interviewed: 33 CI practitioners from across the five case studies and one CI consultant. As CI is an emerging practice in Australia, an interview with the CI consultant was conducted to enable mapping of the broader landscape of CI in the country, to ensure that our findings were contextualized (Tracy, 2010). Two follow-up interviews were conducted via telephone (with backbone organization representatives), as the interviews ran over the allocated meeting time and interviewees indicated they had further comments on the topics of the interview.

The interview data were supplemented with related documentary material comprising reports to funders, community development plans, internal reports, websites and social media pages and council and government documents for each initiative. In addition, two of the authors visited the sites of activities of four initiatives, including an employment programme for indigenous youth, an early childhood development initiative and a training programme for volunteers. Participants in these programmes shared their experiences with us during the visits. This provided us with further contextual data and enabled us to further ‘crystallize’ the data from the interviews (Tracy, 2010: 843).

3.2. Data analysis

The data collected through the interviews and additional documents were analysed using thematic analysis, which entails systematically searching across a data set to find repeated patterns of meaning. This analysis approach corresponded with the exploratory orientation of our study and allowed us to generate unanticipated insights, highlight new concepts and draw attention to similarities and differences (Braun and Clarke, 2006). We adopted a ‘theory-led’ approach (Locke, 2001), shaped by our preliminary literature review and subsequent reading. Initially, we constructed narrative descriptions of each CI initiative that described the geographic and demographic context, antecedents including earlier collaborative initiatives, main actors and intended outcomes. Subsequent analytic (thematic) coding involved the interpretive task of considering meanings, creating conceptual categories and abstracting from the data, and was an ongoing, iterative process and an adaptation of the constant comparative method (Strauss and Corbin, 1998). Open coding, which involved looking for initial themes (or ‘units of meaning’) in the text resulted in a large number (142) of ‘in-vivo’ and theoretically informed categories (Locke, 2001). This initial coding was undertaken by an independent coder experienced in qualitative analysis and cross-checked by two of the authors, using qualitative data analysis software. We then looked for hierarchical relationships and other patterns in these categories, using cross-case comparisons to further refine and reduce them and create second-order categories from which eventual conceptual categories could be developed (Daniel, 2012). For example, the three second-order categories of organizational, relational and leadership factors were constructed from the patterns and relationships identified in earlier rounds of coding. Relationships between these success factors were developed as part of an overall CI model (discussed later).

3.3. The CI initiatives

Table 1 outlines key features of the five initiatives in our study. Each initiative aims to address complex problems of place-based disadvantage. In all of the communities studied, collaborative practices among multi-sector organizations preceded the implementation of CI practices. Two initiatives had previously been awarded federal funding to focus on education and employment due to high levels of disadvantage in the community (Department of Human Services, 2012). This multi-agency and multi-sector programme had delivered some promising results in the various localities; however, federal funding was discontinued in mid-2014. Another two initiatives, which focused on children, health and education, had also participated in a state government–funded programme (Department of Education and Training, 2013). In both cases, key actors in each of the regions were looking for other collaborative mechanisms to address the issues of concern. The CI approach was seen as a useful instrument to continue this work.

Table 1.

Collective Impact cases.

CI initiative	Year commenced	Life cycle^a	Focus	Number of interviewees	Context
I1	2013	Incubation	Well-being and education of children	7	Regional community, ongoing disadvantage, vulnerable families and poor performance on child development indicators. Partners include business, not-for-profit, education, community, civic and philanthropic organizations. Previous experience with multi-agency programmes. Philanthropic funding
I2	2011	Progress	Education attainment and workforce participation	4	Suburban community, high disadvantage and high unemployment. Decline of local manufacturing industries. Partners include government, business and the not-for-profit organizations. Previous participant in collaborative government-funded pilot programme
I3	2014	Start-up	Well-being of children	9	Outer suburban community, high levels of disadvantage and poor performance on child development indicators. Partners include government and non-government agencies, businesses and universities. Previous participant in collaborative government-funded pilot programme
I4	2010	Progress	Community development	4	Outer regional community, place-based disadvantage, decline of local industries, low levels of school completion and high unemployment. Partners from industry, local government, service agencies and education. Previous participant in collaborative government-funded pilot programme
I5	2006	Progress, towards sustainability	Well-being of children	9	Regional community, vulnerable families and poor school achievement. Partners include community service providers, for-profit organizations, universities and government. Sponsorship from local government and some initial philanthropic support

CI: Collective Impact.

See the ‘Data analysis’ section for description.

From the documentary and interview data, we determined the ‘life cycle’ stage of each initiative, using the social enterprise framework developed by Chambers and Edwards-Stuart (2007) as a starting point. We adapted the life-cycle stages proposed by these authors to encompass what we saw as four stages of interorganizational collaboration: Incubation, Start-up, Progress and Sustainability. These stages are also in line with the stages of CI found in the practitioner literature (Hanleybrown et al., 2012; Weaver, 2013). The first stage, Incubation, is characterized by organizational actors connecting over a broad social issue, which at this stage might be quite generic (e.g. unemployment). Typically, there is dissatisfaction with the status quo and a degree of ‘emotional energy’ as parties come together. The Start-up stage entails processes of formalization, including focusing on what will become the common agenda and allocating roles and responsibilities. The Progress stage is characterized by a functioning backbone organization, ongoing communication and the strengthening of interorganizational ties. The final Sustainability stage denotes an initiative that is robust and effective in addressing the issues of concern, with signs of longevity.

4. Findings

4.1. CI landscape

Most participants indicated that the CI framework provides a more structured approach than other forms of interorganizational collaboration with which they were familiar. In the following sections, we review each dimension of the CI framework in the context of our interviews.

4.1.1. Common agenda

Across each of the initiatives, participants stressed the importance of an explicit common agenda for their work. The articulation of an explicit common agenda was seen to provide guidance, consistency and transparency.

There was broad agreement that the process for developing the common agenda needed to entail stakeholder and community engagement as well as ‘drawing all the partners together’ (Wendy, I5). On two occasions, the common agenda came out of a larger community consultation process, organized and driven by the local council. In I4, 100 individuals representing a range of different community and government organizations participated in a 2-day workshop. In addition,

We did community surveys; we went to some of the more disadvantaged communities and did some creative sort of things to get their input. We did workshops with school students, we did online surveys, we did a whole heap of different things. (Brad, I4)

The strategic plan coming out of these types of engagement processes formed the basis of the common agenda. Such processes were seen to allow for the involvement of the community from the very beginning, providing a mechanism for the targeting of needs as articulated by the community and strong ‘buy-in’.

Participants identified three key conditions for the development of a sound common agenda: (a) a focus on the issue targeted; (b) a ‘champion’ to drive the process, provide support and ensure inclusiveness; and (c) willingness of members to ‘stay at the table’. Narrowing down the initiative’s focus was seen as crucial. As one initiative representative explained, ‘We’re always picking our best bets in many ways. Because we can’t do everything’ (Leonie, I5). Similar emphasis was given to leading the development of the common agenda and supporting the person who took on this role. A representative of one initiative, for example, emphasized the need for ‘good will, trust and a willingness to stay at the table’ (Mark, I1). This determination was seen as crucial, as the development of the agenda takes time and could be a difficult and at times ‘frustrating’ process (Mark, I1).

4.1.2. Shared measurement

Our findings indicate the importance of setting up a shared measurement system that is consistent across the different initiatives and life cycle stages. As one initiative representative stated, ‘The whole notion of an ongoing basis, measuring everything that we do has been […] a very crucial thing because it’s not of any value to anyone long term if we don’t do that’ (Ruth, I3). The reasons given for developing such systems related to either communicating success to funders (from government, not-for-profit or private sectors) or providing a means of evaluating their progress and the overall impact of the initiative. A key element of shared measurement appeared to be the ability to come up with a common language.

Although there was broad agreement on the value of a shared measurement system, developing and implementing metrics presented significant challenges for all of the initiatives, partly because of the complexity of the processes they were trying to capture. A representative from initiative I5, for example, stated,

Different players, different sectors […] how do you find a system that doesn’t create extra work for everybody, extra costs? That in itself is quite a sophisticated job just figuring out what the shared indicators are going to be and then the mode of collection. (Robyn, I5)

Another participant observed that developing a shared measurement system was ‘going to be a huge task, like what indicators are we going to use? That’s going to be like having a big database’ (Ruth, I3). This was difficult, even for a long-standing initiative such as I5, which had brought in consultants: ‘We have struggled with collective measurement […]. We started off – we’re thinking, oh, this might be simple. […] We really needed some assistance with collective measurement’ (Debbie, I5).

4.1.3. Mutually reinforcing activities

The need for organizations to ensure their activities were mutually reinforcing was implicit, and participants referred to the importance of cooperation. One long-standing initiative (I5) had developed a formal mechanism for ensuring that each organization’s contributions complemented and reinforced the overarching goal. One participant explained that this understanding

Came about through working out what all of our core strengths were, getting a shared vision, having an outcomes framework to work towards. All of the things that we talk about now in Collective Impact we started doing through that process. (Judy, I5)

This process was important as some of the organizations in the past had competed for funding. Similar emphasis was placed on the need to avoid duplication of efforts. As one participant put it, ‘Without the knowledge of what other people are doing, they can miss or can even be working in opposition, which doesn’t add very much to the collective good of the community’ (Jack, I2).

4.1.4. Continuous communication

In each of the initiatives, participants acknowledged the importance of internal and external communication measures. Communication channels varied from traditional methods (such as phone calls, meetings and emails) to more structured approaches, but in all cases, communication was formally coordinated. For example, one participant observed,

There’s a variety of mechanisms. There is an overarching coordinating group that meets regularly. Its focus is to facilitate the ongoing collaboration and connection. Then there are smaller regular group meetings of particular organizations that are focusing on a particular aspect. There are maybe three or four regular mechanisms of these sorts that meet regularly. (I4, Warren)

Initiative representatives also stressed the importance of external communication channels, such as ‘networking and marketing’ (Anne, I1), as well as time spent on channels such as websites, social media and newsletters. In each case, communication appeared to be a process that had been thought through and coordinated across the initiative, suggesting that the CI framework’s emphasis on continuous and coordinated communication had been taken on board.

4.1.5. Backbone organization

One of the central features of the CI framework is the presence of a purpose-driven backbone organization. From the interview data, it was clear that most of the initiatives were still in the process of developing the backbone role. The backbone organization was referred to in different ways across the initiatives, but our data revealed insights into the process of selecting a backbone organization and its role and importance.

In some cases, the backbone role was filled by a representative from local government, and in others, a separate organizational structure with a designated full-time leader was created. In the case of the local council, it was seen as ‘the neutral player’, with a grounded knowledge in local issues and programmes – although there was always the need to manage the ‘tension’ around ‘ownership’ of the initiative (Sophia, I5). In cases in which a purpose-built organization was established, this came about in recognition of the limitations of previous collaborative approaches. For example, one participant stated,

There was really no one group that was providing – if you like that was listening to all of the individual groups about what they were learning and where they thought they could put solutions. So the idea of this, if you like, is to create an overarching kind of body to try and in a way direct the traffic. (Steve, I4)

The selection process of the backbone appeared to be ‘a conscious decision’ (Warren, I4) for most, but it was not without tensions. One initiative had put in place a structured selection process involving ‘an in-depth interview with the four possible organizations that could do it’ (Ruth, I3). For those initiatives in the Incubation or Start-up stages, the role was still under development.

The backbone in each initiative focused on slightly different aspects: administrative duties, communication and funding opportunities in the case of initiative I5; relationship building for initiative I3; coordination for initiative I4; and measurement for initiatives I1 and I2. One participant explained their backbone structure thus:

I think [the backbone] it’s important in that capacity, that structure that works […] because of that, you can go off and do the work. […] I don’t need to worry about [funding] because I know that it is being managed by other people. All I have to worry about is doing the work that I do and feeding back my results into the main group. (Sophia, I5)

In spite of some differences in function, participants agreed on the importance of the backbone. One participant commented, ‘it’s critical. It’s what actually will make it all work’ (Jim, I3).

4.2. Success factors for CI

Although CI is still taking shape in Australia, many participants had a clear view of the necessary elements for success. The change created needed to be ‘system-wide, cross-sector and […] citizen-inclusive’ (CI Consultant). Success, our findings suggest, is when ‘people have got to the stage where they’re not so much thinking about their organization first; they’re thinking about the community first’ (Brad, I4) and when ‘the sum is greater than the individual parts’ (Debbie, I5). Our analysis gave rise to three broad categories of ‘success factors’ – (a) organizational, (b) relational and (c) leadership – each of which was interrelated in mutually reinforcing ways, as indicated in Figure 1.

Figure 1.

Success factors for Collective Impact.

4.2.1. Organizational factors

References were made to a range of organizational factors that representatives saw as underpinning the success of their initiatives. Those that stood out as being important across the range of initiatives were mechanisms for citizen engagement (or community consultation), a diversity of representation, clear and robust business and governance processes and mechanisms for ongoing review and evaluation.

Our data pointed to the importance of building direct relationships with the communities served by CI initiatives. This should not be taken for granted, and the CI framework provided mechanisms for this kind of community consultation, as we saw in the previous section. Diversity across members and stakeholders in terms of experience, disciplinary background and organizational type was also indicated as important. For example, one participant explained the importance of ‘having the right mix of people involved; otherwise you just bang your head against the wall’ (I4, Steve). For another participant, it was expressed as being core to success:

If we were all the same, I think our chance of success would be diminished. I think our chances of success are heightened because we demonstrate different attributes and different styles of leadership. It makes it more challenging but it increases our chances of success. (Mark, I1)

Having a diverse representation also helped mitigate the ‘siloing’ that many participants saw as characteristic of earlier attempts at place-based collaboration.

The role played by the backbone organization, as discussed in the previous section, and in some cases by steering committees, was overwhelmingly seen as a key element of successful CI, and a key advantage over other forms of interorganizational collaboration. One participant noted that without a backbone organization, ‘instead of being well run and organized, it would be shambolic’ (Wendy, I5). Related to a well-functioning backbone organization was the presence of robust organizational processes as part of an overall need for strong governance. One participant emphasized the need for

Clear governance, a clear methodology and steps for achieving – for recognizing you are delivering Collective Impact […] different checkpoints in terms of the governance structure and the project management in saying, yes, we’ve now achieved a clear governance structure. (Stefan, I3)

Clear and effective processes were seen as necessary to mitigate the tendency for collaborative networks to ‘transcend the bureaucratic rules that keep the system fair’, and in some cases, act unethically (Michelle, I5). Various respondents also indicated the importance of ongoing review and evaluation of the progress of their initiative.

4.2.2. Relational factors

Overwhelmingly, respondents pointed to the centrality of what we define as the relational elements of CI. Key components include the development of positive relationships across the various actors in an initiative, consultative processes with the local communities, the honesty required to meet the challenges of addressing wicked problems and the mutual trust and respect that emerge over time.

Participants articulated the value of being part of a wider CI community: ‘That kind of feeling that it’s a global initiative, that you can all kind of support each other from a distance’ (Wendy, I5). In the more established initiatives, the quality of relationships was a crucial factor for success – and specific attention had been given to building these relationships. For example, one of the initiatives used the technique known as ‘appreciative inquiry’ (Van der Haar and Hosking, 2004) to build these relationships. The outcome, in the eyes of one participant, was

A growing capacity for deeper levels of conversation that create a group wisdom through the way they [CI members] listen and speak to each other. That they transcend the limits of each other’s personalities to understand the capacity each person is sitting with and know how to join that and use it with theirs. (Michelle, I5)

Others emphasized that relationship building takes time. For example, one participant stated, ‘With these sorts of initiatives it’s really important that there is a reasonable timeline. You can’t expect to build strong partnerships and trust in a very short term. It just doesn’t happen. It can’t’ (Jack, I2).

As robust relationships developed over time, participants observed that they became more comfortable with expressing their views honestly. One participant stated,

Best practice in Collective Impact is all about relationships because it’s about having the hard conversations – the difficult conversations – because you’re actually starting to tackle issues that no one’s bothered with […]. It’s about people being honest and saying, look this doesn’t work […]. Now I think we’re at the stage where people know one another, they understand; they can be honest with one another. (Brad, I4)

A corollary of robust and honest relationships was a sense of mutual trust, which participants saw as intertwined with mutual respect. For example, one participant stated that ‘people are respectful of what people at the table put a voice across for’ (Donna, I3). In another case, it involved the various members ‘agreeing not to have a turf war’ (Isobel, I3). This was an ongoing theme. Without trust, our data suggests self-interest prevails and people revert to their own organizational agendas. In these cases, trusting relationships were ‘both the lubricant and the glue […] facilitat[ing] the work of collaboration and […] hold[ing] the collaboration together’ (Bryson et al., 2006: 47–48). Ultimately, trust requires, as one participant explained, ‘a leap of faith […] because early on it’s not always apparent whether it’s going to work and whether there’s going to be opportunity for individuals or organizations’ (Mark, I1).

4.2.3. Leadership factors

Although CI places emphasis on the collective, we found much evidence to support the importance of effective leadership processes across each of the initiatives. Leadership in this context can be defined as the ‘product of interactive dynamics’ whereby ‘interacting agents generate adaptive outcomes’ (Lichtenstein et al., 2006: 4). One of the elements of leadership identified by participants involved championing the CI approach. As one participant stated,

You need at least two or three really committed champions who can sell the concept, who can work around it to bring it to life with people and to explain philosophy – explain how it’s different to all the other networks and collaborative things that we’ve got around which most people admit or think they’re useless. (Jim, I3)

Creation of a common agenda was also seen as an exercise in leadership, in creating the shared vision necessary to ‘bring voices together’ (Maryann, I1).

Leadership was expressed in different forms across the participating initiatives. In three of our initiatives, key individuals were directly identified as leaders by other members. Importantly, though, this leadership was not charismatic or authoritarian but more in tune with what has been termed ‘servant leadership’ (Sendjaya et al., 2008). One participant in I5 described the leader’s style as

She genuinely takes on people’s ideas and wants to incorporate them. So what I’d see is that she’ll bring a group of people together […] People have another idea and she moulds it. It’s not like she’s manipulating it but she genuinely makes it part of where things are going. Yet she’s still a leader […] She’d probably say she isn’t a leader. I think she genuinely authentically doesn’t think she’s the leader. (Leonie, I5)

Our findings suggest that adaptive leadership processes, involving the surfacing of conflict and ‘rich interactions’ (Lichtenstein and Plowman, 2009: 622), were also present. For example, a leader was described as ‘Someone who […] is a connector. Who is able to draw people together, who’s able to bridge different communities, different objectives, different ways of working and find commonality amongst that’ (Maryann and Rebecca, I1).

Notably, there was some recognition of the ‘shadow side’ of CI work, particularly when the self-interest of the various members is masked by the overarching drive towards the collective. One leader spoke about how, over time, she became:

More attuned to the self-interest of people around themselves, their career and their organizations. Factoring that into the process, that’s probably been my biggest learning […] I think that people that drive collaboration they have kind of a blinded view of things. It’s about the group interest and so I really learned to understand self-interest and pay respect to that. (Michelle, I5)

This ongoing tension between self-interest on the part of (some) stakeholders and the overarching concern for ‘genuine CI’ was a key element of the leadership challenges associated with this type of work.

5. Discussion

In this article, we have argued that complex social issues require organizations to work together and advanced network-based interorganizational collaboration as an effective means of doing so. We introduced the CI framework, a relatively new approach to organizing for collaboration and sought to gain insights into how the CI framework is being applied and interpreted in Australia.

Our study has shown that although the application and interpretation of CI among our sample initiatives was not uniform, the CI framework equipped organizations with a ‘blueprint’ to tackle complex social issues. This is to be expected, as a certain flexibility needs to be preserved in the application of the framework to reflect the emergent nature of the initiative’s shared purpose, as well as changes in its internal and external environment (Kania and Kramer, 2013). CI was seen as providing clear processes, thus balancing out a network structure characterized by no formal top-down relationships. However, each of the CI initiatives considered in this study showed elements of ‘deep’ collaboration; in other words, the relationships and interactions extended across all layers of the member organizations, the scope of interactions was broad and the flow of information was multi-directional (Hardy et al., 2003). In this sense, our work enabled us to situate CI as a ‘method’ or technique within which to apply principles of network-based collaboration, rather than a new ‘methodology’ or underlying philosophy.

As with other forms of network-based interorganizational collaboration, CI was found to be characterized by shared ideas and resources, as well as a long-term common agenda that transcends individual interest. We found that an explicit common agenda was an important feature of the framework, as it was seen to provide guidance, consistency and transparency. While the degree of consensus around the precise ‘content’ of an initiative’s common agenda varied, this appeared to be related to the life cycle of the initiative. Difficulties in developing a shared agenda are to be expected because, as Easterling (2013) argues, ‘it is challenging and even unnatural for organizations with distinct missions to come together and develop a shared agenda, especially one that involves systems change’ (p. 69). Consensus was reached across participating initiatives around the need to engage with stakeholders and community in developing the common agenda. While the importance of ‘selecting’ the right organizations to partner with is strongly emphasized by CI advocates (Kania and Kramer, 2013), our study found that partnerships in the initiatives studied emerged from existing arrangements. Continuous communication processes had, for the most part, been considered and coordinated across each initiative, and multiple channels and styles of communication were in place. It is worth noting, though, that communication processes need to be constantly re-evaluated in the light of new information and changes in the environment.

We found overwhelming consensus around the need to develop a shared measurement system. Developing and implementing such metrics, however, presented significant challenges for all of the initiatives. These mostly stemmed from finding shared indicators of success as opposed to individual ones. Indeed, the literature emphasizes that getting partner organizations to agree on shared measurements and finding common metrics that are also representative of distinct operations is challenging. This is exacerbated by the funding structure of social purpose organizations, which often requires reporting to funders on specific dimensions that are difficult to align (Cabaj, n.d.).

Challenges also arose with the backbone organization. We found that identifying the backbone organization within the respective initiatives was not always a straightforward process. The nature of the role of the backbone organization, as well as its importance, was largely agreed upon, although there were slightly different focuses on administrative duties, relationship building, coordination and measurement. While these activities should be common across all backbone organizations, the degree and manner in which each carries out these functions depends on the nature of the initiative and its context. Backbone organizations can take multiple forms and their support functions can be divided among multiple organizations to best suit the environment in which the initiative evolves (Hanleybrown et al., 2012).

Our study found that participants had a strong take on what they believed to be key success factors, and these formed three overarching categories. They were consistent with those identified in the network literature but highlighted additional characteristics critical to the good functioning of CI initiatives. The first category related to organizational elements. These included diversity across CI members, in terms of experiences as well as disciplinary background and organizational type. The ‘mutually reinforcing activities’ component of CI reflects values consistent with network structures; networks indeed operate on the basis of recognising, and creating value from, mutual dependence (Keast et al., 2004). However, while the network literature generally discusses bringing different actors to the table (Keast et al., 2004; Podolny and Page, 1998), with particular reference to various sectors at times (e.g. Agranoff, 2006; Keast and Mandell, 2014; Mandell, 2001), CI puts particular emphasis on engaging parties that complement each other. In addition, robust organizational processes, an effective backbone organization and ongoing review and evaluation were all seen as vital. Our study showed that these elements provide CI with key advantages over other forms of interorganizational collaboration.

Second, our study enabled us to confirm the importance of the relational aspects of interorganizational collaboration, in concordance with extant research. The need to build positive relationships across the various actors and stakeholders in an initiative, the need to engage and consult with the local communities, the need for honesty and the mutual trust and respect that emerged over time were all seen as critical. This is an important feature of the CI framework, as trust is considered essential to the functioning of any network-based collaboration, including the CI initiatives featured here, as it consolidates member commitment to the initiative as well as their belief that their interests are equally valued.

Third, although CI places emphasis on the collective, our study has demonstrated the importance of effective leadership processes across each of the initiatives and that leadership styles are akin to servant leadership. This ongoing tension between self-interest on the part of (some) stakeholders and the overarching concern for ‘genuine CI’ was a key leadership challenge associated with this type of work. Finally, our findings suggested that successful CI requires a commitment to the longer-term time frames that allow the relational and measurement challenges inherent in this type of interorganizational collaboration to be addressed.

6. Conclusion

Our study allowed us to situate the CI framework within its rightful theoretical roots – network-based interorganizational collaboration – and to provide a theoretically robust study of CI processes that is missing in the field. We have found that the CI framework provides a useful method for implementing principles of network-based collaboration, rather than a fundamentally new ‘methodology’ or underlying philosophy. It provides guidance on how to exchange resources for greater impact, as well as a common language to work together better and thus focus on the process of collaboration rather than its outcomes. Our qualitative case studies show that CI supports existing theory as well as extant literature. However, while research in network structures and interorganizational collaboration points to the need for trust and commitment, CI brings further emphasis to relational conceptualisations of interorganizational collaboration. Indeed, the quality of the relational aspect of interorganizational collaboration is an important dimension of success, and the CI framework puts the development of relationships at the core and provides guidance on how to do so.

In this article, we have drawn on research in network structures and interorganizational collaboration in order to anchor our explanations of CI. While we note the increasing interest in network-based forms of collaboration, we argue that further clarification of how best to achieve successful collaborative relationships in the social purpose system was needed and suggest that the CI framework may provide such guidance. The CI framework shows much promise, but it is important to recognize that CI is in the early stages of application in Australia and therefore further research will be needed. For example, once this mechanism becomes more established, longitudinal studies would undoubtedly further our empirical understanding. Furthermore, situating CI within the network management literature will also extend its theoretical resonance. With the right leadership and a shared commitment to extra-organizational goals, this structured approach to network-based interorganizational collaboration has the potential to make a significant difference in cases of place-based disadvantage and possibly in other cases of complex and systemic societal problems.

Footnotes

Final transcript accepted 22 March 2017 by Peter Liesch (AE International Business).

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

1.

Calculated as AUD172 billion government contribution to social services (SCRGSP, 2013), plus AUD70 billion of not-for-profit income (Productivity Commission, 2010).

2.

Based on 2013 GDP for Australia of AUD1520.944bn (Australian Bureau of Statistics, 2012–2013).

3.

A conference organized by the Centre for Social Impact in collaboration with Social Leadership Australia in Sydney and Melbourne in February and July 2014, respectively.

4.

An initiative launched by the Centre for Social Impact and Social Leadership Australia to assist communities in coming together to address social issues by applying the CI framework (Collaboration for Impact, 2014).

References

Agranoff

(2006) Inside collaborative networks: Ten lessons for public managers. Public Administration Review 66: 56–65.

Anderson

(1999) Complexity theory and organization science. Organization Science 10: 216–232.

Australian Bureau of Statistics (2012–2013) Australian system of national accounts (cat. no. 5204.0). Available at: http://www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/5204.02012–13?OpenDocument (accessed 20 January 2014).

Australian Institute of Health and Welfare (AIHW) (2014) Australia’s Health 2014 (Australia’s Health Series No. 14, Cat. No. AUS 178). Canberra, ACT, Australia: AIHW.

Australian Treasury (2010) Australia to 2050: Future Challenges. Intergenerational Report 2010. Canberra, ACT, Australia: Commonwealth of Australia.

Bovaird

(2008) Emergent strategic management and planning mechanisms in complex adaptive systems. Public Management Review 10: 316–340.

Boyce

(2013) Collective Impact: Aligning organizational efforts for broader social change. Journal of the Academy of Nutrition and Dietetics 113: 495–497.

Braun

Clarke

(2006) Using thematic analysis in psychology. Qualitative Research in Psychology 3: 77–101.

Bryson

Crosby

Middleton Stone

(2006) The design and implementation of cross-sector collaborations: Propositions from the literature. Public Administration Review 66: 44–55.

10.

Cabaj

(n.d.) Collective Impact and shared measurement: Tough but necessary. Available at: http://tamarackcci.ca/resource-library/evaluation/collective-impact-shared-measurement (accessed 4 January 2015).

11.

Centre for Social Impact (2016) Australia’s social purpose market: Understanding funding flows and exploring implications. PricewaterhouseCoopers Australia. Available at: http://www.csi.edu.au/media/uploads/127037316_Social_Purpose_Market_v5.pdf (accessed 12 January 2017).

12.

Chambers

Edwards-Stuart

(2007) Leadership in the Social Economy. London: School for Social Entrepreneurs.

13.

Chau

(2014) Australian NFPs falling behind in transparency. Available at: http://probonoaustralia.com.au/news/2014/03/australian-nfps-falling-behind-in-transparency (accessed January 2015).

14.

Collaboration for Impact (2014) The search. Available at: http://www.collaborationforimpact.com/the-search (accessed 10 January 2014).

15.

Dal Molin

Masella

(2015) Networks in policy, management and governance: A comparative literature review. Journal of Management and Governance 20: 1–27.

16.

Daniel

(2012) Sampling Essentials. Thousand Oaks: SAGE.

17.

Department of Education and Training (2013) Best start: Supporting families and communities. Available at: http://www.education.vic.gov.au/about/programs/health/pages/beststart.aspx (accessed 12 February 2014).

18.

Department of Human Services (2012) Better futures, local solutions: Grants program guidelines. Available at: https://www.dss.gov.au/sites/default/files/files/grants-funding/Part_C_BFLS_Place_based_FMP_Services.doc (accessed 28 September 2015).

19.

Easterling

(2013) Getting to Collective Impact: How funders can contribute over the life course of the work. The Foundation Review 5: 67–83.

20.

Edmonson

Hecht

(2014) Defining quality Collective Impact. Stanford Social Innovation Review 12: 6–7.

21.

Gajda

(2004) Utilizing collaboration theory to evaluate strategic alliances. American Journal of Evaluation 25: 65–77.

22.

Gephart

Robert

(2004) From the editors: Qualitative research and the ‘Academy of Management Journal’. The Academy of Management Journal 47: 454–462.

23.

Hanleybrown

Kania

Kramer

(2012) Channeling change: Making Collective Impact work. Stanford Social Innovation Review. Available at: https://ssir.org/articles/entry/channeling_change_making_collective_impact_work (accessed 20 September 2015).

24.

Hardy

Phillips

Lawrence

(2003) Resources, knowledge and influence: The organizational effects of interorganizational collaboration. Journal of Management Studies 40: 321–347.

25.

Huxham

Macdonald

(1992) Introducing collaborative advantage: Achieving inter-organizational effectiveness through meta-strategy. Management Decision 30: 50–56.

26.

Irby

Boyle

(2014) Aligning Collective Impact initiatives. Stanford Social Innovation Review 12: 15–16.

27.

Johnson

(2012) Housing first ‘down under’. European Journal of Homelessness 6: 183–191.

28.

Kania

Kramer

(2011) Collective Impact. Stanford Social Innovation Review 9: 36–41.

29.

Kania

Kramer

(2013) Embracing emergence: How Collective Impact addresses complexity. Stanford Social Innovation Review 11: 1–7.

30.

Kania

Hanleybrown

Splansky Juster

(2014) Essential mindset shifts for Collective Impact. Stanford Social Innovation Review 12: 2–5.

31.

Keast

Mandell

(2014) The collaborative push: Moving beyond rhetoric and gaining evidence. Journal of Management and Governance 18: 9–28.

32.

Keast

Mandell

Brown

. (2004) Network structures: Working differently and changing expectations. Public Administration Review 64: 363–371.

33.

Lichtenstein

Plowman

(2009) The leadership of emergence: A complex systems leadership theory of emergence at successive organizational levels. The Leadership Quarterly 20: 617–630.

34.

Lichtenstein

Mandell

Brown

. (2006) Complexity Leadership Theory: Management Department Faculty Publications. Lincoln, NE: University of Nebraska-Lincoln.

35.

Locke

(2001) Grounded Theory in Management Research. London: SAGE.

36.

Mandell

(2001) Collaboration through network structures for community building efforts. National Civic Review 90: 279–287.

37.

Mandell

Keast

(2008) Evaluating the effectiveness of interorganizational relations through networks. Public Management Review 10: 715–731.

38.

Mandell

Steelman

(2003) Understanding what can be accomplished through interorganizational innovations. Public Management Review 5: 197–224.

39.

Minnery

Greenhalgh

(2007) Approaches to homelessness policy in Europe, the United States and Australia. Journal of Social Issues 63: 641–655.

40.

Muir

Edwards

Gray

. (2010) The national evaluation of the Communities for Children initiative. Family Matters 84: 35–42.

41.

New South Wales (NSW) Department of Education and Communities (2011) Education and communities: Connected communities strategy. Available at: https://www.det.nsw.edu.au/media/downloads/about-us/news-at-det/announcements/yr2012/connected-communities-strategy.pdf (accessed 26 September 2015).

42.

O’Neill

Graham

(2013) Collective Impact in South Australia. Available at: https://collectiveimpactaustralia.com/2013/07/19/collective-impact-in-south-australia/ (accessed 12 January 2014).

43.

Pinkse

Kolk

(2012) Addressing the climate change – Sustainable development nexus: The role of multistakeholder partnerships. Business and Society 51: 176–210.

44.

Podolny

Page

(1998) Network forms of organization. Annual Review of Sociology 24: 57–76.

45.

Powell

(1990) Neither market nor hierarchy: Network forms of organization. In: Staw

Cummings

(eds) Research in Organizational Behavior. Greenwich, CT: JAI Press, pp. 295–336.

46.

Pro Bono Australia (2014) Collective Impact. Available at: http://www.probonoaustralia.com.au/news/category/4567 (accessed 15 January 2015).

47.

Productivity Commission (2010) Contribution of the Not-for-Profit Sector (Research Report). Canberra, ACT, Australia: Productivity Commission.

48.

Provan

Lemaire

(2012) Core concepts and key ideas for understanding public sector organizational networks. Public Administration Review 72: 638–648.

49.

Provan

Milward

(2001) Do networks really work? A framework for evaluating public-sector organizational networks. Public Administration Review 61: 414–423.

50.

Reitan

(1998) Theories of interorganizational relations in the human services. Social Service Review 72: 285–309.

51.

Rittel

Webber

(1973) Dilemmas in a general theory of planning. Policy Sciences 4: 155–169.

52.

Rowley

(2012) Conducting research interviews. Management Research Review 35(3–4): 260–271.

53.

SCRGSP (Steering Committee for the Review of Government Service Provision). (2013) Report on Government Services 2013. Canberra: Productivity Commission.

54.

Sendjaya

Sarros

Santora

(2008) Defining and measuring servant leadership behaviour in organizations. Journal of Management Studies 45: 402–424.

55.

Strauss

Corbin

(1998) Basics of Qualitative Research: Techniques and Procedures for Developing Grounded Theory. Thousand Oaks: SAGE.

56.

Thompson

(2014) Collective Impact: Funder, heal thyself. Stanford Social Innovation Review 12: 8.

57.

Tracy

(2010) Qualitative quality: Eight ‘big-tent’ criteria for excellent qualitative research. Qualitative Inquiry 16: 837–851.

58.

Travers

(2012) Qualitative interviewing methods. In: Walters

(ed.) Social Research Methods. 3rd ed. South Melbourne, VIC, Australia: Oxford University Press, pp. 225–253.

59.

Van Beurden

Kia

Zask

. (2011) Making sense in a complex landscape: How the Cynefin framework from Complex Adaptive Systems Theory can inform health promotion practice. Health Promotion International 28: 73–83.

60.

Van der Haar

Hosking

(2004) Evaluating appreciative inquiry: A relational constructionist perspective. Human Relations 57: 1017–1036.

61.

Vinson

Rawsthorne

(2015) Dropping off the Edge 2015: Persistent Communal Disadvantage in Australia. Richmond, VIC, Australia: Jesuit Social Services/Catholic Social Services Australia.

62.

Weaver

(2013) Navigating Collective Impact. Available at: http://www.fsg.org/blog/navigating-collective-impact (accessed 7 December 2014).

63.

Weber

Khademian

(2008) Wicked problems, knowledge challenges, and collaborative capacity builders in network settings. Public Administration Review 68: 334–349.

64.

Williams

(2005) Cooperation by design: Structure and cooperation in interorganizational networks. Journal of Business Research 58: 223–231.

65.

Yin

(2013) Case Study Research: Design and Methods. Thousand Oaks: SAGE.