Collaborative governance of public health in low- and middle-income countries: lessons from research in public administration.

Multisectoral governance, one of many terms used to describe collaborative, cross-boundary approaches to solving complex public problems, is being applied broadly in several policy arenas, most notably in environmental and natural resource management, but increasingly in public health in multiple settings and scales around the globe. This paper explores how to transfer knowledge about collaborative governance to challenging public health settings found in low-income and moderate-income countries (LMICs). This paper presents a general background on collaborative governance, summarises some relevant empirical findings on the performance of collaborative governance and lays out some of the challenges and considerations for thinking about improving collaborative public health governance in LMICs.

Collaborative governance is working in a variety of cross-sector policy arenas and research has confirmed and extended practitioners’ experience to improve performance.

This paper highlights the potential and the challenges for designing and managing cross-sector collaborative governance in the context of public health provision in low-income and moderate-income countries (LMICs).

When applying collaborative governance to cross-sector public health approaches in LMICs, it is recommended that one:

Takes a systems approach that acknowledges a complex, dynamic context.

Uses a design approach informed by a comprehensive institutional and sociopolitical assessment.

Focuses on the multiple leadership demands of cross-sector collaborative governance.

Introduction

Multisectoral governance, one of many terms used to describe collaborative, cross-boundary approaches to solving complex public problems, is being applied broadly in several policy arenas, most notably in environmental and natural resource management, but increasingly in public health in multiple settings and scales around the globe. This paper explores how to transfer knowledge about collaborative governance to challenging public health settings found in low-income and moderate-income countries (LMICs). Its primary goals are to provide a general grounding in the field of collaborative governance as developed, point to relevant empirical findings on the performance of collaborative governance and lay out some of the challenges and considerations for thinking about how to improve collaborative public health governance in LMICs.

It is not an overstatement to suggest that the field of public administration in the USA is being transformed by its increasing reliance on multisector collaborative governance. No longer is ‘government’ the primary engine of change in matters of public policy, management and service delivery. For the past 20 years and more, the public’s business has increasingly been carried out by a complex array of hybrid institutional arrangements across public, private and non-profit sectors.1 2 This has occurred in large part in response to the ‘hollowing out’ of government where reduced personnel, capacity and resources have necessitated a shift to contracting out, partnering and leveraging support for other multisector performance.3 The corresponding rise in the non-profit sector has been fed by this shift from government to governance.4

Those who study public administration have clearly responded to this transformation as evidenced by the predominance of recent scholarship in contract management, interagency cooperation, public-private partnerships, public service networks, collaborative management and collaborative leadership. Concerns over accountability and performance have also heightened as the efficiency, effectiveness and equity of these multiorganisational, cross-jurisdictional arrangements become harder to measure and manage.

Collaborative governance has become a useful covering term for the study and practice of cross-sector collaboration. Whether ‘sector’ refers to the public, private for-profit and non-profit arenas, or to different public policy domains, the concepts, challenges and opportunities for cross-boundary collaborative systems are similar. A variety of frameworks have developed to chart different approaches to this emerging phenomenon, ranging from network analyses that emphasise the structure of relationships among organisations,5 6 to process models that focus on interpersonal dynamics and capacity building,7 to negotiation approaches that emphasise bargaining and conflict management,8 9 to contingency performance models that incorporate antecedent conditions and essential inputs for productive collaborative outcomes.10–12 All of these conceptual lenses contribute to our understanding of collaborative governance writ large.

The earliest academic use of ‘collaborative governance’ focused on how private or corporate sector expertise and knowledge could be drawn on to improve performance in the public sector.13 14 Associated early scholarship in Europe explored distributed collaborative governance with a focus on different subnational approaches to public-private partnerships and other governing arrangements that incorporated corporate and community stakeholders.15 ‘Collaborative governance’ was further developed by Ansell and Gash in 2008 in their widely cited article in the Journal of Policy Administration Research and Theory, where they analysed 137 documented cases from a broadly ranging literature.11 They defined collaborative governance as ‘a governing arrangement where one or more public agencies directly engage non-state stakeholders in a collective decision-making process that is formal, consensus oriented and deliberative and that aims to make or implement public policy or manage public programmes or assets.’11, p. 544

Emerson et al 10 and Emerson and Nabatchi16 built on Ansell and Gash’s definition but expanded it to cover a broader suite of agents, structures, processes and actions that enable collaboration across organisations, jurisdictions and sectors. They emphasise the cross-boundary nature of what Kettl described as ‘the collaborative imperative’, and do not limit collaborative governance to government-initiated efforts.4 Rather, collaborative governance includes institutional forms that extend beyond the conventional focus on the public manager or public sector. Furthermore, collaborative governance may, to varying degrees, involve broader public participation and civic engagement when public education and support is an essential component for effective collaborative planning and implementation.

Specifically, Emerson et al define collaborative governance as ‘the processes and structures of public policy decision making and management that engage people across the boundaries of public agencies, levels of government and/or the public, private and civic spheres to carry out a public purpose that could not otherwise be accomplished’.10, p. 3 This broad definition of collaborative governance provides the basis for their integrative framework that is enabling comparative and aggregate empirical analyses across diverse theoretical, normative and applied perspectives. This framework is depicted in figure 1 and will be further parsed in the ‘What we know from research on collaborative governance’ section.

Figure 1

Integrated framework for collaborative governance regime. Source: Emerson and Nabatchi.16

What we know from research on collaborative governance

It is important to underscore that collaborating across boundaries began in the field, not in the lab. Strides in practice have been made around the world for many years without the benefit of theorists or conceptual frameworks. Most of the lessons learnt have come from cutting-edge practitioners and managers. Scholars are still catching up with what some would call a governance movement. But in the past couple of decades there has been a substantial increase in empirical work by scholars in public administration and management. Much of the practice wisdom is being validated, while important unexpected observations and questions are being raised as well. Past limitations on research methodology (eg, focus on single cases, limited time frames for study, difficulty in measuring process variables and conceptualising performance measures, etc) are slowly being surmounted and we have a growing number of large-N studies and more robust analytic approaches to aggregate data being use in studies of collaborative governance.11 17–20 What follows is a description of recent empirical findings from public administration research on collaborative governance organised in accordance with the Emerson and Nabatchi framework.

System context and drivers

Collaborative governance is situated within a dynamic system context that can include resource or service conditions, policy and legal frameworks, socioeconomic and cultural characteristics, network characteristics, political dynamics and power relations and the history of conflict, among other dimensions.16 In particular, prior relationships and existing networks as well as the institutional context matter as they shape opportunities and constraints, and influence if, how and when collaborative governance unfolds and operates.21–24

Scholars have studied the specific conditions or drivers that are essential to the formation of collaborative systems or, what Emerson and Nabatchi refer to as collaborative governance regimes (CGRs). Four factors have surfaced in the research as significant combined drivers of CGRs:

Uncertainty about the nature of a problem can drive groups to work together to understand and define the problem6 22 and how to approach it in order to reduce, diffuse and share risk.

Interdependence among organisations who are unable to accomplish goals on their own turn to other organisations to foster collective action through mutual reliance.25–28

Consequential incentives that make acceptance of the status quo undesirable drive cross-boundary collaboration.11 These incentives may be internal pressures (salient issues, resource needs, interests or opportunities) or external pressures (situational or institutional crises, threats or opportunities).16

Initiating leadership matters in motivating the preliminary engagement of participants in a potential CGR. Boundary-spanning capacities, in particular, can be helpful in framing issues for diverse participants and creating the right conditions to launch collaboration dynamics.28 29

Collaboration dynamics

Once initiated, the dynamics of cross-boundary collaboration unfold as iterative interactions between behavioural elements (principled engagement), interpersonal elements (shared motivation) and functional elements (capacity for joint action).16 Research suggests that collaboration dynamics foster desired actions and outcomes and that each of these elements contribute to performance.

Principled engagement

People collaborate and when they engage in a purposeful manner, their primary collaborative actions or behaviours span the discovery of their common and different interests; the joint definition of the problem or challenge they face; open deliberation about their shared interests and how to effectuate change of some kind and shared determinations or decisions they make (short-term and long-term, substantive and procedural) along the way.16 30 These collaborative behaviours of principled engagement can be reinforced and strengthened through a number of factors, according to empirical research, for example:

Skillful communication is understood to be the substrate of collaboration.31

Collaboration fosters individual and group learning.8 Collective learning, the process itself and the products or new knowledge generated, contributes to collaborative performance.32 33

Conflict should be expected among diverse groups and organisations, especially around differential status, but conflict may be more salient at the outset rather than later in the process.34–36

Most scholars and practitioners advance some form of consensus decision rule (although not always necessarily full unanimity) for collaborative determinations that are fairer, more durable and effective.37–39

Shared motivation

Principled engagement fosters and is strengthened by the interpersonal dynamics of shared motivation. The motivation of individuals and organisations to come to the table and stay at the table hinges on the relational dimensions of trust, mutual understanding, legitimacy and commitment.13 Consistent with the research on social capital, interpersonal relationships matter and that starts with trust.40 41 Trust has long been a sine qua non of collaboration.6 42–44 It lays the foundation for mutual understanding of each other’s common interests, and of their distinct and important differences. This recognition of unique and common perspectives and interests contributes to an appreciation for the internal legitimacy of the group26 45 46 and ultimately to the necessary commitment of individuals and organisations to the collective action required.11 Underscoring the value of these relational elements of shared motivation are research findings, such as that:

Trust building is essential, whether one starts with or without trusting relationships.47–49

Legitimacy—internal and external—contributes significantly to successful collaborative performance.50

Capacity for joint action

Collaboration dynamics also call on functional elements of procedural and institutional arrangements, leadership, resources and knowledge. Comparative studies have underscored the importance of collaborative capacity to collaborative actions and outcomes51 and as reinforcing principled engagement and shared motivation over time.13 Procedural norms of reciprocity and rules for self-governing are essential to cross-boundary governance.26 52 53 Protocols and more formal institutional arrangements are important supplements when the collaborative arena or network is complex with many participants.3 6 54 Leadership continues to be a central capacity11 55 that includes facilitators and managers, and the participants themselves as they represent their organisations or constituencies, public decision makers and technical and scientific experts.56 57 Knowledge, both explicit and tacit, may be the currency of collaboration16 and is essential for collaboration partners to assemble and share to inform their determinations and produce effective collaborative outcomes.58 Finally, resources, be they financial, in-kind, logistical, staffing, technical expertise, etc, have long been recognised as both essential to productive collaboration and more easily leveraged among the participants and their sponsors.5 26

Additional empirical findings of interest here include the following:

Collaborative structures tend to be more adaptive and changeable than conventional institutional structures.59–61

Changes in leadership and participants over time need to be anticipated and managed.62–64

Integrative leadership characteristics can contribute to productive collaboration.29 65 66

Differential power relations and dominating power exerted through formal authority are important conditions to address.67 68

Collaborative actions, outcomes and adaptation

Collaborative governance regimes are instrumental by definition. They must produce public value and benefit participating organisations, those that have sponsored or funded them, the directly affected stakeholders and the public at large. Through their collaboration dynamics, CGRs generate actions that in turn lead to outcomes and eventually to adaptation. Depending on the context, purpose and type of CGR, collaborative actions will vary, ranging from strategic outputs such as securing external endorsements; providing information or training to constituents or the public; enacting policy measures, laws or regulations and marshalling external resources—to direct action—such as deploying staff; sitting or building facilities; issuing permits; cleaning or restoring environments; carrying out new management practices; monitoring implementation and enforcing compliance.10

The recent findings from empirical research on the performance of collaborative governance regimes are instructive. In a 2015 multicase study of marine and freshwater ecosystem restoration partnerships in the Puget Sound, Scott and Thomas found evidence supporting the contribution of principled engagement and capacity for joint action to reported increases in consultation, planning or implementation among CGR participant organisations and within larger networks.69 They also found that within principled engagement, three specific facets—awareness of other organisations’ interests and values, increased face-to-face communication and the use of common language—are associated with increases in network ties. Their measures of joint capacity also were associated with increases in network ties. Based on their study, the authors suggest that collaboration dynamics reduce transaction costs, which in turn helps form and strengthen ties between organisations in the CGR and those in other networks.

In another comprehensive study, Ulibarri looked at the effect of collaboration dynamics and its individual components on outcomes across 24 hydropower relicensing cases under the Federal Energy Regulatory Commission’s alternative or integrated licensing processes.18 Based on survey research with 270 participants, Ulibarri found evidence of reinforcing relationships between the three components of collaboration dynamics (principles engagement, shared motivation and capacity for joint action) and of their effect on decision making, participant satisfaction and the perceived quality of the license decision and predicted environmental and economic changes.

In a related study, Ulibarri compares three different cases with high, medium and low ratings on collaboration dynamics and found significant differences in the reported quality of the resulting environmental decisions.70 She also finds that high-quality collaboration led to jointly developed and highly implementable operating regimes designed to improve numerous resources, whereas lesser-quality collaboration resulted in operating requirements that ignored environmental concerns raised by stakeholders and lacked implementation provisions.

Finally, Scott, in one of the first major outcome performance studies of its kind, demonstrates through an examination of 357 watersheds in the USA, that collaborative watershed groups directly contribute to significant improvements in water chemistry and in-stream habitat conditions, as measured objectively by watershed data, not participant reports.20

Challenges and considerations for applying collaborative governance to public health in LMICs

In their paper in this volume, Bennett et al enumerate multiple challenges in applying collaborative governance generally and specifically to health in LMICs. Collaborative governance can be a resource-intensive practice, dependent on skillful leadership and process management, hard to implement when competition for resources is prevalent and when our understanding of the effectiveness of investing in these multisectoral approaches is limited as well. They point out that collaborative governance is even more challenging in LMICs with weak states, limited institutional infrastructure that is often highly hierarchical, political instability, limited resources and skills, limited leadership and social and political systems that often depend on patronage or ingrained corruption. All these challenges contribute to the lack of incentives to cross boundaries to work together, to try to innovate, to share accountability, let alone learn from practice through evaluation. It is rather extraordinary, given all these difficulties, that collaborative governance in LMICs has been attempted at all!

A reasonable reaction might be to turn back from the aspiration of collaborative governance and the higher bar of shared decision making and instead refocus on communication, coordination and cooperation—simpler, more straightforward approaches that keep the command and control systems in place, keep roles and responsibilities clear and reduce the messiness and risk of power sharing. This reticence might well be appropriate in certain settings, where crisis and emergency responses are required without the benefit of collaborative emergency preparedness. That said, it would be useful to test the short-term advantages of these safer approaches in other health issue contexts against the potential long-term benefits to the participants, the local communities and the social and institutional systems they rely on. If, indeed, shared responsibility for health improvements, empowerment of individuals and groups, strengthening local leadership and leveraging resources for joint action are important, then investing in CGRs would well be worthwhile.

It should be underscored here that there are also strengths and advantages that can be drawn on to support the effective design and implementation of collaborative governance approaches in this arena, not the least of which is the importance of the public health problems in LMICs. There is considerable salience and immediacy to these problems which can serve as an international, regional and in-country motivator for change. There are also multiple interdependent sectors engaging in solving these problems, including international governance bodies, country governments, NGOs, private corporations and individual entrepreneurial public investors. In every sector, there are committed leaders and advocates for collaborative engagement at different scales and specifically in LMICs. If collaborative governance approaches have been seen to add public value in other settings, then it seems reasonable, if not imperative, to further experiment and try to replicate this performance in this critical policy arena.

There is no question that collaborative governance is hard, even in higher-income countries. And it can be costly in terms of resources and time, and if a collaborative effort fails, the effects can be damaging and long-lived. Participants will be less likely to work together in the future, setting back any gains made in building social capital and collaborative learning. So, first rule of thumb—do no harm.

The challenge then is how, when appropriate, to make collaborative governance feasible in LMICs, where different political, economic and cultural systems prevail and where arguably there is less capacity for joint action. ‘When appropriate’ is an important caveat, as just mentioned, and cannot be as easily determined if we conceptualise the choice as simply binary, that is, to collaborate or not to collaborate. There are different types of collaborative governance (self-initiated; third-party convened and externally mandated) and they respond to different contexts and conditions. One type may be more appropriate than another. Or the four drivers of collaboration (uncertainty, interdependence, consequential incentives and initiating leadership) may not be in place, but perhaps they could be in time with proactive convening work among the parties. Or the presenting problem may call for integrating several different approaches, only one of which may be through a cross-boundary system.

The following general guidance is offered when studying or considering the potential value and limitations of applying cross-sector collaboration in LMICs.

Take a longer view and a systems approach

As we know, collaborative governance regimes are complex, multilevel systems that change over time. They combine structures and processes and agency into dynamic systems that can be formed in different ways and can evolve over time along different trajectories (figure 2).

Figure 2

Collaborative system.

One can think of these various system dimensions as contributing incentives, inspiration, interconnections and inter-relations as depicted in figure 2. There is no one right way of viewing these complex, social phenomena. The only wrong way is to view them through too narrow a lens. Cross-sector collaborative governance systems require mobilising and managing institutions, networks, processes and leadership. All of these dimensions are subject to change over time, institutions being the slowest to change, while leadership and processes are most dynamic and vulnerable to external shifts and events. Collaborative governance structures fit this dynamic environment and are more responsive and adaptive than formal organisational structures.62 But this requires of conveners, funders and managers a wide-angle perspective, patience and flexibility to balance, integrate and orchestrate these moving parts over time.

The collaborative governance of Uganda’s HIV response illustrates both the dynamic and evolutionary nature of multisectoral collaboration, and the complex ways in which the dimensions identified above may interact. Uganda’s response to HIV/AIDS, during the late 1980s and 1990s was widely acclaimed, but also rare in the extent to which it adopted a multisectoral approach.71 When the disease was first identified in Uganda in 1986, President Museveni responded by creating a National Committee for the Prevention of AIDS (NCPA), which he chaired and was situated within his own office. The NCPA held monthly meetings that brought together multiple ministries, donor representatives and increasingly over time, civil society organisations. In 1990, the government sought to institutionalise this approach through the establishment of Uganda’s AIDS Commission and its Secretariat, which was responsible for developing a new strategy for a multisectoral response to AIDS control.72 While this new Commission was tasked with coordinating activities, sharing information and monitoring and evaluation across sectors, it struggled with these tasks, perhaps lacking the clear authority that the NCPA, housed in the President’s office had possessed. Due to the disappointing performance of this Commission it was disestablished in 1997, but only to be re-established when an influential donor required a structure of this nature in order to be able to disburse funding.71 Despite the shifting institutional arrangements, the collaborative commitment to address the challenge of HIV across multiple sectors, and the networks that were established early on across ministries as diverse as health, defence, education, involving both government and non-government actors, were key to rapidly mobilise the response and address issues of stigma and discrimination which plagued HIV/AIDS programmes elsewhere in sub-Saharan Africa at this time.73

Use a design approach

All collaborative governance regimes start with an itch, a tension, a challenge, sometimes an opportunity. Thinking about the condition or problem from a design perspective leads inevitably to questions like—What do we know about this condition or problem? What do we need to know? Whom should we ask? How can they help us? Who should be involved in solving the problem? In other words, a careful assessment of the condition or problem and the system context in which it exists and was created must be conducted. Part of that system context is the political economy and the nature and sources of power and authority that can be brought to the table (or can hinder the initiative). Are they sufficient, and if unbalanced (which is always the case), how can one rebalance the power dynamic through design?

A second set of design questions focuses on the appropriateness of a collaborative approach—What is the added value or ‘collaborative advantage’ of collaboration versus other alternatives? Are the conditions sufficient to initiate a collaborative governance regime? Are there consequential incentives? Is there uncertainty? Is there a recognised interdependence among the stakeholders involved? Is there identifiable leadership to initiate collaboration? Is there sufficient capacity to engage partners or what would it take to assure that capacity?

There are also practical considerations and operational constraints that must be taken into consideration when designing for collaboration governance. The capacity for joint action develops over time but there are early transaction costs that must be overcome for CGRs to get underway. There are institutional requirements, be they laws, regulations, contracts or norms, often at multiple levels, to understood and comply with. In some cases, new institutions need to be created. There are often jurisdictional requirements to address. There are time frames required of collaborative governance that need to be accommodated. Important external deadlines or schedules, such as review periods, funding deadlines, election cycles, need to be considered. There are resource issues, staffing and logistics to plan for. These operational matters may be handled up front or in stages, depending on the nature of the initiating leaders and conveners. For self-initiating CGRs, the core participants share the support of these functions. For third-party assisted CGRs, a disinterested convener or funder steps up to handle the initial transaction costs of the collaboration. With externally directed CGRs, it is most often thegovernment that makes provision for collaboration and guides or induces participants to engage.

The Integrated Child Development Scheme (ICDS) in India offers insight into the challenges that may arise if a design approach is not taken, and inadequate attention paid to local issues of power and authority. The ICDS was initiated >40 years ago with a mandate to provide six connected services: immunisation, supplementary nutrition, health check-ups, referral services, preschool non-formal education and nutrition and health information. These services are supposed to be provided by community-level workers focused on nutrition and childhood development (anganwadi workers (AWW)) and health (auxiliary nurse/midwives (ANM)). As a flagship project, the government has invested heavily, spending approximately US$5.4 billion on the programme between 2000 and 2010.74 While there is some evidence of positive impact of the ICDS programme, India remains one of the worst performing countries in terms of childhood nutrition and a recent evaluation indicated serious and persistent design flaws that have undermined the programme, and proved quite intractable.75

Historically, evaluations have highlighted the lack of infrastructure for the scheme, and failure to provide adequate food commodities; however, the 2011 evaluation explored further the underlying reasons for these failures, drawing attention to:

the lack of awareness among beneficiaries—in one assessment only 3% of potential beneficiaries could identify a specific centre or worker associated with the scheme;

ineffective coordination committees, and the fact that frequently such committees were centrally concerned about how they as individuals could benefit from the Scheme;

widespread corruption, leakage of funds and patronage;

the fact that AWWs are only minimally trained (they receive just 3 months training), and ill equipped to coordinate with other services, and if AWWs were to perform fully their duties (from supporting nutrition to preschool education), they would face excessive demands on their time.

While the lack of coordination between the health sector workers (ANMs) and the AWWs working for the ICDS scheme is clearly problematic, more deeply rooted challenges related to design of the scheme and its fit with the local context, notably the lack of accountability of the scheme to local communities, the power relations between the largely female AWWs and ANMs on the one hand, and village leaders on the other hand, and the significant resources made available by the government for food commodities, have often critically undermined programme functioning.

Focus on leadership

Do not misspecify or underestimate the leadership requirements of collaboration. Across all the research on collaborative governance, the need for effective leadership is found to be critical. An excellent study by O’Leary et al on the skillset of collaborative managers offers valuable insights into the attributes and skills required to effectively manage in these complex settings.76 Individual attributes, such as openness, patience and self-confidence, and interpersonal skills, such as being a good communicator and listener, were mentioned most frequently among senior level federal officials in the USA responding to survey and interview questions about effective collaborative leaders. Surprisingly, these ‘soft’ skills were viewed as essential for successfully collaborative leaders, well ahead of group process skills, such as facilitation, negotiation and collaborative problem-solving; strategic thinking skills and substantive and technical expertise and knowledge. This makes sense in light of the research on the importance of trust and trusting relationships to collaboration previously noted.30 47–49 O’Leary et al are now beginning to study civil service officials in other countries to see if such findings are culturally specific.

These findings underscore that collaborative leadership is more than skillful execution of tasks. It is about building relationships and trust with standing as well, important elements that are needed to initiate collaboration and to foster in partners over time. Effective collaborative leadership must be demonstrated and shared by the convener or manager of a collaboration and by all the participants who must represent their organisations, networks or constituencies. Multiple leaders and multiple kinds of leadership skills are needed throughout the development and sustainment of a collaborative governance regime. If collaborative leadership is lacking at the various scales where collaboration is operating, then there may need to be some front-end, possibly sustained investment in leadership training, mentoring and knowledge/awareness building before moving forward.

Research also tells us conflict management skills are important for effective leadership in collaborative governance.34–36 By definition, collaborative governance engages participants with different interests and values, roles and resources and relationships to the services being sought. While coming together for a common purpose and finding common ground, these differences are not meant to be stifled or suppressed, but rather surfaced and expressed so that the full group can understand the full dimensions of the issue at hand and the differential impacts and challenges being faced. Competition and power differentials can exacerbate these differences. Multicultural settings are also likely to present additional conflict challenges. Handling these potential conflicts by acknowledging and honouring differences is an essential skillset for shifting gears into joint problem solving for mutual gains.28 38 56

One example of the centrality of leadership to effective multisectoral collaboration comes from the process of negotiating the Framework Convention on Tobacco Control (FCTC). The FCTC is the only treaty developed under the auspices of WHO, and was established through an extensive consultative process involving all member states, as well as representatives of Ministries of Health, Trade and Industry, Finance and state-owned tobacco producers.77 The Brazilian government was widely acknowledged as providing critical leadership for the treaty.78 Despite the fact that Brazil hosted a sizeable tobacco industry it had already taken multiple progressive steps to curb tobacco use. It gained credibility and standing among other countries on the basis of its own tobacco policies, as well as the role that it had previously played in brokering other global health agreements, for example, around access to antiretroviral drugs for people living with HIV/AIDS. Multiple individuals from the Brazilian government took leadership roles in the FCTC process.78 The former coordinator of the Brazilian National TB Control Programme was recruited to lead the WHO’s Tobacco Free Initiative, and the Intergovernmental Negotiating Body that conducted the FCTC negotiations was chaired by one, then another, Brazilian diplomat. The Brazilian delegation to the negotiations were also critical in advocating strongly for key aspects of the convention, such as the use of large graphic designs on cigarette packs, and the establishment of regulatory bodies to oversee tobacco content and emissions. Analysts of the FCTC case have commented on how critical support from the Brazilian Foreign Affairs ministry was to the success of the negotiations, and in particular have drawn attention to the diplomatic and coalition building skills of the top diplomats who chaired the negotiations. These individuals spearheaded Brazil’s strategy to build regional consensus about the FCTC within Latin America, and then build linkages across regions. The same group of diplomats was also critical in drawing civil society into the negotiation, which ultimately played a critical role in pressuring governments to support the treaty.

Conclusion

The complex challenges for delivering public health in LMICs around the world are daunting. Cross-sector collaboration offers a promising path forward, but comes with challenges of its own that require informed design strategies, management skills and leadership.

This overview of collaboration governance as studied by public administration scholars summarises key empirical findings from a variety of policy contexts outside the public health field. While general guidance is offered on the application of collaborative governance to public health provision in LMICs, there remains a gap in evidence-based research in this context. There are promising examples of collaborative governance at work in this arena. However, more systematic research is needed. It is hoped that this volume will stimulate a future research agenda to carefully evaluate the prospects and performance of collaborative governance in public health delivery in LMICs. The potential for creating public value and solving systemic public health problems is substantial and well worth the long-term investment in this understudied area.