Abstract
As many of the challenges facing society are too complex to be addressed by single organizations working alone, nonprofit organizations are increasingly working in collaboration with public authorities. The governance of nonprofit–public collaborations is important for their effectiveness, yet it remains poorly understood. Drawing on case study research, this article examines and develops an extant conceptual model developed by Takahashi and Smutny that seeks to explain the formation and demise of nonprofit collaborations in terms of “collaborative windows” and the inability to adapt initial governance structures. The research finds that while initial governance structures are an important constraint on development, they can be adapted and changed. It also suggests that the development of collaborations is not only influenced by changes in the collaborative window but also by how key actors in the collaboration respond to important internal tensions.
Keywords
Introduction
Given the complexity of many social, environmental, and economic problems facing communities, nonprofit organizations are increasingly collaborating with public authorities. Nonprofit research has yet to address these developments satisfactorily. As noted by Stone and Sandfort (2009), “research on nonprofit organizations does not fully consider how the policy environment shapes organizational operation and performance and shapes how actors act strategically to advance their organizational interests” (p. 1070). Similarly, Renz (2006) discusses how research on nonprofit governance too often equates governance with what boards do. He advocates a new focus on inter-organizational governance processes that occur as organizations work together to address social problems. That is the focus of this article.
Within organizations, governance structures and processes are shaped by legal and regulatory requirements. The governance of collaborations is more elusive, as they are often established without any clear legal form or body in charge, and the relationships between partners are subject to change (Stone, Crosby, & Bryson, 2010). Public collaborations are often highly dynamic and even “chaotic” (Huxham & Vangen, 2000; Stone, Crosby, & Bryson, 2013), as they must respond to complex and changing policy environments and deal with internal paradoxes and tensions (Provan & Kenis, 2008; Saz-Carranza & Ospina, 2011). The governance structures of collaborations are therefore more fluid, changing in response to internal and external drivers, as well as to participants’ attempts to manage inherent tensions (Stone et al., 2010).
It is therefore important to gain a better understanding of how the governance of collaborations changes over time. As noted by Provan and Kenis (2008), “Systematic research on network evolution is needed, focusing especially on how the governance of public networks emerges (whether by mandate or choice), and how it changes over time” (p. 238). Similarly, Stone and colleagues (2013) suggest, “Scholars have much to learn about collaborations’ governance systems, especially with regard to their ability to adapt to these ever-changing contexts.” (p. 250). The research reported here is intended to address this gap.
Drawing on case study research, this article refines an existing conceptual model aimed at explaining the formation of and changes in the governance of nonprofit collaborations. It builds on work by Lober (1997) and by Takahashi and Smutny (2002). These authors extend Kingdon’s (1984, 2011) seminal work, which explains the formation of public policies in terms of the opening up of policy windows and the actions of policy entrepreneurs. They argue that the formation of collaborations can be similarly explained in terms of collaborative windows and the actions of collaborative entrepreneurs. Takahashi and Smutny extend the model even further to explain the short-lived nature of many collaborations. They suggest that “initial governance structures emanating from particular collaborative windows and entrepreneurs limit their adaptability and portend their short-term demise” (p. 181).
Takahashi and Smutny (2002) test their model using a case study of a collaboration between three community-based nonprofit organizations that provide services to people with HIV. While they conclude that the case supports their conceptual model, they call for additional research to “explore whether this conceptual argument holds for other partnership experiences and for other social service contexts” (p. 181). This study therefore assesses the utility of this model in a very different context, through a longitudinal case study of a public–nonprofit partnership aimed at neighborhood regeneration in deprived areas of one U.K. city. 1 The Head of the Regeneration Team (who we will call the team director for short), an employee of the City Council, initiated the collaboration and acted as a key coordinator. The case study traces the development of the collaboration from its inception, focusing particularly on an attempt by the team director to redesign its governance structure.
The article is structured as follows. In the next section, we define what we mean by partnership, collaboration, and the governance of inter-organizational collaborations. We then describe the initial conceptual framework developed by Takahashi and Smutny and how it builds on previous work by Lober and Kingdon. This is followed by a description of the research methodology employed and the analysis of the case study. We conclude that the model developed by Takahashi and Smutny provides only a partial explanation of how the collaboration was formed and developed, and that it is overly pessimistic. In particular, our research suggests that while initial governance structures pose important constraints on development, they can be adapted and changed. This suggests that change is influenced not only by changes in the collaborative window, but also by the responses of key actors in the collaboration to internal tensions. We also consider the implications for nonprofit organizations and community groups participating in collaborations with more powerful public authorities, as well as for nonprofit researchers.
Defining Terms
Many terms have been used to describe configurations of organizations that voluntarily agree to collaborate. This is confusing, and it impedes conceptual clarity (Cropper, Ebers, Huxham, & Ring, 2008). Clear definitions are therefore needed. We use the terms collaboration and partnership interchangeably to refer to formalized, joint-working arrangements between organizations that remain legally autonomous while engaging in ongoing, coordinated collective action to achieve outcomes that none of them could achieve on their own. In the literature, these arrangements are often conceptualized as “networks” (e.g., Agranoff & McGuire, 2001; O’Toole, 1997; Provan & Milward, 2001; Van Bortel, Mullins, & Rhodes, 2009).
The term governance is even more elusive. It is rooted in a Latin word meaning to steer or give direction, but it is used in several different ways, both within and across disciplines (Klijn, 2008; Kooiman, 1999). One useful way of distinguishing between different usages involves the level of analysis at which the concept is applied (Kooiman, 1999).
In this article, we focus on the inter-organizational level, examining how collaborations between organizations are governed. In developing a definition for this level of analysis, Stone et al. (2010) draw on a definition of governance developed by Hughes (2010): “Governance is about running organizations, about steering as in the original derivation, how to organize, and how to set procedures for an organization to be run” (p. 88). They note that governance in collaborations is more “elusive” than organizational or corporate governance is, given the absence of a “clear entity in charge.” Governance is nevertheless critical “for a partnership to survive and produce results” (p. 312). As suggested by Stone et al. (2010), governance is about “making collective decisions about important issues, including the purpose of collective action, strategies for achieving purpose and oversight and accountability mechanisms” (p. 310).
Provan and Kenis (2008) argue that the governance of networks is important for their effectiveness, although this topic has been neglected in research. They state, a focus on governance involves the use of institutions and structures of authority and collaboration to allocate resources and to coordinate and control joint action across the network as a whole. . . . Unlike organizations networks must be governed without the benefit of hierarchy or ownership. (Provan & Kenis, 2008, pp. 231-232)
They distinguish governance from issues of a more operational nature: “These interactions are distinct from operational links, which are often dyad-based including referral, sharing of information, and joint programs” (Provan & Kenis, 2008, p. 231).
Building on these definitions, we propose that the governance of collaborations entails the design and use of a structure and processes that enable actors to set the overall direction of the collaboration, and that coordinate and allocate resources for the collaboration as a whole and account for its activities.
Collaborative Windows, Collaborative Entrepreneurs, and Governance Structures
A crucial step in forming a new collaboration is to decide how the partnership will be governed. Drawing on a review of research in the field, Provan and Kenis (2008) suggest that the effectiveness of any structure depends on various internal and external contingencies. Takahashi and Smutny (2002) note that such contingencies are often viewed in general terms rather than as conditions operating in particular times and places (p. 167).
It is thus important to consider how temporal and spatial conditions influence the way collaborations are formed, governed, and developed. Lober (1997) and Takahashi and Smutny (2002) develop and refine Kingdon’s (1984, 2011) theory of policy windows to explain the formation of collaborations. Kingdon explains how government policies are formed, proposing that three largely independent, temporal streams run through the political system: a problem stream, a policy stream, and a political stream. The problem stream consists of issues or situations that interest groups identify as “problems” to be addressed. The policy stream consists of policy proposals advocated by various groups. The political stream consists of various influences on the political system (e.g., public opinion, the media, and elections). Kingdon argues that whenever these different streams converge, a “policy window” opens, presenting an opportunity to adopt new policies. For this to happen, however, policy entrepreneurs (either individuals or groups) must recognize that the window has opened and have the skills to exploit the opportunity and gain support for their proposals.
Lober (1997) develops Kingdon’s model to explain the formation of inter-organizational collaborations. Lober retains the problem stream, but adapts Kingdon’s other two streams. The policy stream becomes possible “solutions” to the problem rather than policies. The political stream is broadened to include social and economic factors affecting the issues being addressed (hereafter, the PSE stream). Lober adds a fourth stream—the organizational stream—that encompasses changes in organizational and industry behavior regarding the issues being addressed. According to Lober, convergence in these four streams can create the conditions for forming a collaboration (i.e., a “collaborative” window rather than policy window). For this to happen, however, the opportunity must be exploited by a collaborative entrepreneur. For Lober, as well as for Takahashi and Smutny, the collaborative entrepreneur resembles the policy entrepreneur. Collaborative entrepreneurs act as the catalyst for forming collaborations by working across organizational boundaries to join organizations and identify solutions to problems. There are subtle differences. In Kingdon’s model, policy entrepreneurs are passionate advocates for their favored policies, possibly making the system more amenable to the policy, pushing the policy, and coupling the various streams to help create policy windows that they can exploit. The collaborative entrepreneurs described by Lober and by Takahashi and Smutny do not appear to be so active in creating the conditions for collaboration, being more likely to exploit collaborative windows once they emerge. Lober describes the role of collaborative entrepreneurs only in general terms, as having “conceived of the project, selected participants and lobbied them to participate” (p. 19). Similarly, Takahashi and Smutny (2002) suggest that collaborative entrepreneurs must recognize the collaborative window, tap geographically defined networks to engage suitable partners, and bring them together to form the collaboration (pp. 168-169). The role of the collaborative entrepreneurs thus resembles that of the collaborative convener or coordinator who has long been recognized in the literature on inter-organizational collaboration (e.g., Gray, 1995).
Takahashi and Smutny (2002) adopt and refine Lober’s four-stream model of collaborative formation. They add a “spatial dimension” to the idea of temporal streams, which constrains the nature of collaborative windows in three ways. First, the partners’ areas of operation are subject to spatial restrictions. Second, this restricts the organizations and networks to which collaborative entrepreneurs have access. Third, spatial constraints affect factors that facilitate collaboration (e.g., opportunities to meet face-to-face to build trust).
Takahashi and Smutny (2002) extend their model beyond the formation stage to include the operational stage of collaboration. They argue that collaborative entrepreneurs “initiate alliances among partners using specific initial governance structures that fit with the participants and the features of the collaborative window” (Takahashi & Smutny, 2002, p. 169). They further suggest that this initial governance structure seriously constrains the future adaptability and resilience of the partnership, because “organizational inertia” and the “time-consuming nature of collaborative governance” make these structures resistant to change. They suggest that collaborative entrepreneurs and other partners in the collaboration may not “have the skills to maintain, sustain, or adapt the collaborative partnership’s initial governance structure to changing temporal and spatial conditions after the collaborative window closes” (Takahashi & Smutny, 2002, p. 169). They therefore propose that features of a collaboration’s formation contain the seeds for its demise in a relatively short time, as initial governance structures fail to adapt. The main features of Takahashi and Smutny’s conceptual model are shown in Figure 1.
Conceptual model of the formation and development of collaborations (Takahashi & Smutny, 2002).
In our study, we assess the applicability of the model developed by Takahashi and Smutny to explain the formation and development of a collaboration between public agencies and various nonprofit and community organizations aimed at neighborhood regeneration in one U.K. city. 2 The City Council initiated the collaboration in 2009 as a means of implementing its neighborhood-regeneration strategy. (The City Council is the principal tier of local government providing a range of services including housing, planning, education, social services, waste collection and management, and transport.) This strategy and the subsequent collaboration aimed to reduce the gap between the city’s most and least affluent neighborhoods by regenerating the most deprived neighborhoods. A framework based on Kingdon’s theory of policy windows is particularly appropriate, as this collaboration arose from a public policy initiative.
Method
This study follows a longitudinal, case study research design (Tsang, 2013; Yin, 1989). We gained access to the Head of the Regeneration Team at the City Council through a nonprofit organization, which we shall call Neighbourhood Action (NA), which was familiar with our research. The City Council contracted with NA to undertake neighborhood development and support the neighborhood-regeneration program. We first met with the person who was responsible for the neighborhood-regeneration program in July 2011. At this point, this team director was considering how to redesign the governance structures and processes of the collaboration. We agreed that the research would focus on capturing the challenges, successes, and experiences concerning the governance of the collaboration, as a tool for considering the redesign of the governance arrangements. We also emphasized our desire to examine changes in governance structures and processes over time.
Data were collected in two phases. The first phase took place in 2011, focusing primarily on the governance of the collaboration as a whole. The second phase was carried out in 2012, focusing on collaboration in one of the neighborhoods, thus providing insights into the development of the collaboration at this level. Additional information was also gathered on the development of the governance of the collaboration as a whole.
Data were collected in several ways including semi-structured interviews, participation in and observation of workshops, observation of meetings, documentation, informal conversations, emails, and meetings with the regeneration team director, the Director and Deputy Director of NA, and other key participants
In Phase 1, we conducted 10 interviews, as indicated in Table 1. The number of interviews was determined by pragmatic considerations. The team director wanted to use findings from this phase to inform deliberations about redesigning the governance structure in Autumn 2011. This was what was practical within this period. We also collected and reviewed documentation relating to the collaboration, including the City Council’s neighborhood-regeneration strategy and its collaboration-initiation document, a City Council review and an external review of the collaboration (both from 2010), and regeneration plans for two neighborhoods in the collaboration. We analyzed the documentation to build a chronological description for the case (Yin, 1989). This also allowed supplementation and triangulation with data from interviews. We submitted a report on this phase of the research to the team director and the group convened to discuss redesigning the governance structure, and it was circulated to interviewees.
Participants in Interviews in Phase 1 of the Research.
Access proved more challenging in Phase 2. The team director was concerned about timing. In light of anticipated cuts and restructuring, she saw no further benefits for the Council, expressing concerns that the research might raise false expectations. We emphasized our objective of examining how the collaboration was developing at the neighborhood level. A second phase was eventually agreed in July 2012.
The second phase of research involved eight interviews, focusing on one neighborhood (see Table 2). Seven of the interviewees were primary actors at the neighborhood level. We also interviewed the Deputy Director of NA, to gather a different perspective on how the collaboration was developing, in addition to gaining access to relevant community-development workers in the neighborhood. We collected a full set of minutes from the neighborhood steering group’s (SG) meetings, thus providing further information about the development of collaboration in the neighborhood and triangulation with the interview data. Findings from this phase were also fed back through a meeting of the steering group.
Participants in the Interviews in Phase 2 of the Research.
Data from the case were analyzed using a combination of pattern matching and explanation building (Yin, 1989, pp. 109-115). Data were matched against the theoretical model proposed by Takahashi and Smutny. For aspects of the case that did not seem adequately explained by the model, we engaged in an iterative process of explanation building, looking for themes that might provide further explanations and comparing them with other data from the case (Yin, 1989, pp. 113-115).
Analysis of the Case
The analysis of the case study is divided into three parts. The first analyzes the formation of the collaboration and the establishment of an initial governance structure. The second describes how the governance structure and scope of the collaboration changed from its formation in 2009 to the end of the study in Autumn 2012. The third part analyzes these changes. The model developed by Takahashi and Smutny suggests that organizational inertia and the collaborative entrepreneur’s lack of skill lead to a failure to adapt the initial governance structure to changes in the collaborative window and, ultimately, the demise of the collaboration. Evidence from this case suggests that this does not provide a complete explanation of what happens. We propose additional explanatory factors concerning tensions that need to be managed in collaborations.
Formation of the Collaboration and Initial Governance Structure
The collaboration to implement the City Council’s neighborhood-regeneration strategy was formed when a collaborative window opened due to the convergence of the problem, solution, organizational, and political, social, and economic (PSE) streams.
The problem stream: National and local governments in the United Kingdom have long recognized that some neighborhoods across the country suffer from multiple deprivations. The U.K. Government’s Department for Communities and Local Government (DCLG) and its predecessors have published statistics on local measures of deprivation in England since the 1970s. A 2011 report notes that 5 million people in England were living in the most deprived areas in 2008, and that 98% of those deprived areas were urban (DCLG, 2011a, p. 1). Moreover, 56% of local authorities contain at least one area among the 10% most deprived areas in the country. Although the city where the research took place is relatively affluent, it also contains several deprived areas. According to the City Council’s 2008 neighborhood-regeneration strategy, these deprived areas were growing in scale and intensity, and were likely to deteriorate further due to rising unemployment.
The PSE stream contained several strands favorable to neighborhood regeneration. In 2001, the U.K. government established a National Strategy for Neighbourhood Renewal with the vision that “within 10 to 20 years time no one should be disadvantaged by where they live” (as quoted in Amion Consulting, 2010, p. 6). The strategy had two main goals: to reduce various indicators of deprivation in the poorest areas and to narrow the gap between the most deprived areas and other areas in the country. Unlike previous strategies, the national strategy emphasized “locally-determined measures” and the role that local public bodies could play in reversing decline (Amion Consulting, 2010, p. 6). A 2010 evaluation of the strategy notes that, despite some narrowing of the gap between the most and least deprived neighborhoods, “the most deprived neighbourhoods were still a long way behind,” and deteriorating on several indicators, including violent crime and mortality (Amion Consulting, 2010, p. 40). The evaluation also notes that these changes took place during favorable economic times and that the economic slowdown was beginning to have a negative effect. During this period, greater academic and media attention had been paid to the growing negative impact of inequality on health and other indicators of well-being (e.g., Wilkinson & Pickett, 2009).
In another important trend in the early 2000s, the New Labour Government increasingly emphasized the need for government partnerships to tackle social and economic problems in society (Newman, 2001; Osborne, 2010). At the local level, public authorities were encouraged to establish cross-sector partnerships to address a wide range of problems.
The policy/solution stream within the City Council was influenced by various complementary policies. It had a sustainable community strategy, which emphasized the need to tackle problem areas in the city. This emphasized the importance of building the capacity of neighborhoods and helping them develop a strong sense of community. The Council’s growth strategy also recognized the need to support community and economic infrastructure. The Council’s neighborhood-regeneration strategy was expected to provide the neighborhood focus for these broader strategies. Nineteen neighborhoods were identified using the Index of Multiple Deprivation. The plan called for forming an initial pilot group with three neighborhoods, extending the program each year to include other neighborhoods. The strategy hinged on the engagement of communities and other partners from the public, private, and nonprofit sectors, emphasizing that top-down neighborhood regeneration would not work. It envisaged that partners from different sectors would work with residents in each neighborhood to form neighborhood steering groups to develop their own neighborhood plans to guide new and existing investments and services. It suggested that local parish and town councils would assume leadership roles in neighborhoods in their areas. (The city was divided into several local parishes and towns with elected councils. These bodies have a consultative role in planning, and some bear responsibility for community centers or other local facilities, albeit with relatively few resources.) The plan called for the City Council to provide resources to encourage community engagement and help neighborhoods develop Neighbourhood Plans.
The organizational stream: A wide range of organizations operating in the various neighborhoods of the city influence the social, economic, and physical well-being of each area. The exact composition varies by neighborhood, but typically includes the City Council, parish and town councils, the police, schools, the city college, health services, nonprofit organizations, local businesses, and community groups, including residents’ associations and community centers. Many of these organizations responded to a City Council consultation regarding the development of the neighborhood-regeneration strategy.
The City Council’s director of regeneration acted as the collaborative entrepreneur, forging the neighborhood-regeneration strategy and coordinating its implementation. Although forming the collaboration was part of the director’s job, we identify important similarities with the entrepreneurial role, as characterized by Lober and by Takahashi and Smutny. As noted previously, however, this case differs somewhat from Kingdon’s policy entrepreneur. The director used her position in the corporate directorate, which includes the Council’s senior executives, to gain access to many relevant contacts within the Council, as well as with other public bodies, local nonprofit organizations, and community groups. She used these networks to identify and engage suitable people to join the collaboration. The collaboration had a limited budget (beyond existing staff or services provided by the various partners) to support the work of the regeneration program. Also, she acted entrepreneurially to generate new resources, engaging with NA to encourage them to align their community-development work (funded by the City Council) to help define and achieve the social goals of the neighborhood-regeneration program.
The neighborhood-regeneration program was initially established with a three-tier governance structure designed by the team director. This consisted of neighborhood steering groups in each neighborhood, accountable to a performance group, which was accountable to a sponsor group, with service and support from the City Council’s regeneration team. The original purpose and composition of these bodies is presented in Table 3. The regeneration team also commissioned a nonprofit organization, part of a larger national federation, whose mission is to develop sustainable communities, to help the neighborhood steering groups develop the neighborhood plans.
Initial Governance Structure of the Collaboration.
Changes to the Governance and Scope of the Collaboration
In contrast to the case analyzed by Takahashi and Smutny, this case involved major changes to the governance structure of the collaboration over time (illustrated in Figure 2). The first change occurred in 2010, when the sponsor group stopped meeting. The team director considered it no longer necessary, as the focus had shifted toward delivering plans and achieving outcomes: “having those two tiers was proving to be unnecessary” and “it had served its purpose towards the end . . . [it] had become much more performance managed . . . are we delivering outcomes.” The performance group and each of the neighborhood steering groups continued with quarterly (instead of monthly) performance group meetings. The program was also expanded at this time to cover three additional neighborhoods.
Changes in the governance structure of the neighborhood-regeneration collaboration resistance to change.
The governance structure changed again in Spring 2011, when the director expressed concerns about the efficiency and effectiveness of the performance group deciding to discontinue it and review the governance structure. The director was particularly concerned that the group’s large size was transforming it into an unfocussed “talking shop.” A further change occurred when NA assumed the role of supporting the neighborhood steering groups. The director’s advocacy of their role in the collaboration helped NA to renegotiate resources with the City Council for their activities.
In July 2011, NA convened a meeting of the program partners to share ideas about developing economic, physical, social, and human capital in the neighborhoods. At this meeting, the director announced that the governance structure would be revised in the autumn. Coincidentally, the first phase of the research also began at this time, and the director hoped that it would inform the review.
In September 2011, the director convened a small group (consisting largely of senior council officers and the research team) to help consider options for revising the governance structure. Perhaps significantly, people from the neighborhoods and parish/town councils were not invited to these meetings. After two meetings and much discussion, the director decided that it would be more effective not to replace the performance group. In effect, the regeneration team was left to coordinate the work across the neighborhoods, with the neighborhood steering groups planning and coordinating work at the neighborhood level.
By Autumn 2012, the second phase of the research revealed that the neighborhood steering groups had continued, with varying levels of activity, although the program had not been extended to further neighborhoods as originally planned. Membership in the steering groups was open to residents, nonprofit organizations, local councilors, and relevant staff from public agencies working in the neighborhood (e.g., police, schools, city and town councils). Meeting attendance fluctuated widely, declining over time, particularly by residents. At this time, much of the regeneration team’s efforts focused on another neighborhood, where a major housing-regeneration program was being planned. Further loss of impetus was feared without the strong backing of the team.
Since its inception, the neighborhood steering group we studied had developed a range of initiatives, including improved landscaping and play facilities on the estate, a community newsletter, increased awareness of training and health initiatives, and the organization of various community events. Another indirect benefit was that the collaboration enabled relationships to develop between steering group members and public officials, thus allowing people to know whom to contact when problems arose. Because of the limited financial resources available to support the steering groups, most changes were on a small scale, often emerging from improvements in the use or coordination of existing services. Residents were frustrated that more costly physical-regeneration activities had not been possible.
Collaborative Windows, Internal Tensions, and Change
Takahashi and Smutny (2002) suggest that the demise of collaborations can be explained by failures to adapt to changes in the collaborative window, due to the resistance of initial governance structures to change and the collaborative entrepreneurs’ lack of skills to manage these changes. The evidence presented below suggests that this is only a partial explanation and that additional factors influenced the change process.
Changes in the Collaborative Window
Several important changes in the collaborative window occurred in the period 2009-2012. In particular, changes in the PSE stream had important implications for public agencies involved in the collaboration, as well as for the collaborative entrepreneur’s priorities.
The collaboration started just after the global financial crisis of 2008. To cut the public spending deficit, the new coalition government (elected in 2010) launched a program of austerity measures. It also introduced other changes with a potential impact on neighborhood regeneration. New powers were given to local communities, allowing them a greater say in developments in their areas (DCLG, 2011b). The government also relaxed previous financial restrictions on councils, which had effectively prevented them from building new public housing, thus encouraging them to resume these efforts.
These changes in the PSE stream strongly influenced changes in the organizational and solution streams, suggesting that the different streams are less independent than suggested by Kingdon, and subsequently by Lober and by Takahashi and Smutny. In the organizational stream, the City Council, like many other public bodies, had to consider how it could manage with fewer resources. This led to cuts in the regeneration team, which was transferred from a central directorate in the City Council to a directorate responsible for housing. Other public agencies in the collaboration were also facing cuts or frozen budgets, which influenced their commitment to the collaboration. One public partner noted about the collaboration: “For me, it is very small in my head. I’m trying to save all my services from being decommissioned at the moment, that’s the biggest problem for me.” The lack of resources also meant that some problems that neighborhood plans had identified as priorities could not be tackled.
Changes in government policy also influenced the local-level solution stream. Given the new freedoms granted to councils, the City Council’s Housing Directorate pursued a major new public–private partnership in 2012 to redevelop one of the neighborhood-regeneration areas and to pilot the implementation of the new neighborhood-planning policy (DCLG, 2011b). Much of the time of the regeneration team was reallocated from the collaboration to these new priorities. Nevertheless commitment continued within the neighborhood we studied to maintain the collaboration, although there was concern that the reduced commitment from the City Council might threaten their future. The original plan to extend the neighborhood-regeneration program to additional neighborhoods was dropped.
While changes in the collaborative window help explain some of the longer term changes in the collaboration, they do not explain the changes in governance structures occurring in the period 2009-2012, particularly regarding the collaborative entrepreneur’s focus and the decline in commitment of other public sector partners. Understanding these changes requires examining internal tensions and challenges that arose, particularly with respect to the performance group. These tensions transcend organizational inertia and lack of skill.
Internal Tensions and Challenges
The performance group experienced several inter-related tensions reflecting the different expectations and goals of participants (thus hindering the group’s effectiveness), as well as tension between efficiency and inclusiveness.
Existing research suggests that bringing together organizations with different goals and ways of working can generate problems in formulating common goals, possibly impairing effectiveness and generating collaborative inertia (Vangen & Huxham, 2012). Participants had differing expectations about the role of the performance group and its potential contribution to what they perceived as the goals of the collaboration. This was exacerbated in the early stages of the group, when the focus was inevitably on overseeing the establishment of neighborhood steering groups and the development of neighborhood plans: I think when it started out, it was very much process led, very much about managing risk, there was an awful lot of risk management talked about, and it was really about delivering on City Council policy on regeneration.
Over time, this changed, and others came to see it as a monitoring group and place for new ideas: “Performance monitoring group is everything that they’ve decided to do. They monitor and obviously fresh ideas or thoughts or whatever else can be brought in there.” Other participants considered the discussions far too operational, preferring to discuss more strategic issues: “I think the intention was to be strategic, but they weren’t; they were still operational.”
These disagreements manifested themselves in different views regarding the value of the neighborhood plans. Some perceived them as “hugely important” and as something that “should guide the regeneration in the way that residents wish, within the Strategic Policy of [the City Council] and other relevant legislation.” Others saw them as addressing short-term issues: “[They] were just about dog mess and things like that . . . it’s not transformation . . . there was no sustainability . . . so actually it was just purely cosmetic.” Some even questioned the entire approach to regeneration: The approach is very much this is the problem let’s get a plan and get on with it . . . if you were to manage something from a [partner] perspective you’d . . . do more around planning, thinking things through and deciding what might be best.
There were also tensions over priorities. Some performance group participants perceived a tension between the priorities of the City Council and those of the residents. As one participant noted, “It felt very ‘us and them’ in terms of the parishes and the voluntary and community sector and the council.” This led some participants to feel trapped between these two positions: We’re very much about delivering on behalf of the residents, so for us it’s about championing their cause to . . . [the] Council, and actually getting [them] to recognize that their desirable outcomes may not be the same as the residents, and somehow for them to marry those two together. It’s almost like I’m piggy in the middle.
This also generated concerns about the performance-management processes established by the Council: “You start trying to word proposals in a way that ticks the boxes” (i.e., doing and reporting things to meet criteria specified by the Council rather than what is actually needed or happening).
From a structural perspective, another fundamental tension was between efficiency and inclusiveness, which is common to collaborations (Provan & Kenis, 2008). The performance group was a relatively inclusive, large group comprising officers from the main public agencies, nonprofit organizations, local parish and town councilors, and representatives from the neighborhood steering groups. This raised concerns about the group’s efficiency and effectiveness in making decisions. Another concern was that the performance group tended to become mired in paperwork and operational detail: all it had was reams and reams of paper . . . just an absolute nightmare . . . if you are not an operational manager with the absolute evidence at your finger tips you can’t get down to that level of detail.
There were also concerns that people from different neighborhoods were not interested in what was happening in other neighborhoods. Combined with the diversity of goals and priorities, this made the performance group appear as a talking shop with no clear sense of purpose. The team director commented, “It’s quite big and that makes it much more conversational . . . I just want to get in there, do it and make the decisions and go.”
Interestingly, an earlier independent external review of the neighborhood-regeneration program in 2010 concluded, “The Performance Group was not highly valued by participants; it was described as ‘ponderous,’ ‘bureaucratic,’ and the risk-based approach to analysis was disliked and rarely understood, even by some senior attendees.” Such frustrations over efficiency and effectiveness led the team director to discontinue the meetings and look for possible alternative governance structures. Despite detailed consideration of alternatives, concerns over the difficulties and costs of collaboration at this level, particularly if more neighborhoods were added, led the director to decide to shift toward a lead-organization structure, in which the regeneration team coordinated the work across neighborhoods. Shortly thereafter, changes in the collaborative window (as discussed above) altered the priorities of the team. The neighbourhood steering groups thus continued with little support from the team, and the program was not expanded to new neighborhoods as originally planned.
Discussion and Conclusion
Nonprofit organizations are increasingly collaborating with public authorities to address pressing social problems. Nonprofit research has yet to reflect this new reality adequately (Renz, 2006; Stone & Sandfort, 2009). It is therefore becoming increasingly important to understand the circumstances under which these collaborations are formed and how they develop, as well as their likely consequences for nonprofit organizations.
Building on the work of Kingdon (1984, 2011) and Lober (1997), Takahashi and Smutny (2002) propose a model to explain both the formation and demise of collaborations. They tested this model in a single case study, calling for further studies to replicate the findings in other settings. We have attempted to replicate their findings in the case of a public–nonprofit partnership concerned with neighborhood regeneration. Our results support important aspects of the model. They confirm that convergence in the PSE, problem, solution, and organizational streams opened a collaborative window, which the collaborative entrepreneur exploited to form the collaboration. Nevertheless, our results suggest that the model developed by Takahashi and Smutny requires further refinement to understand collaboration’s development and the changes in its governance.
First, our findings suggest that the four streams comprising the collaborative window are not independent, as stated in the previous models, but interdependent. In particular, once the collaboration was formed, changes in the political, social, and economic stream influenced both the solution and organizational streams. As Kingdon (2011) acknowledges, the degree of independence of the various streams has been a topic of debate in the literature on policy windows (pp. 227-229). Central government policy changes in the PSE stream influenced the solution stream at the local level, where the new freedoms granted to councils to start building new public housing (i.e., rental housing owned by councils) particularly influenced the City Council’s approach to neighborhood regeneration. This generated new priorities for the regeneration team. Funding cuts also led to changes in the organizational stream, with public bodies involved in the collaboration having to make cuts, decommission services, and restructure. This reduced the priority of the collaboration for some of the public partners, constraining the resources available for deployment in the neighborhoods, which in turn raised concerns that this would undermine the neighborhood steering groups and resident involvement.
Second, referring to the short-lived nature of many collaborations, Takahashi and Smutny (2002) suggest that the inability to adapt initial governance structures results in their demise. They attribute this inability to adapt to organizational inertia and the collaborative entrepreneur’s lack of skills to maintain and adapt the initial governance structure. Our research suggests that this model is overly pessimistic about the ability of collaborations to change their governance structures. Our results suggest that the collaborative entrepreneur was able to change the initial governance structure, although it was not easy. Changes in the four streams, which influenced the priorities and commitment of different partners to the collaboration and the resources available to achieve its plans, were more important to the collaboration’s long-term future than were difficulties encountered in changing how it was governed.
Third, and perhaps most importantly, our findings suggest that collaborations also face important internal tensions and emergent challenges that must be addressed by those who govern and manage the collaboration. These tensions might also lead to changes in governance structures and processes. In particular, we have identified two key tensions. The first concerns the tension between “efficiency and inclusiveness” (Provan & Kenis, 2008). This led the collaborative entrepreneur to change the collaboration’s governance structure at the city level to improve efficiency, constituting a shift away from a participative, shared governance model toward a lead-organization model (Provan & Kenis, 2008). The second tension concerns goals and ways of working, which led some partners to reduce their commitment or withdraw from the collaboration. It also complicated agreements concerning the role of the performance group. The identification of these internal tensions is particularly important, as it suggests the need to extend the conceptual model developed by Takahashi and Smutny to connect with other important strands of theory about collaborations. These theories suggest that the dynamics of collaborations are driven by important tensions, which must be managed if the collaboration is to be successful (Huxham & Vangen, 2005; Provan & Kenis, 2008; Saz-Carranza & Ospina, 2011; Stone et al., 2010). Figure 3 outlines the main features of this extended conceptual model.
Refined conceptual model of the formation and development of collaborations.
For nonprofit organizations and community groups, understanding what lies behind the dynamic nature of collaborations and their governance arrangements might help them advance their goals when collaborating with more powerful public authorities. Our findings suggest that local councils are likely to play a dominant role in deciding the composition, scope, and governance structure of collaborations that they initiate. This ability to establish the structure, terms, and agenda for collaborations can give councils considerable power. As various other studies have suggested, these collaborations or networks are unlikely to constitute partnerships of equals (Entwistle, Bristow, Hines, Donaldson, & Martin, 2007; Lowndes & Skelcher, 1998).
Nonprofit organizations should also recognize that local governments exist within a complex and changing national policy and economic environment. As reflected in our findings, this can lead to changes in the “collaborative window” at the local level, changing the priorities of public partners, perhaps altering their commitment to the collaboration and even leading to its decline or demise. Nonprofit organizations are advised to be aware of these potential dynamics and risks when engaging in public–nonprofit collaborations.
These possible disadvantages should be weighed against potential benefits. In our case study, the collaboration did open new opportunities for residents, community groups and local nonprofit organizations to be involved in shaping plans for regeneration activities in their neighborhoods, and it provided access to new resources for realizing improvements. Perhaps inevitably, resource constraints meant that some larger projects were not possible, but there were tangible benefits. Less obvious benefits were also realized, as the collaboration enabled relationships with council officers and other service providers to develop, which could be used to address other problems and issues that arose.
Finally, our findings have important implications for researchers interested in the governance and effectiveness of nonprofit organizations. They suggest that the focus of research might usefully shift from the single organization to encompass the inter-organizational level and the collaborations and alliances in which nonprofit organizations engage to achieve social change. Although organizations and groups may deliver services individually, they are often coordinated and resourced through a wider network of relationships (Renz, 2006). Governance and leadership processes thus do not occur only at the organizational level (e.g., within an organization’s board), but at multiple levels. As this research shows, however, governance at the inter-organizational level is often highly dynamic and fraught with tensions. It can therefore be particularly challenging to achieve the potential advantages of collaboration, often resulting in collaborative inertia or demise (Huxham & Vangen, 2005). One important challenge for nonprofit researchers is therefore to gain additional insight into the formation, development, and effectiveness of public–nonprofit collaborations, and how the tensions and challenges they face can be overcome. We believe that the model resulting from this research is a step in this direction.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors gratefully acknowledge funding from the Economic and Social Research Council (Grant RES-073-27-0033), which part funded a Management and Business Development Fellowship for John Paul Hayes that enabled him to work on the research reported here.