Transitioning to network governance in the social services sector: the blending and segregating mechanisms in the hybridization process

ABSTRACT

The literature on network governance in social services comprises static analyses and has not investigated the complications of implementation. This study describes a case in Hong Kong in which the public implementation of networked governance in the social services sector began chaotically due to conflict among network actors, with clashes between values associated with the new network governance model and values inherited from earlier, entrenched models. The study highlights how a set of both blending and segregating mechanisms emerged during the governance transition, respectively facilitating and also preventing the hybridization of different logics towards a coherent set of practices.

KEYWORDS:

• Network governance
• implementation
• hybridization
• public service delivery network

Introduction

Over the past few decades, public administration in the social services has undergone major reforms. Starting in the 1980s, multiple governments have introduced the New Public Management (NPM) reform, which aims to shift the social services management that was once pertinent to the public administration (PA) style into more business-like arrangements (Simonet 2015; Laffin 2016; Moberg 2017). Whereas the NPM reform has swept through multiple social welfare domains, including those of health, education, housing, and many others (Simonet 2015; Møller and Skedsmo 2013; Sprigings 2002), empirical work has suggested that the reform only yielded varied successes across regions (Haque 2007; Moberg 2017). There have also been growing criticisms of the NPM reform due to its overreliance on market logic to solve social issues and its failure to capture a collaborative approach in value creation (Hartley 2005; Colin, Osborne, and Bond 2014). Gradually, commentators have argued that the NPM discourse has lost its appeal, and they have advocated a new paradigm: New Public Governance (NPG).

Rather than emphasize economic efficiency, the central notion of the NPG anchors on inter-organizational relationships (Klijn and Koppenjan 2012). Based upon network theory, the NPG refers to a set of principles for reconfiguring the power relationships among public agencies, private organizations, and community groups, in the pursuit of public value (Healey et al. 2017; Jessop 1995; Torfing and Triantafillou 2013; Bovaird 2005). Theorists portray the NPG as an alternative public sector management discourse that is distinct from the NPM and the PA (Osborne 2006). Network governance is the central notion in the NPG, and it prescribes a networked mode of organizing, with emphasis on the participation of non-public actors and a non-hierarchical collaborative approach in policymaking and service implementation (VanKersbergen and Van Waarden 2004; Rhodes 1997).

A growing body of literature has tapped into network governance, or NPG, in contemporary social welfare systems (Yanto, Shang, and Roy 2021; Barraket et al. 2021; Pestoff and Brandsen 2010; Barraket, Keast, and Furneaux 2015). Proponents have underscored the beneficial prospects emanating from this new mode of governance in the social services sector (Barraket et al. 2021). However, the majority of the existing work has been static analyses in which little attention was paid to the process of governance transition. Prior research, on the other hand, has illustrated governance transitions as complex processes (Iacovino, Barsanti, and Cinquini 2017; Wiesel and Modell 2014), and successful implementation of a new mode of governance may not always take place (Wegner and Verschoore 2021; Carmine, Nasi, and Rivenbark 2021). This complexity calls for a careful, process-based exploration to reveal the process of governance reform in the social services sector and to evaluate the extent to which the prospects of a newly implemented NPG mode of governance can be effective.

This study aimed to fill the knowledge gaps by examining a case illustrating how the government in Hong Kong has extended the concept of social innovation into policy action –– the Social Innovation and Entrepreneurial Development Fund (SIE Fund) –– and how that effort has contributed to the evolving NPG-inspired governance practices in the local social services sector. Using a case study paradigm, this study prospectively tracked this public effort in co-working with intermediaries to re-configure the governance of the local social services sector into a network mode of organization. Taking the theoretical lens that governance transition is a complex process (Wiesel and Modell 2014), this study contributes to the existing literature by outlining how sets of blending and segregating mechanisms interacted in the hybridization of governance logics over the course of the reform. Collectively, the study’s findings generate new theoretical and practical insights that can inform the public governance transition in the context of social services sector management.

Before elaborating on the study’s design, context, and findings, the next section provides a brief overview of the different sets of governance logics and the theoretical underpinnings of this work. The subsequent sections discuss the key findings and theoretical and practical insights derived from this study.

Background

The extant literature has highlighted the key differences in emphasis, forms of governance, performance measures, and several other features among the three different sets of governance logics–– PA, NPM, and NPG (e.g. see Osborne 2006; Wiesel and Modell 2014). To elaborate, PA has been characterized as focusing on bureaucracy and legitimacy, and as emphasizing compliance with the rule of law (Dunleavy and Hood 1994). Decision-making is conducted largely within the bureaucracy, and service delivery is centralized and hierarchical. The governance model in PA is procedure-focused and emphasizes regulation and strict input-control (Considine 2001).

In contrast, NPM stresses the use of market-centric principles and embedded competition as the means for improving the efficiency of public services. Unlike PA, NPM promotes a decentralized mode of organization, and it separates policy-making and service implementation by measures such as outsourcing. The NPM has been identified as a deregulation process (Meyer and Hammerschmid 2006). Rather than focusing on input-control, it advocates the use of output-based indicators to assess intra-organizational efficiency and financial results (Hood 1995).

For NPG, the discourse has shifted even further, to the governing of a networked society and to co-production with non-public sectors for policy-making and service implementation (Stoker 2006). Network governance is a central notion in NPG, which depicts a collaborative, networked mode of organizing in pursuit of public values (Provan and Kenis 2008; Sørensen and Torfing 2016). New public governance signifies a replacement of the hierarchical policy system by reconfiguring that system into a relatively horizontal and co-participatory structure. Effective policy network coordination is prescribed as a key role of the state (Osborne 2006). Control practices in NPG exhibit a greater concern on inter-organizational processes and the degree to which public agencies jointly meet the needs of the citizens.

Current literature contains two assertions regarding the process of public governance transition. Whereas one assertion denotes the transition as being largely manifested as a linear development (Entwistle and Martin 2005; Christensen 2012), a relatively more dominant view contends that the transformation is a complex process which constitutes a combining and mixing of the new governance logics with the previously entrenched ones (Wiesel and Modell 2014). Layers of governance logics are conceived as being in contact with each other in the wake of the reform and as causing elements of various logics to integrate (De Vries and Nemec 2013; Iacovino, Barsanti, and Cinquini 2017). During the early phase of an NPG reform, conflicts and tensions arising from the inconsistent practices ascribed to various governance logics are predicted to be likely, and the extent to which those dynamics dissipate or subside will depend upon the manifestation of a set of blending and segregating mechanisms that either foster or hamper the hybridization of the different governance logics. In other words, the outcome will depend on whether the competing values and practices pertaining to the different sets of governance logics integrate and in so doing yield a set of context-specific, coherent heuristics (Wiesel and Modell 2014). Drawing on previous work (Haveman and Rao 2006; Jacobs 2005), the desired blending mechanisms refer to planned and even to unintended actions that ultimately facilitate the various logics to hybridize into new configurations of governance practices. In contrast, segregating mechanisms are the institutional arrangements that prevent hybridization from taking place. In this sense, the absence of blending mechanisms and the presence of segregating mechanisms in the reform process –– leading to a failure to hybridize –– suggests a resulting governance structure that would lack coherence and stability and would be torn by a continued co-existence of inconsistent and competing logics perpetuating tension and conflicts.

In the context of social services sector management, a growing body of work recommends the prospects of network governance or governance practices akin to the NPG discourse (Park and Wilding 2014; Massey and Johnston-Miller 2016; Barraket et al. 2021; Sørensen and Torfing 2013). Those workers have advanced a viewpoint that organizations such as social enterprises and community groups can serve as key network actors, and this new mode of governance can help attract non-traditional resources for service delivery, harness the productive capabilities of civil society and public sectors to resolve complex problems, and can stimulate new organizational combinations and insights for improved policy and programme design (Yanto, Shang, and Roy 2021; Barraket et al. 2021; Park and Wilding 2014). However, the extant claims were largely built upon static empirical evidence and did not specifically consider the potential complications arising during the implementation of network governance. Whether introducing an NPG model into the social services fields in various geopolitical settings will result in the projected prospects requires further examination.

To address those knowledge gaps, this study has drawn on the experience of a recent policy initiative brought forward by the government in Hong Kong as an attempt to reconfigure the governance model of the local social services sector into one that is more closely aligned with the NPG discourse. This study was able to reveal the hybridization process of the Hong Kong policy case –– that is, how the values and practices ascribed to various governance logics that co-existed in the course of the reform integrated and illuminated the complexity that was manifested over the course of the reform process. This study aimed to identify the set of blending and segregating mechanisms that took place in the transition process and to delineate their relationships in the creation of a new hybridized governance model. In so doing, this study provides a nuanced view of the linkages between the implementation and the outcomes of applying NPG as a management approach to the social services sector.

This policy case, the SIE Fund, represents a public attempt to implement an intermediary-facilitated network mode of governance in the local social services sector, under the framing of social innovation. Drawing from network governance literature, the governance model of the Fund can be conceived as a purpose-oriented network, defined as ‘a network comprised of three or more autonomous actors who participate in a joint effort based on a common purpose’ (Carboni et al. 2019, 210), and with intermediaries as the network administration organizations (NAOs). This study underscored that intermediaries, characterized as entities that occupy a space between at least two other parties (Honig 2004; Armanios et al. 2017), can constitute a major blending mechanism in the process of governance reform. Specifically, owing to the nature of intermediaries in possessing the mediating and co-coordinative capacity to bring together disconnected actors in a collective pursuit of public values (Barraket 2020; Moss 2009), their intermediation work fills the structural holes in the network and may both facilitate actors who have previously been ingrained with different sets of values and also foster a preference for incrementally developing a shared interpretation of policy priorities and standards through collective learning. In turn, that facilitation may contribute to the emergence of a coherent set of collaborative practices –– that is, to a new, hybridized, context-specific mode of governance.

Hong Kong, a city in the Eastern Asian region, has positioned itself as an international hub of innovation.¹ Discussion of innovation in social services delivery is not new in the local context, but the term social innovation has only recently been popularized in the local policy discourse and strategized as a tool for alleviating poverty (Chan et al. 2019). A recent study suggested that Hong Kong’s policy framing of social innovation shares similarities with that in other geopolitical settings, where the term signifies public attempts in social services reform (Chan, Chui, and Chandra 2021). This policy framing on social innovation is closely aligned with the technocratic stream (Montgomery 2016), in which the term is perceived to be politically neutral and emphasizes the market sphere in resolving complex social problems. Such a technocratic framing of social innovation is also evident in Anglo-American settings and European settings (Simon, Roy, and Hulgård 2021; Baptista et al. 2019). Thus, the experiences drawn from this study’s policy case could shed light on the governance transition process in other regions where social innovation is also been using as a governance approach for instituting social service sector reforms.

Study design

A case study research design was adopted, given that the purpose of this work was to unfold the complexity of the policy case and advance the theoretical understanding of the hybridization of governance logic in the reform process (Yin 2009). Drawing from the anthropological paradigm, the author traced the naturalistic development of the SIE case for a six-year period, from 2015 to 2021. To compile a holistic investigation, data were collected through a multitude of methods: document reviews, participant observations, open-ended and semi-structured interviews, and a structured survey. The study involved both a retrospective investigation and a prospective ongoing investigation of the case. Details of the materials collected are summarized in Table 1.

Table 1. Materials collected for analysis².

	Phase I (Jan – Dec 2016) Description of the setting and details of the Fund’s network governance arrangement	Phase II Tracing of the process of the reform and uncovering of the complexities that emerged during the process
		1^st half (Jan 2017 – Dec 2017)	2^nd half (Jan 2018 – May 2021)
Document review^a	Documents on the Fund’s directions and strategies Proposals of the Intermediaries	Relevant documents emerged over this period (e.g. assessment reports of the intermediary’s services)
Participant observations	Notes from the Fund’s Board meetings (n = 4)	observational notes of the Fund’s Board meetings and the intermediaries’ implementation
Interviews	Transcripts of exchanges with the members of the Fund’s task force and secretariats (n = 10)	Transcripts of exchanges with task force members (n = 20)^b secretariats of the Fund (n = 14)^c managers of the intermediaries (n = 12)^d project holders (n = 32)^e funding applicants (n = 15)^f other relevant informants not directly involved in the Fund’s implementation network (i.e. local policymakers and other service providers in the local field) (n = 10)^g	Transcripts of exchanges with task force members (n = 14) secretariats of the Fund (n = 14) managers of the intermediaries (n = 12) project holders (n = 20) funding applicants (n = 10) other relevant informants not directly involved in the Fund’s implementation network (n = 10)
Survey	–	–	Structured survey with the project holders (n = 103)

The data collection was categorized into two phases. Phase I involved familiarization with the policy setting, identification of the underlying rationales of the reform, and sketching out the newly proposed governance model of the Fund. Then, data collection moved on to Phase II and prospectively traced the transition process, largely through extensive interviews. Using a theoretical sampling approach (Eisenhardt 1989), a wide range of stakeholders were interviewed and the data collection stopped when theoretical saturation was reached. Stakeholders interviewed were those who were directly involved, such as the policymakers (i.e. task force members of the SIE Fund), implementors (secretariats of the SIE Fund and the managers of the intermediaries), and end-users (i.e. funded social service providers, a term used interchangeably here with project holders), as well as those who were indirectly involved, such as non-funded project applicants and local social service providers who did not exhibit an interest towards the Fund. Whereas the data collection was a continuous process, conceptually it could be interpreted as involving two consecutive rounds of interviews. The first round of the interview process focused more on establishing the baseline situation of the early implementation of the reform (i.e. what the tensions and dynamics were that emerged in this setting, and why) and on formulating the blending and segregating mechanisms involved in the hybridization process.³ The second round of the interview process was a prospective, ongoing investigation with the main purpose of identifying the potential changes and/or inertia experienced in the reform process.⁴ In each round of interviews, some informants were interviewed multiple times for data clarification.

In parallel, the second half of phase II also included a survey that translated some of the observations drawn from the earlier phases of the interviews into a standardized questionnaire for examining certain intermediation works that were intended to create blending in the Fund’s governance model. Respondents were all project holders who were affiliated with the intermediaries.⁵ Of the 132 eligible informants, 103 (78%) completed the survey.⁶ Throughout phase II, the author also participated repeatedly in the Fund’s internal discussions and reviewed any additional documents deemed relevant to the case.

Qualitative content analyses were conducted on the collected documents and the observational notes (Bowen 2009). The author and two masters-degree-prepared research assistants coded the data independently and then performed a thematic analysis using the following steps: familiarizing themselves with the data, generating the initial codes, searching for themes, reviewing the themes, and defining and naming the themes (Braun and Clarke 2006). Findings relevant to the research questions were then discussed. Where there were discrepancies, the issues were discussed among the coders until a consensus was reached. It is worth highlighting that the analysis was not done in a static fashion, it was more closely represented as a continuous work by undertaking analyses at multiple sequential timepoints prospectively along the reform process. Each point of analysis was conceptualized with a pre-design and a post-design. Specifically, the author used the newly emergent materials to construct an updated portrayal of the governance reform and then compared that portrayal with the baseline understanding that had been formulated from the previous analyses. Through a series of those pre- and post-comparisons, the analyses attempted to identify what had changed and what remained unchanged over the course of the reform, and also to identify the underlying causes.

Findings derived from the qualitative thematic analyses for the previous timepoints were used to inform the formulation of questions for subsequent interviews at later timepoints. At the later stage, some of the results derived from the analyses during the earlier timepoints were explicitly presented and discussed with the relevant informants, such as the Task Force and the Secretariats of the Fund, as part of the triangulation procedure. Following the survey, descriptive statistics were conducted on the quantitative data. Analysis of the quantitative data specifically came at this stage to verify and challenge the author’s propositions put forth in the earlier phase. Finally, this series of analyses were aggregated and regrouped back into two periods, albeit arbitrarily, and are presented in this work as (1) the period of early implementation of the reform and (2) the emergence of hybridization. Ethics approval was obtained from the Human Research Ethics Committee of the University of Hong Kong (reference no. EA160620).

This study

Before elaborating on the findings, I shall summarize here the backdrop of this policy case. In general, the provision of state welfare in Hong Kong emerged after the colonial era and can be roughly categorized into two streams. One refers to the provision of large-scale welfare infrastructures such as public housing, a healthcare system, education, and social security, and the other is the provision of personal care social services such as childcare, elderly care, family services, and crisis support, among others (J. Wong 2017; Hung 2012). The interest of this study pertained to the latter group. Specifically, provision of these social services emerged following the late 1970s, and between the late 1970s and the 1990s, such services were largely delivered by local non-profits based on the financial subsidies offered by the government through the subvention arrangement. During that period, the relationships between the government and the non-profits were largely top-down, and the non-profits were considered to be the bureaucratic arm of the government (Lee 2005). The subvention system relied on input-based regulatory control, and those practices were characterized as creating inflexibility in resource deployment and being administratively cumbersome (Leung 2002; Lee 2012).

Starting in the late 1990s, in an attempt to contain the rapid growth in welfare expenses, the government initiated major NPM reforms that introduced a large number of market mechanisms into the social services sector (Leung 2002; Lee 2012). A hallmark of that reform was the introduction of lump-sum grants as a partial replacement of the subvention arrangements. Contracting out in tandem with the competitive bidding was another strategy to privatize social services. These market-like arrangements gave rise to a rapid expansion of the local social enterprises sector. However, many of these social enterprises were found to lack financial viability, given that their survival depended heavily on continuous public funding. This NPM reform ultimately was criticized for failing to improve the state funding dependencies in the provision of social services (Leung 2002; Lee 2012).

Beginning in the twenty-first century, policy language focusing on collaborations between the government, the business sector, and the community in social service provision has increasingly emerged (Nip 2010; Lee et al. 2005). Notably, the government emphasized on ‘playing the role of facilitator to promote the tripartite partnership’.⁷ Emerging at this period and concurrently with a growing discourse on social innovation in both the local and international context, the SIE Fund was established by the HKSAR government and was announced in the government’s 2014 policy address. First budgeted with HK$500 million, an amount that was later doubled, the SIE Fund was a notable public funding scheme to support local social innovative initiatives to provide human-centred services for the underprivileged. Organizations taking up the publicly funded portion comprised both non-profits and private corporations and typically identified themselves as social enterprises. Whereas the Fund showed no explicit preferences in resource allocation by types of organizations, the Fund was more appealing to the latter group. An earlier study highlighted that the Fund facilitated a notable group of new players entering the local social services sector, a large number of whom came from the private sector and had little or no prior experience in the local social services field (Chan et al. 2019).

The network governance model of the SIE Fund

The SIE Fund adopted a public board system in which the members (termed task force members) are community elites selected from various sectoral backgrounds (e.g. non-profits, private corporations, philanthropies, social enterprises, academic institutions, think tanks, legislators). That group functions as the policymakers (i.e. steering the Fund in strategic directions) as well as the gatekeeper of the public resources (i.e. vetting the funding). The secretariats of the Fund, who are all public managers, are involved in the policy formulation and are the Fund’s administrative arm. From the author’s fieldwork and exchanges with the task force and the secretariats of the Fund, there were extensive indications of the Fund’s inclination to cultivate a collaborative social services sector. Emphasizing that the ecosystem was networked space, the Fund consistently framed itself as a ‘network facilitator’ in its internal reports. Cross-sectoral collaboration was constantly stressed, and some members of the task force explained that their vetting decisions were focused on ‘whether the applicants can utilize the funding to mobilize collaborative work in social service delivery’.

Another stream of evidence pointing to an NPG transition can be embodied by the Fund’s use of intermediaries as part of their governance system. Figure 1 illustrates the Fund’s three-layer governance structure under the Innovative Programme (IP) scheme, which is its key initiative for supporting the provision of community-led social services. Wherein the policy space can be conceptualized as three nested networks (i.e. at the Fund level, the intermediary level, and the project level). To elaborate, one can conceptualize that the SIE Fund had moved away from the typical two-layer funder-grantee system (adopted in most of the other local public social service Funds) and had adopted a three-layer funder-intermediaries-grantees structure. Major initiatives of the Fund, such as the Innovation Programme (IP) Scheme, the Applied Research Scheme, and other flagship initiatives are all embedded in this intermediary-led network mode of governance.

Figure 1. The network governance model of the SIE Fund.

As it was forged earlier, the Fund’s governance model is a purpose-oriented network (Carboni et al. 2019), with the intermediaries, represented by the solid squares in Figure 1, being the NAOs who are contracted to manage their network members (i.e. the project holders). Specifically, the intermediaries brought in their existing knowledge and social capital (e.g. networks with NGOs and philanthropic foundations) to support their network members. Their typical work involved organizing the funding scheme, providing tailored advice and support, sourcing and securing matching funds, delivering an incubation programme and networking, and monitoring the project. The Fund described this intermediation arrangement as an ‘experimentation’, given that at the time of its establishment, this three-layer administrative system had rarely been used by other local public funding schemes in the social services context.

As will be further elaborated, the intermediaries mediated the vertical relationship between the secretariats and the project holders.⁸ They also connected the project holders with different players and stakeholders in the Fund’s system and the larger SI ecosystem, as part of their incubation support. Apart from that, the intermediaries worked together to leverage their complementary strengths and resources. They collaborated formally (e.g. via referrals) and informally (e.g. via information and knowledge sharing) to provide incubation support for the project holders.

As the social service providers contracted by the SIE Fund, the project holders, represented by the solid circles in Figure 1, each proposed their own service scope (social innovation projects). Within that network, these project holders brought in their innovative ideas, expertise, and social networks (e.g. social service collaborators are represented as solid triangles in Figure 1), combined with the financial and incubation support of the SIE Fund and the Fund’s intermediaries, to execute the mandate of the Fund (i.e. relieving poverty) via their social service delivery. Both the intermediaries and these project holders entered into a contractual agreement with the Fund, with a prespecified service scope. Payments were in a lump-sum form and were scheduled for disbursement in multiple instalments at prespecified times. However, the payments would only become effective when the service providers’ performances –– that of both the intermediaries and the project holders –– were deemed satisfactory upon the Fund’s approval. In other words, the Fund’s governance model leaned towards being top-down and hierarchical, and owing to resource dependence, the Fund had regulatory control over other actors in the network. The findings presented next were mainly drawn from the Fund’s IP scheme.

Early implementation (late 2015 to middle 2017)

For the early implementation phase of the governance system, the data provided clear empirical support substantiating that the transitioning would be complex and that a layering of practices pertaining to different sets of governance logics (PA, NPM, NPG) would co-exist. Specifically, input controls, such as itemized budgeting, were still adopted in the Fund’s governance system. The Fund also had tight risk control, including frequent financial audits and multiphase funding instalments. In addition, elements resembling managerialism were evident –– as one of the major principles of the Fund stated, ‘Risk assessment is important while promoting innovation so as to maintain the prudent use of public money’. Cost-effectiveness and public accountability were stressed.

During the early implementation phase, tension owing to clashes of the values and practices of competing governance logics were widespread. The author repeatedly encountered discouraged project holders, many of whom had come from a commercial background and were new to the SIE Fund space, who complained about the bureaucracy of the Fund. In contrast, the secretariats of the Fund were discontented over the poor quality of the reporting of funding progress.

There were so many reporting works –– i.e. operational reports, financial reports, and audit reports, etc. It was unanticipated and overwhelming, and frankly, the amount of funding I asked for was not much. The administrative workload was undue. There were also so many back-and-forth in the process. This [the bureaucracy] is why I would never ask for government funding in the future. (Reflective account from a project holder)

The [project holders’] reports are just not up to standard. No way this report can pass our internal standard of requirement. We have to revert the report back and seek a refinement. Unless the report seemed acceptable, the financial control w[ould] not provide the next funding instalment. (Sharing from a Secretariat of the Fund)

Tensions were also detected between the intermediaries and the Fund, and especially the members of the task force. Whereas the task force criticized the intermediaries for failing to identify innovative and high-potential projects, and for not being good at their incubation work, the intermediaries rejected those accusations and retaliated against the task force for being overly harsh and ambiguous in their decision making, and also for being so high up in the management chain that they failed to recognize the intermediaries’ frontline efforts.

Do you find the [intermediaries] doing decent work? Many projects that they identify are not innovative. We rejected [the funding support for] them very often. They were supposed to be the good gatekeepers and they weren’t doing their work properly. (Sharing from a task force member)

They [the task force] are too harsh. They need us to identify projects [that have a] good business model, [are] innovative, and will have social impact. These three many times do not co-exist. If there are such projects, they may not consider applying for the Fund. There are other funders who are more lenient. (Exchange with managers of the intermediaries)

I don’t understand how the task force interprets social innovativeness. Projects that we found innovative were all rejected. They are being too risk-averse (Exchange with a manager of an intermediary)

The task force had never seen what we have done in action. On what grounds [did] they come to this accusation? (Exchange with managers of the intermediaries)

Project holders also commonly failed to appreciate the intermediaries’ support, instead perceiving the intermediaries to be unwitting disciplinarians and enforcers of the Fund who tried to control them tightly.

Intermediaries aren’t helpful. They are the government’s pawns to make us work. When we encountered operational challenges, what they did was just keep asking whether we ha[d] achieved the performance indicator. No help [was] offered, whatsoever. (Complaints from a project holder)

In regard to the project holders’ execution of social services, many of them expressed having experienced substantial challenges. Many voiced that the incubation support they received was insufficient and that there was widespread distrust from the outsiders, because they were new in the network space (i.e. in a non-profits-dominated social services field).

I regretted to do this project. I found [it] so difficult to find partners [for service delivery]. I sense that if you are not non-profits, they are not willing to partner with you. I felt that I [was]a laboratory rat. O[n o]ne side the Fund [wa]s pushing me to do it [for fulfiling the KPIs], and on the other side no one [potential partners for service delivery] would trust [any]one. I think the system just fail[ed] us. (Reflective account from a project holder)

The evaluation study of the SIE Fund at that time highlighted that ‘there were considerable problems (tension and conflicts) at this embryonic stage’.

Emergence of the hybridization (late 2017 to early 2021)

As the reform continued to unfold, some of the problems gradually diminished –– for instance, the emergent hybridization of governance logic did lead to a more coherent and harmonious set of values and practices. Furthermore, multiple rounds of administrative streamlining led to recognizable accounts of a more flexible exchange system between the Fund and the project holders. Notably, the number of complaints that the project holders filed against the intermediaries and the Fund declined, and evidence grew of an increased mutual understanding between the parties.

A range of intermediations conducted by the intermediaries clearly contributed to the emergence of a more coherent governance system. In other words, this study recognized that the intermediaries acted as key agents to facilitate blending activities throughout the governance transition process. Their blending work was achieved in two ways: (1) they conducted vertical mediations between the secretariats and the project holders (their network members), and (2) they performed horizontal coordinative work between the project holders and the community stakeholders.

First, data collected from all levels showed that the mediating work of the intermediaries facilitated an increased trust and mutual understanding between the secretariats and the project holders. Project holders expressed an increased understanding of the need for reporting for public accountability, and the secretariats also expressed a greater acceptance for removing some procedures that were conceived as red tape. The intermediaries reportedly assisted in efforts for streamlining and at the same time provided additional support to the grantees. Appreciation from both sides towards the intermediaries increased in response to their effectiveness as middlemen. In the survey, 78% of the 103 respondents (project holders) recorded a satisfactory or favourable opinion towards the mediation support that the intermediaries provided (see appendices).

The reporting arrangements were streamlined with the help of the intermediaries. There are fewer complaints on the workload of the administrative work. Also, we are happy that the quality of the reports, overall speaking, improved. We have [fewer] quality issues to deal with and we are aware of the active intervention provided by the intermediaries. (Extracts from interviews with the Secretariats)

I am happy with the reporting workloads. They are manageable. Managers of the intermediaries are also helpful. (Reflections of a project holder newly entering the system at the 3rd year of the implementation)

Apart from mediating the vertical relationships, the intermediaries were also crucial in coordinating the horizontal networks, for instance by facilitating the development of partnerships in social services delivery. Across the study period, the intermediaries reflected their accumulated experience for their incubation work, such as in organizing workshops to grow the project holders’ capacity. Interviews with the project holders supported this assertion. In addition, the survey reflected that a notable proportion of the respondents found the incubation support by the intermediaries to be satisfactory, although that varied in accord with different aspects of the support.⁹ Intermediaries also highlighted their enhanced efforts to make network referrals according to the grantees’ needs. They reported that they enhanced cross-intermediary collaborative efforts to organize a mingling of activities that could facilitate cross-project fertilization. Interview data reflected that these intermediations helped the new actors (project holders who were new in the local social services field) assimilate into the existing structure, providing those actors with various connected ways to integrate. The survey results showed that in the respondents’ development of key networks, 77% of the 103 respondents reported having received key help from the intermediaries’ support for information and knowledge exchange, 69% from their co-creation of social services, and 53% from their access to potential new sources of financial support.

We have been far more involved in supporting the grantees. When they encountered some issues., e.g. legal, we would leverage our networks to find pro bono support for them. Our target is to bridge them to the help they need. (Reflection of a manager of the intermediaries in the 2nd year of model implementation)

I didn’t have networks with the non-profit sector. An intermediary linked me up with several that they are well connected. Eventually I partnered with two. Now they are our important community partners. (Extracts from interviews with project holders).

Through repeated communication, the secretariats sensed an increased level of trust and acceptance from the task force members towards the new governance arrangement. To some extent, the intermediation work that facilitated the transition of governance was limited not only to the intermediaries’ work but also involved active engagement of the public managers.

We actively explained the work of the intermediaries to the internal superiors and the members of the task force. They also read the evaluation report. They seemed to have less scepticism. (Extracts from interviews with Secretariats)

Persistence of the segregating mechanisms

Even though over the course of time the reform had led to a more coherent governance structure, the data still suggested the presence of inconsistent or competing logic. One source of that inconsistency appeared to be attributable to the different values across the sectors: the public versus the private sectors. Despite notable efforts to streamline the administrative procedures, complaints from the project holders still were not uncommon and came most noticeably from those with a strong background in the private sectors. Those complaints quoted the procedural controls, such as the requirement for itemized budgeting and fixed-period instalments, as a substantial inconvenience. Managers of the intermediaries argued that the inflexibilities delegitimized their work and contributed to a negative feedback loop that deterred potentially interested parties from participating, thus weakening the Fund’s attempt to become a network facilitator.

There has been quite an improvement on the processing time for the reporting approval. However, there are still other [things] the Fund has not solved. There will still [be] itemized budgeting. When project holders hope to revise their budget, they will have to go through another round of approval. These arrangements are inflexible. Frankly speaking, they affected our work. Project holders complained [that we were] not being helpful but actually it was related to the Fund’s arrangement. We have little to work on but just to comfort the project holder. (Reflections from managers of the intermediaries)

The Fund may not [be] fully aware [of] the negative effects of the administrative arrangements. Due to the negative experiences, some stated that they will never apply [to] the SIE Fund again. The word will spread and that will deter others from joining. (Reflections with managers of the intermediaries)

In addition, tensions emerging from the hierarchical governance of the Fund diminished but never entirely dissipated, with intermediaries sometimes highlighting their frustration towards the fluidity in decision making among the task force’s top administrators.

Over the years, some members left and some new members entered. Their funding preference, from my encounters with them, seemingly changed and they may not be entirely explicit to us. We have to (re-)learn the game rules. (Reflections of managers of the intermediaries)

Tensions also still arose between the intermediaries and the project holders. Specifically, a small but notable group of project holders reflected that the assistance provided by the intermediaries was not supportive, and they typically reported that a mismatch remained between what they needed and what the intermediaries offered. For their part, some intermediaries blamed budget constraints for their inability to provide a more preferable level of support, stating that they had to prioritize their work of fulfiling the key performance measures that had been stipulated in their contracts and that relied more on output-based monitoring than on outcome-based measures.

The budget we had gotten from the Fund required us only to give the vetting support, and we were not ‘obliged’ to provide other additional help. (Defence coming from a manager of the intermediaries)

We have to finish our vetting paperwork. We have a key performance measure to meet [in the number of projects we vetted]. (Explanation provided by a manager of the intermediaries)

The findings illuminated the fact that governance reform is a dynamic and complex process. Table 2 summarizes the blending and segregating mechanisms that influenced the process of hybridization, in this case study.

Table 2. The blending and segregating mechanisms that manifested over the course of the governance reform.

Mechanisms	Influences on the Fund’s implementation of a network governance reform of the provision of local social services
Blending activities of the intermediaries	facilitating mutual understanding and trust between the public agencies and the policy clients through vertical mediation assisting horizontal partnerships between network members (project holders) and actors in the social service fields (e.g. non-profits) by active coordination, and also enabling improvement of the capacities of their network members via incubation support
Segregating mechanisms	value differences among the actors (i.e. public versus non-public participants) caused by the multisectoral nature of the Fund’s implementation network feelings of uncertainty, and concerns about perceived fluidity in decision-making by task force administrators, due to the top-down governance structure

Discussion

Through prospectively tracing a policy case in Hong Kong for a six-year period, this study illuminated the process of a public attempt to implement a new mode of governance in the local social services sector by introducing a network-based structure akin to the NPG paradigm. This research echoed previous work illustrating that such governance transitions are unlikely to be a simple linear replacement, and instead are far more dynamic and complex and involve combining the values and structures of the newly introduced with those of the previously entrenched –– that is, forming a hybridization of governance logics (Wiesel and Modell 2014). As this case illustrates, despite the intention to create a new, networked mode of social service governance, administrative arrangements pertaining to the earlier models of PA and NPM, such as itemized budgeting and managerial-based performance control, continued to co-exist. In addition, although the reform contributed in some ways to a decentralization and diversification of the local social services sector, it also reinforced a hierarchical imprint that the newly entrenched participants, such as actors coming from a private sector background, might not favour. This study unveiled a chaotic period during which tensions were not uncommon among the network actors –– the early phase of the hybridization –– which preceded the emergence of a more cohesive set of values and practices –– a new, hybridized governance structure.

This study added to the literature by expanding beyond the existing static discussion of the NPG in social services management. This work incorporated the temporal considerations and underscored the complexity that emerged during the implementation of a new governance structure that was oriented towards a network governance, or NPG. The findings of this study contribute to the literature by offering a more balanced view of the prospect of transitioning to an NPG, and they underline how implementation issues can come into play and influence the potential of this form of governance (Carmine, Nasi, and Rivenbark 2021). This work recommends closer examinations of the implementation of an NPG in social services management, and it outlines the need for careful deliberation on how to make a new mode of governance in order to yield the beneficial prospects that are connected with an NPG (Barraket et al. 2021; Yanto, Shang, and Roy 2021). As was found in this case, where historical legacies are in place for installing a new mode of governance, how the logics of NPG combine with the previous, entrenched practices, and how they unfold via the hybridization process to become a context-specific hybrid mode of governance in a social services context, all deserve deeper examination.

This work does not refute the prominence of the NPG model in social service sector management; instead, it stresses a closer examination of the implementation process. In doing that, this study generated new knowledge by outlining what made the network governance model work in this policy case –– the blending of mechanisms that facilitated the hybridization of different logics into a coherent set of values and practices. Specifically, it underscored the salience of the intermediaries acting as key network actors in mediating the vertical and horizontal relationships in this social services system. Existing research on governance transitions in the social services sector has infrequently touched on the roles of intermediaries, but relevant insights can be drawn from the literature in other domains of governance reform, in which prior work has articulated that intermediaries perform significant hidden work, such as institutionalizing new modes of regulation, filling in the institutional gaps arising from the reform, and building up and stabilizing new relations (Moss 2009; Williamson 2014; Fransen and LeBaron 2019). Similar revelations have also been found in the literature on social innovation, with intermediaries described as playing a transformative role in the process (Moore, Westley, and Brodhead 2012; Jenner 2016; Barraket 2020; Jae-Yun and Yoon 2021). Towards that end, this work adds to the literature by enriching the conceptual discussion on the role of intermediaries in the growing discourse of network governance in social services sector reforms.

Network governance literature offers crucial insights into how the intermediaries, or the NAOs, could have acted as effective network managers in the public service implementation networks. Occupying the structural holes in the network (Burt 2004), the intermediaries could have a translational advantage for communicating with different network actors by understanding what resonates and what does not. The network coordination by the intermediaries in such situations as this policy case between the Fund and the project holders could improve interorganizational trust and mutual understanding, and that in turn could reduce transaction costs, enhance stability in relations, and facilitate commitment across the network actors (Emerson, Nabatchi, and Balogh 2012). Also, the brokerage role of the intermediaries could facilitate the transfer of information and knowledge among the disconnected members, and as a whole it could enhance network performance (Huang 2014). Intermediaries might also be particularly salient in empowering their network members via resource championing, by transferring legitimacy through the mechanism of borrowing (Perry-Smith and Vittorio Mannucci 2017). The literature also has pointed out that various characteristics of the NAOs, such as network composition, structures, and the coordinative strategies employed by the network managers, could affect the efficacy of the network governance (Provan and Kenis 2008; Huang 2014; Herranz 2008). That might explain the variations that this study observed in the incubation performance among the intermediaries. One line of future work, therefore, would be to explore the coordinative approaches of the intermediaries, in relation to the constellations of their networks that would either facilitate or hamper the implementation of network governance in the social services sector.

A number of studies have highlighted the differences in values across the sectors (Saz-Carranza and Longo 2012; Feeney 2008). The dissimilar-purpose thesis suggests that the public sector tends to attach the greatest value to authority, has a mandate to work for the public good, and has a long-term horizon, whereas the private sector values the market as the source of control, is customer-driven, and has a short-term horizon (Van Wart 2013). Herranz’s (2008) work highlighted that these sectoral differences matter in network management and require network managers to deploy different strategies, and he proposed a continuum from a passive approach to an active approach in an effort to handle actors with different sectoral orientations. He also pointed out that coordinating a complex multisector-networked environment is an emergent challenge in public management, given that public service implementation networks are increasingly multi-organizational and multisectoral. In this study’s policy case, however, the Fund’s contribution in diversifying the local social services sector by facilitating the participation of actors from the private sector in their delivery of social services actually made the governance more challenging because the entire network became more multisector-based. The persistent tensions observed between the Fund (public sector) and the project holders, many of whom had a background in the private sector, appeared to relate to their differences in values, as ascribed to their sectoral origins. The Fund’s preference for a hierarchically-based directive administration because it provided more authority control was viewed as rendering inflexible the coordinative strategies of other network actors, and thus as constituting a segregating element that prevented further integration in the Fund’s governance structure.¹⁰ This observation echoes the assertions that the conventional hierarchical tradition of public management does not apply to multisectoral forms of network governance (O’Toole and Kenneth 1999; Agranoff and McGuire 2001), and it calls instead for an integrative approach towards the management of complex multisectoral networks (Herranz 2008).

Overall, this study was a rigorous effort to unfold the complexities of public governance transitions in the social services sector in an East Asian setting. Through constructing a contextually rich case, this research adds novel insights into the process of re-configuring social services sector management into a networked organizational mode. Furthermore, it illuminates the blending and segregating elements that occur over the course of such a transition and that influence the implementation of the new governance system. Practical wisdom on facilitating the emergence of a coherent set of governance practices was also revealed. However, the single-case-study design limits the findings’ generalizability, thus warranting future studies that will take a process-oriented approach to explore the implementation of an NPG approach in the social services sector.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplemental data

Supplemental data for this article can be accessed at https://doi.org/10.1080/14719037.2022.2132280.

Additional information

Notes on contributors

Chee Hon Chan

Chee Hon Chan is Assistant Professor of the Faculty of Social Science and the Programme of Data Science and Policy Studies at the Chinese University of Hong Kong. His research interests are policy innovation, governance network, and social innovation.

Notes

1. “International Innovation and Technology Hub”, HKSAR Government, October 2021, https://www.policyaddress.gov.hk/2021/eng/pdf/publications/14-5/06_International-Innovation-and-Tech.pdf; assessed on 30^th June 2022.

2. a: All of the documents were made available directly from the Fund. b: :Recruitment of task force members for interviews was conducted through referrals from the SIE Fund. The task force members’ participation in the interviews was entirely on a voluntary basis. However, because the task force members were made up of policy elites from multiple sectors, including corporate businesses, the non-profits sector, philanthropic foundations, social entrepreneurs, academicians, and public managers from other units of the HKSAR government, the researcher attempted and was successful in obtaining at least one the task force member from each background field for formal interviews and/or informal exchanges. All of the interviewed task force members participated in the project vetting and policy deliberation of the SIE Fund. c: Interviews and exchanges were conducted with all levels of the secretariats, ranging from the Secretary-General of the Fund, to senior managers, to officers who had interactions with the intermediaries and the project holders. d: Managers of the three intermediaries were interviewed. In phase II, since the service of intermediary C was terminated by the Fund, no personnel were recruited for interviews from this particular intermediary in phase II. The interviewed managers of the intermediaries ranged from the unit-in-charge to the frontline officers. They all had been involved in the exchanges with the Fund and the project holders, although with varying degrees of interaction and different focuses. Most of the managers also had been directly involved in the incubation activities, involving planning, organizing, and administering the incubation programme. e: For interviews with the project holders, recruitment began by going through the list of funded projects of the SIE Fund. Next, the researcher categorized those projects by their main service nature, project stage, and affiliated intermediaries (see Appendix B). Then, the maximum variation principle was applied for the sample selection. All interviewed informants were project applicants who were directly involved in their project implementation (i.e. social services delivery). They typically were interviewed when they had just completed their funded projects. They all had first-hand experience interacting with the managers of the intermediaries and with the secretariats of the Fund. f: Informants were referred directly by the intermediaries. These informants were all project applicants, but they had not successfully obtained a funding support. g: The researcher began by asking for referrals by the task force and secretariats, and personal contacts, to recruit potential informants who were not directly involved in the operation of the Fund (but who at the same time had substantial seniority in the field), and hence would be able to provide comments on the SIE Fund and the local social service sector reforms. Then, a snowball sampling strategy was used in which the researcher asked the interviewed informants to refer potential candidates. Finally, interviewed informants included legislators, senior government officials, chief-executive officers of some large-scale non-profit organizations, leaders from the business field, and several senior managers in the local social services sector.

3. Interviews were conducted either in Cantonese or English. The interviews conducted with the implementors, end users, and non-funded project applicants began by exploring the involvement of the informants in the Fund’s governance process (i.e. in what parts of the implementation was the individual involved?). Then, the interviews solicited the informants’ experiences, both positive and negative, during their participation. At that point, the interviewer (i.e. the author) tapped on the rationales underpinning the individual’s appraisals and also enquired about the potential changes she perceived that could improve the governance process. The interviews with the policymakers (task force members) focused on exploring the rationales of the governance design and explored the expected benefits and potential risks associated with this governance transition. The interviews with local social service providers focused on soliciting their perceptions and observations of the changes that were made to the introduction of the SIE Fund, in the existing governance of the local social services sector. Interviews typically lasted approximately 45 to 75 minutes and were audio-recorded and then transcribed. When audio-recordings were deemed not appropriate (e.g. during quick exchanges), the interviewers (mainly the author) took written notes. The majority of the interviews at phase I were conducted face-to-face.

4. The 2^nd-round interviews shared similarities with those in the 1^st-round. For instance, the interviews with the implementors, end users, and the non-funded project applicants also began with enquiries about their degree participation in the Fund’s governance process. That again was followed by soliciting the informants’ experiences and observations during their participation. In a departure from the 1^st round, additional questions were incorporated to identify the changes in the Fund’s governance over the course of the implementation at that point. Specifically, the interviewer (i.e. the author), who had seen some of the observations and appraisals reported by the informants during the 1^st-round of interviews, asked the informants whether, and to what extent, they shared similar experiences or had similar observations and appraisals. Plausible causes of the changes (and non-changes) in the Fund’s governance over the course of the implementation were solicited. The 2^nd round interviews with the local social service providers (i.e. providers outside the SIE Fund’s network space) incorporated elements of data triangulation. Whereas the beginning of those interviews also asked the informants how they perceived the SIE Fund to have affected the local social service sector governance over the course of the transition, the interviewer (the author) again presented some of the observations and assertions derived from qualitative thematic analysis of the 1^st round of interview data and asked the local providers whether they (dis)agreed with those assertions or observations, in regard to the Fund’s influence on the governance of the social service sector. In the 2^nd round interviews with the policymakers (the task force), the elements of data triangulation were even more explicit. Whereas the 1^st-round of interviews explored the underpinnings of the design of the Fund’s governance model, the 2^nd-round interviews had a greater focus on discussing the observations made from a thematic analysis that was based on the 1^st-round interview data. The purpose of that approach was to learn how the local providers agreed (and disagreed) with the first-round assertions and also to explore their conjectures about how the observations might tie into the design and implementation of the Fund’s governance model.

5. Characteristics of the intermediaries are summarized in Appendix A.

6. Characteristics of the 103 respondents are shown in Appendix B. All project holders were invited to take part in a survey at the time that their projects were completed. For instance, when a project commenced in January 2019 and their implementation period was 12 months, this study sent the project holders an invitation via email (and contacted them via intermediaries) in December 2019 inviting their participation in the survey. Once they accepted, a trained research assistant either called or directly visited the project holders to administer the survey, which was a standardized survey with one part that explicitly sought the project holders’ appraisals of the intermediaries’ support. (Exact wordings of the eight items deriving from the qualitative findings and the item’ responses are summarized in Appendix D). The survey typically took 30 minutes to complete.

7. This contention can be exemplified from the statement made by The Secretary for Labour and Welfare of the HKSAR government, stating that “The Government strives to introduce new thinking, new motivation and new values to solve social problems, by channelling Hong Kong’s rich social capital into positive supporting forces and promoting greater collaboration among the community, the business sector and the Government … In working with the elite of society, he wishes to transform the concept of tripartite collaboration into action, and promote a new culture of helping the disadvantaged.” A press release of the HKSAR government, https://www.info.gov.hk/gia/general/201101/06/P201101060151.htm; assessed on 30^th June 2022).

8. Intermediaries within the Fund’s IP Scheme were contracted to process and vet funding applications, and they made recommendations for the SIE Fund’s later-stage vetting and endorsement. During the funding period, they also handled reporting submissions and communications between the SIIs and the SIE Fund.

9. Qualitative reflections from the project holders regarding intermediation support, and also the corresponding survey findings, are summarized in the appendices C and D.

10. After the compilation of this paper, an exchange between the author and the Secretariats of the Fund revealed that the Fund is committed to review its governance model and is considering flattening the hierarchical structure by delegating greater regulatory control to the contracted intermediary agencies.