Whole-of-society approach for public health policymaking: a case study of polycentric governance from Quebec, Canada

Transcription

1 Ann. N.Y. Acad. Sci. ISSN ANNALS OF THE NEW YORK ACADEMY OF SCIENCES Issue: Paths of Convergence for Agriculture, Health, and Wealth : a case study of polycentric governance from Quebec, Canada Nii A. Addy, 1 Alain Poirier, 2 Chantal Blouin, 3 Nick Drager, 1,4 and Laurette Dubé 1 1 McGill Centre for the Convergence of Health and Economics (MCCHE) and the Desautels Faculty of Management, McGill University, Montréal, Québec, Canada. 2 Institut national de santé publique du Québec, Montréal, Québec, Canada. 3 Agri-food and health, Institut national de santé publique du Québec, Québec City, Québec, Canada. 4 London School of Hygiene and Tropical Medicine, London, England Address for correspondence: Nii A. Addy, Desautels Faculty of Management, McGill University, 1001 Sherbrooke Street West, Montréal, Québec, Canada H3A 1G5. nii.addy@mcgill.ca In adopting a whole-of-society (WoS) approach that engages multiple stakeholders in public health policies across contexts, the authors propose that effective governance presents a challenge. The purpose of this paper is to highlight a case for how polycentric governance underlying the WoS approach is already functioning, while outlining an agenda to enable adaptive learning for improving such governance processes. Drawing upon a case study from Quebec, Canada, we employ empirically developed concepts from extensive, decades-long work of the 2009 Nobel laureate Elinor Ostrom in the governance of policy in nonhealth domains to analyze early efforts at polycentric governance in policies around overnutrition, highlighting interactions between international, domestic, state and nonstate actors and processes. Using information from primary and secondary sources, we analyze the emergence of the broader policy context of Quebec s public health system in the 20th century. We present a microsituational analysis of the WoS approach for Quebec s 21st century policies on healthy lifestyles, emphasizing the role of governance at the community level. We argue for rethinking prescriptive policy analysis of the 20th century, proposing an agenda for diagnostic policy analysis, which explicates the multiple sets of actors and interacting variables shaping polycentric governance for operationalizing the WoS approach to policymaking in specific contexts. Keywords: diagnostic policy analysis; multi-stakeholder partnerships; polycentric governance; public health policy; whole-of-society Introduction: Addressing governance challenges in public health policymaking As civil society and private sector actors partner with governments, employing participatory and multistakeholder approaches to policymaking for addressing societal issues, 1,2 challenges in governance have emerged. Specifically, in public health policymaking, adaptable and dynamic governance is needed to address the complexity and rapid change that characterizes interactions between diverse stakeholders, who sometimes have conflicting goals and priorities in the face of scarce economic resources, whether in industrialized or emerging economies. 3 For example, in promoting healthy diets as part of global public health policies, effective governance of partnerships involving multinational food companies, which may have previously profited from marketing unhealthy foods, and community organizations across various countries presents a real challenge. Our fundamental task in the 21st century is to address such challenges, in contexts where societies are also attempting to implement universal healthcare policies, 4 while heeding calls of global organizations such as the World Health Organization (WHO) for adopting a whole-of-society (WoS) approach, involving stakeholders from international agencies, states, for-profits, and community organizations. 5 Yet, questions remain about how policymaking processes are governed in the WoS approach, especially across contexts that vary over space and doi: /nyas Ann. N.Y. Acad. Sci (2014) C 2014 New York Academy of Sciences.

2 time. For governing the multiplicity of actors in WoS policy efforts, Dubé et al. propose polycentric governance, 6 8 in which there are multiple centers of decision making occurring simultaneously, 9 as individual and collective actors from governments, markets, and civil society, across levels (i.e., global, national, provincial/state, and local) and domains (e.g., health, agriculture, and information technology) self-organize and experimentally create institutional arrangements on the basis of their diverse expertise and perspectives, developing public health solutions over time. 10 The purpose of this paper is to highlight a case of how polycentric governance underlying the WoS approach is already functioning, while outlining an agenda for analysis to enable adaptive learning for improving such governance processes. Drawing upon a case-study from the Canadian province of Quebec, we employ empirically developed concepts from extensive, decades-long work of the 2009 Nobel laureate Elinor Ostrom in the governance of policy in nonhealth domains such as the provision of public safety and the field of environment 8 to analyze early efforts at polycentric governance in policies around overnutrition, highlighting interactions between international and domestic, state and nonstate actors, and processes. We use information from primary and secondary sources, including studies conducted by this paper s authors, 11,12 reports and academic literature, as well as knowledge from this paper s authors in Quebec s public health field. Our paper complements other work that is being conducted on developing roadmaps for WoS, including studies of innovative approaches to multistakeholder partnerships (MSPs) for addressing nutrition, 13 models for the engagement of private sector, 14 and the challenges of developing metrics and analytics in such efforts. Whereas the other WoS roadmap papers noted here focus on specific initiatives, we analyze the broader-scoped, decades-long project of policymaking, highlighting the challenge of governance, and the rethinking of policy analysis that is needed as part of a WoS paradigm shift that Dubé et al. propose. 15 In the sections that follow, we highlight the multiple, cross-sectoral, multilevel centers of decisionmaking in policymaking over time, first analyzing the emergence of the broader policy context of Quebec s public health system in the 20th century, with the whole-of-government (WoG) approach adopted by the early 2000s. Second, we present a microsituational analysis of the WoS approach for Quebec s 21st century policies on healthy lifestyles, emphasizing the role of governance at the community level. Third, given our case-study, we argue for departing from prescriptive policy analysis, which has been the dominant paradigm in the social sciences during the 20th century, 8 and propose an agenda for diagnostic policy analysis, which explicates the multiple sets of actors and interacting variables shaping polycentric governance for operationalizing the WoS approach to policymaking in specific contexts. Framework for analyzing polycentric governance of policy In exploring polycentric governance underlying the WoS approach for public health policymaking, this paper responds to the call by Ostrom and her colleagues for empirical social science analysis that more completely captures the complexity of policy processes in an increasingly democratic world. 16 Polycentricity is grounded in decades-long empirical research, from 1960s to present including hundreds of case studies, lab and field experiments, and meta-analyses conducted and replicated in different contexts to understand the governance of public policy in nonhealth domains, such as the water industry, and public safety in communities; and common-pool resources, such as lakes, fisheries, forests, and irrigation systems The research found that, contrary to dominant ideas at the time about the effectiveness of centralized governance for getting diverse actors to cooperate in collective action, the multiplicity of governance structures in polycentricity led to more effective public service provision and use. It was not necessarily the decentralized nature of governance that led to greater effectiveness, and it was not possible to specify general reasons as to why governance was effective in the polycentric systems studied. Rather, the interaction of broader contextual and microsituational variables which drive individual behavior during policy implementation led to governance effectiveness, highlighting the importance of diagnostic policy analysis for understanding polycentric systems in any given context (Fig. 1), 8,31 as opposed to determinist approaches. Specifically, within the broader context and microsituations, individual and collective actors interacting repeatedly with each other during policymaking Ann. N.Y. Acad. Sci (2014) C 2014 New York Academy of Sciences. 217

3 Figure 1. Broader contexts and microsituations shaping polycentric governance of policy. 3 learnandadopt norms internal valuations that are negatively or positively related to their actions, such as dishonesty when interacting with others in a given situation, 32 while the level of trust acts as a rule of thumb a heuristic to guide choices that actors make among alternative norms (e.g., trusting another actor promotes honesty in interactions with them), leading to increased or reduced cooperation and with net benefits that are realized influencing subsequent interactions. 31 Thus, we argue that during public health policy development and implementation in a given context, individuals initiating cooperationinrepeatedsituationsaspartofmsp decision-making units may gain a reputation as being trustworthy, with others reciprocating, further increasing the levels of cooperation, and with benefits accruing to MSP members, and subsequently reinforcing levels of trust and reciprocity in implementing public health policy. 20th century: Emergence of the broader context In analyzing the broader context within which public policies emerge, E. Ostrom 8 notes that given multiple levels of contexts in which social systems can be embedded, analysts should identify the appropriate system and variables to address a given question. 33 With our focus on questions of governance for policymaking, among variables that Ostrom notes, those considered in the Quebec case include the social, economic, and political settings within which policy emerges; the use/provision of the public good (i.e., actors involved in health care); the ecosystems related to the public good; and the governance systems (Table 1). A historical perspective highlights how the interactions of these variables at global, national, provincial, regional, and local levels during the 20th century shaped the broader context out of which Quebec s public health policies emerged by the early 21st century. During the period from the 1920s to the 1970s, Quebec underwent socioeconomic transformation, with a transition from nonstate (i.e., Catholic Church) to a state provision for health. Particularly, in the 1960s, the province embarked on a Quiet Revolution (Révolution tranquille), leading to the rise of a welfare state. The state gained economic power by nationalizing production of hydroelectricity, and took a more central role in financing and providing social services, including health and education. The leadership of individual and collective actors across sectors and levels was vital in creating institutional mechanisms for collaboration between state and nonstate actors during this period. 34 By the 1970s, the users of the health system expanded as Quebec adopted universal health insurance, driven by federal social welfare approaches, in which the federal government provided matched funding to provinces that adopted its policies. 35 From the 1970s to the 1990s, multisectoral approaches to health policies were promoted by global, federal, and provincial actors, 36 with interactions between factors highlighting linkages between health and other ecosystems, such as technology, and food and agriculture. 37 A significant global event that influenced policies in Quebec and other regions of the world occurred in 1986, when the first of a series of conferences towards achieving health for all by 2000 was held in Ottawa, Canada. 38 There, leaders from around the world highlighted the need for multisectoral approaches to tackling health issues, 39,40 captured in the Ottawa Charter for Health Promotion, which took inspiration from WHO s 1946 constitution; 41 the 1974 Marc Lalonde Report that had been spearheaded by a former Canadian Minister of Health (a Quebecer); 42 and the 1978 Alma-Ata Declaration. 43 Around the world, including Quebec, the Ottawa Charter became a reference point for public health actors as they attempted to place health on the policy agenda 218 Ann. N.Y. Acad. Sci (2014) C 2014 New York Academy of Sciences.

4 Table 1. Summary of the broader context of Quebec public health, 20th to early 21st century Sets of variables Quebec case Social, economic, and political settings Economic development and emergence of a welfare state by 1970s. Government policies prioritizing public health through late 20th century and early 2000s. Political factors and financing gaps leading to partnerships between governments and private foundations to fund public health policy implementation in early 2000s. Public health actors Interactions between leadership of global, federal, provincial, and regional actors in expanding public health through universal coverage. Norms of state provision of health care. Low state financing levels for public health did not reflect importance noted in rhetoric. Related ecosystems Technology (e.g., hydroelectricityproduction); finance; education; food and agriculture. Governance system Multiple government agencies engaged in public health, across levels. Multiple nongovernment organizations (private foundation, provincial, regional, community organizations, few for-profits). Network structure (multinodal with government and private foundations at a high level; community networks as central actors at a low level). Local collective-choice autonomy (community networks are recognized as key decision-making centers). Monitoring and sanctioning processes (formal and informal monitoring, including holding public officials accountable). of other related ecosystems, 44 in alignment with Health in All Policies (HiAP) efforts globally. 45 From a multilevel perspective, the period saw a transformation of governance for policy implementation in Quebec from hospital- to regional-level networks to facilitate interaction with other sectoral regional bodies; as well as the development of health priorities that set the stage for further formalizing multisectoral approaches. Specifically, research efforts at provincial, federal, and global levels informed the development of a Public Health Act (PHA) in 2001 and a provincial action plan, 46 establishing the policy context for 21st century public health approaches in the province. 47 The PHA spelled out intersectoral actions, with governance roles at multiple levels: provincial actors were to coordinate with regional counterparts, who were in turn to develop action plans and coordinate with community actors, and vice-versa, for policy implementation. The adoption of the PHA by the 2000s exemplified a WoG approach, defined by the OECD as one where a government actively uses formal and/or informal networks across the different agencies within that government to coordinate the design and implementation of the range of interventions that the government s agencies will be making in order to increase the effectiveness of those interventions in achieving the desired objectives. 48 The main program implementing the PHA at local, regional, and provincial levels was the Public Health Program (PHP) adopted in 2002, for the period The PHP specified intersectoral action among its strategies, while also noting the need to strengthen the capacity of individuals and vulnerable groups, and the development of communities to achieve collective public health goals. 50 Conflicting priorities Analysis of public health in Quebec highlights that despite the perceived benefits of multisectoral and multilevel approaches, implementation was faced with challenges. Particularly, as has characterized WoG attempts elsewhere, reconciling sometimes conflicting priorities of various actors often led to policy incoherence In Quebec s health system, Ann. N.Y. Acad. Sci (2014) C 2014 New York Academy of Sciences. 219

5 a major challenge has been to reconcile conflicts in the priorities of public health experts who focus on prevention, versus clinicians priorities, such as treatment; as well as differences in the priorities of nonhealth sectors, particularly in the face of competition for limited financial and other resources. 54 For example, annually, about 265 million dollars, representing only 2.3% of the health and social services budget, had been devoted to prevention, suggesting the low priority it had been accorded. In , public health leaders negotiated a commitment by government to increase the percentage, in alignment with prioritizing public health. 46 However, political and economic factors interacted to challenge the system, ultimately transforming its governance structures. Despite the commitment of the then governing party (Parti Québécois) to fund prevention and the implementation of the PHP, its loss in elections that were held in 2003 precipitated an audit, which indicated budget deficits, resulting in budget cuts to the health sector, and leaving the implementation of the public health policy underfunded. 55 As we present in the section that follows, the inability of government actors to develop coherent public health policies that were adequately resourced and implemented led to the formal engagement of private sector and civil society actors in partnerships to promote public health. The involvement of such nonstate actors reflected global trends that called for government partnerships with the private sector and civil society for health promotion, 56 with benefits and challenges. In addition to nonstate actors supplementing financial and other resources from the state, the former potentially provide monitoring capacity in holding public officials accountable for addressing public needs, even beyond election cycles. Yet, the complexity of governance potentially increases, with conflicts in priorities of health and nonhealth systems becoming compounded by differences between state and nonstate actors, issues that our microsituational analysis of the Quebec case highlights in the next section. 21st century: Microsituational dynamics of polycentric governance Given the context within which public health was situated in Quebec by the beginning of the 21st century, how did polycentric governance for the WoSapproachoperatefromamicrosituationalperspective? In the Quebec case, global, national, and provincial level policymaking processes, regional and local level development of action plans, and MSPs at various levels exemplify the multiple loci of decision making in the polycentric system that characterized the province s WoS approach. With various potential nodes for microsituational analysis, we focus on the implementation of Quebec s governmental action plan (GAP) on lifestyles and weight-related problems, which the government developed within the framework of the 2001 PHA. The GAP, which spanned the period , acknowledged complexity and the roles of multiple stakeholders, expressing the need for multisectoral expertise from state and nonstate actors, thereby further building upon 20th century multisectoral approaches. Illustrating the recognition of the embedded nature of human action in complex systems, the GAP involved eight different ministries and their partners, and identified 21 of 75 actions related to healthy eating each led by a specified government agency under two main themes: (i) food production, transformation, and distribution; and (ii) food security. The GAP also identified healthy eating actions in four settings: family and daycare services, schools, municipalities, and catering. It acknowledged that young adults food choices are mediated by the proximity of shops that sell healthy foods, necessitating the development of community programs that make healthy foods more accessible. We analyze the deployment of the GAP, guided by the framework for multisituational analysis that Poteete, Janssen, and E. Ostrom provide. 31 They note six microsituational variables that are associated with effective polycentric governance, as they have been found to increase trust and cooperation. 8 They are summarized for the Quebec case in Table 2, and elaborated upon below. All the factors potentially interact, contributing to increasing or decreasing the effectiveness of polycentric governance in policy development and implementation. First, where each actor recognizes that the marginal per capita return (MPCR) of cooperation for themselves and others participating in partnerships is high, they are more likely to cooperate in MSPs for policy, as they recognize the value that each participant respectively brings. In the Quebec case, despite the stated desire of engaging individuals from nonstate sectors of society in publichealth decision making and action, such broad 220 Ann. N.Y. Acad. Sci (2014) C 2014 New York Academy of Sciences.

6 Table 2. Summary of microsituational factors for 21st century Quebec public health policy Sets of variables Marginal per capita return (MPCR) of cooperation Reputation of participants Time horizon Feasibility of communication with full set of participants Entry or exit capabilities Agreed-upon sanctioning capabilities Quebec case Private foundation, government, community organizations have high MPCR, thus they cooperate to implement policy In the context of a welfare state, the government has a reputation as leader in policy implementation. Private foundation that initiated MSPs with government has been gaining reputation as cooperative and trustworthy, given history of supporting community organizations. Empirical questions remain as to what the perceptions of reputation are among the various state and nonstate actors. Ten-year commitment by private foundations and governments MSPs include private foundation, government, and community groups (although community groups are diffuse and relatively less represented at high-level administration); private sector (i.e., for-profits) are mostly absent. Repeated face-to-face communication occurs most frequently at community level. Decision makers from provincial, federal, and global levels are relatively removed from community-level participants. Community groups face constraints in capacity, time, and resources that facilitate communication Participants can enter and exit partnerships at low cost. Community groups develop their own sanctions. participation was largely not realized until lack of funding led to the engagement of private sector actors, suggesting a high MPCR of cooperation that the various stakeholders recognized in themselves and in each other, providing motivation for these various stakeholders to cooperate. Notably, while deploying its policies on lifestyles and weightrelated problems without committed financing in the 2000s, the government was publicly called to commit resources to public health by the Lucie and André Chagnon Foundation the largest private family foundation in Canada, with $1.4 billion in assets, established by a Quebec family that had been financially successful from developing a telecommunications firm, and chose to contribute to public good in the province. a The foundation was willing to invest millions of dollars into obesity prevention if the government was to invest an equivalent sum. Given the context of Quebec as a welfare state, partnership with government was in the foundation s best interest for achieving its goals, as noted a See rait-of-the-founding-family.aspx by representatives of the foundation (interview of August 14, 2013 with Jean-Marc Chouinard, Vice- President of Strategy and Partnerships and Alexis Gagné, Strategic Analyst): We found that in our Quebec context, collaboration with government was unavoidable, as it has resources already engaged in some large scale programs around healthy lifestyle. Also, partnership with government provides some legitimacy. We partnered with government to work on demonstrations along with them, as that is more powerful than an approach where we try to demonstrate the effectiveness of initiatives on our own before seeking their approval. The partnership between the Chagnon Foundation and the government also reflect the second and third sets of microsituational factors considered for this case: even where there is no prior interaction, having knowledge of a participant s positive reputation from past history increases cooperation; and longer time horizons provide participants with reason to anticipate that they will benefit from Ann. N.Y. Acad. Sci (2014) C 2014 New York Academy of Sciences. 221

7 cooperating in MSPs. In the context of Quebec as a welfare state, the reputation of the government as the key provider of public health made it an obvious leader in partnership with the Chagnon Foundation. In 2002, as part of an agreement between the Chagnon Foundation and the Quebec government, a nongovernmental organization, Québecen Forme (QEF), b was established as a mechanism for channeling funds to support community level actors to conceive, develop, and implement activities for healthy living at the local, regional, and provincial levels, with an annual contribution of 4 million Canadian dollars. Within the framework provided by the GAP, the Chagnon Foundation entered into negotiations with the Quebec government, and the two parties undertook to contribute a total of 480 million Canadian dollars over a 10-year time horizon, from 2007 to 2017 towards GAP deployment, c promoting healthy living among children and youth in the province. A key feature of the partnership is a focus on community-based governance, with 75% of funding focused on community-level, or local projects. A fourth set of microsituational variables regards the feasibility of communication between the full set of participants around policies. Where such communication is feasible in a MSP, notably through face-to-face interactions, facial expressions, actions, and words are used by participants to judge others trustworthiness in cooperating around policies. In the Quebec case, as in other contexts, there is heterogeneity in participation at various levels, and there have been limits in communication between potential partners, especially community and private sector actors. At a higher level, as exemplified by the provincial administration of QEF, governance is by an administrative board comprising six members each from the government of Québec and the Chagnon Foundation, with no direct representation by community groups, which are diffuse. However, a focal aspect of partnership is at the community level, where there are local partner groups formally called Regroupements locaux de partenaires b See tory.aspx & rapports-annuels.aspx c See Québec en Forme (2012) Historique: quebecenforme.org/a-propos/historique.aspx Figure 2. Sectors of Québec en Forme s partner organizations (2011). (RLP) that are cross-sectoral groups of organizations supported and funded by QEF on the basis of a strategic plan developed from a self-diagnosis of their community. The organizations are motivated to form local partnership groups in order to obtain financial resources and technical support, as well as to achieve their respective RLP objectives, 57 with a mandate to evaluate the resources and needs of the community, and develop and coordinate the implementation of local physical activity and healthy food programs funded by QEF. The RLPs typically include municipal organizations, agencies concerned with health and social services, education, recreation and sports, voluntary organizations, as well as private sector actors, such as businesses. However, there are very few private sector organizations among these partner organizations, an issue that we discuss later. For example, in 2011 only 1% of 2923 QEF partner organizations were from the private sector (Fig. 2). Community-level partnerships have especially been instrumental in enabling repeated face-to-face interactions between MSP participants, including individuals from civil society and government (e.g., at municipal level), and the private sector to a lesser extent (e.g., private sector partners such as vendors of healthy foods), enabling trust, and with goals converging over time. As part of a study being conducted in 2014 to better understand decisionmaking processes in Québec s community groups, 12 one community leader notes the following about the 222 Ann. N.Y. Acad. Sci (2014) C 2014 New York Academy of Sciences.

8 Table 3. Québec en forme (QeF) partnerships, Number Local partner s groups Administrativeregions Municipalities Amount invested (millions of Canadian dollars) NA emergence of shared goals from repeated interactions in a local partnership group (interview of April 17, 2014): When we started working together in 2010 we had different goals. Also, we did not know each other. We also had questions about what we were supposed to be doing. Over the years we now have some shared understanding for improving physical activity and eating. I also know the people better. Now when I am talking to people in my own organization about what we need to do to support the work of the local partnership group I feel more confident. We are still learning how to do things. Face-to-face interactions also occur at other levels. At the regional level, collectives of the local partners groups are organized into INTERCALs, whereas at the provincial level there are more than 30 projects being financed in the promotion of healthy food and active lifestyle. Projects at the provincial level are developed by provincial organizations with specifications or issues defined by QEF. QEF also collaborates with research institutions in conducting evaluations of its activities, to develop and share knowledge, to disseminate results, to raise the consciousness of a new clientele and the involvement of new players. 58 The fifth and sixth microsituational set of factors, entry/exit capabilities of partnerships and agreedupon sanctioning capabilities are also related in the Quebec case. The capability to enter or easily leave situations in MSPs ensures that participants recognize that if they do not reciprocate cooperating actions, cooperators may leave. Also, where participants themselves have agreed-upon sanctions regarding behavior in the MSP, they are more likely to abide by them. Consistent with ease of entry into partnerships and autonomy that community groups have in developing their sanctioning rules, there has been expansion of community level partnerships in QEF, as shown by summarized statistics for the period (Table 3). d Whereas there were 35 local partner groups comprising 700 organizations as of 2007, as of 2012 the number had grown to 152 local partner groups made up of 3197 organizations, with a reach extending to almost all communities of Québec: 1067 of the province s over 1100 local municipalities. Further, annual funding in 2011 had reached 16.6 million Canadian dollars, up from 5.7 million in Although the impact of the community-level partnerships on public health outcomes is only currently being assessed, their expansion suggests that community mobilization is on the rise. Also, there has been expansion of partnerships at the provincial level. Two additional MSPs were created between the Quebec government and the Chagnon Foundation in 2009: Avenir d enfants e and Réunir Réussir, f with multisectoral and multilevel governance similar to QEF. The multiplicity of stakeholders in the three partnerships which exemplify partnerships around active living and healthy eating in the province illustrates the complexity faced in engaging individual and collective actors in MSPs as part of a WoS approach. In the next section, we discuss the challenges and propose how to move forward. Discussion: challenges and ways forward In WoS efforts to develop and implement policies around healthy living, the dynamism of the d Data as of March in each given year for , and June each year for 2011 and e See nes-enfants.aspx & gouv.qc.ca/documents/lr/f_4_0022/f4_0022_a.htm f See partnerships/reunir-reussir-student-retention.aspx Ann. N.Y. Acad. Sci (2014) C 2014 New York Academy of Sciences. 223

9 complex governance systems in which decision makers are embedded continues to pose a challenge in achieving convergence in economic and health outcomes, and calls for rethinking policy analysis. In the Quebec case studied, we find that individual and collective actors across multiple sectors at global, national, provincial, regional, and local levels experiment with governance as they develop and deploy public health policies in emergent processes over decades. Given our findings, we argue for departing from the rational, planned processes assumed in prescriptive policy analysis, which has been the dominant paradigm in the social sciences during the 20th century, 8 and is characterized by attempts to find best practices for public health governance. As Ostrom and her colleagues found in nonhealth domains, there was little utility in prescriptive analysis, as it was not possible to specify general reasons as to why governance was effective in the polycentric systems studied. 8,31 Their decades-long research showed that the multiplicity of governance structures in polycentricity led to more effective public service provision and use. Thus, in this paper, it has not been our intention to claim that this case of polycentric governance analyzed is more successful than any other form of governance, nor do we attempt to specify general reasons for success. Rather, given the multiplicity of decision-making units evident in this case, polycentricity best characterizes governance in the WoS approach to public health policy. Insights for understanding polycentric systems come from diagnostic policy analysis, 33 which explicates the multiple sets of actors and various interacting variables shaping effective governance of policymaking in specific contexts. Across contexts that may vary greatly, what is common is that as decision makers develop solutions to challenges of public health, including its governance, other problems emerge, requiring further adaptation. It is vital to understand how adaptations emerge in the multiple units of decision making, whether state, business, or community led. Indeed, the notion of community takes on the meaning of the smallest unit of collective decision making around policies. In Quebec, where a social welfare state had emerged by the end of the 20th century, a WoS approach to public health has been state led. Lessons learned from our analysis of the broader context and microsituations in Quebec can be applied to understanding the functioning of polycentric governance for policy in other contexts, where the private sector or civil society may play a more dominant role. For example, studying the health system of the industrializing country of Bangladesh, Ahmed, Evans, and Standing 3 noted that the pluralistic governance structures were not a planned strategy, but emerged from a dynamic combination of forces ranging from the legacy of traditional care systems, to the enterprise of the private sector and a permissive and weakly regulated public sector. 3 In contexts as diverse as these, interactions between variables may differ, and as problems emerge in the specific setting, there is a need for adaptation. From the broad contextual perspective, in Quebec s WoS efforts, political and economic factors continue to interact at multiple levels to affect governance of public health policy. For example, in 2012 the government of the Quebec Liberal Party that had been in office since 2003 lost elections. The government of the Parti Québécois that took over power led a process for developing a provincial prevention policy, convergent with the 2005 Bangkok Charter that promotes private public partnerships for funding preventive health. 56 The newer health policy was to go through consultation in the fall of 2013 for adoption in However, elections were called again in early 2014, and a new government of the Quebec Liberal Party emerged in April 2014, led by the Minister of Health and Social Services, Philippe Couillard, under whose leadership the MSP, Quebec en Forme had been created. The implications of the 2014 elections on policies are yet to be known, although economic factors have again come to the fore as the new government highlights balancing its budget as a priority, with the possibility of changes to health expenditure. In the face of reduced funding for public health, there may be a need to turn to the private sector as has been the case in the past. Given the successful mobilization of financial resources from the private Chagnon Foundation in partnership with government for deploying policies on healthy lifestyles, attempts may be made to engage other private sector actors in addressing social needs, with further challenges potentially emerging. One such challenge is how best to develop monitoring and sanctioning processes where there is active involvement of private sector actors, such as food companies, which could potentially cheat by 224 Ann. N.Y. Acad. Sci (2014) C 2014 New York Academy of Sciences.

10 not meeting health standards for making bigger profits. Yet, in the above scenario, multiple interactions between broader contextual and microsituational variables from global through local levels may shape the emergent outcomes. At the global level, with increased engagement of the private sector, concerns about conflicts of interests and priorities of state and nonstate actors are emerging, captured in discourse by media, 59 as well as academics, 60 and ultimately reflected in policies and actions adopted in MSPs. For example, perceived conflicts of interest led to the following decision by WHO (2012): [We] may engage with the private sector on occasion, but according to WHO policy, funds may not be sought or accepted from enterprises that have a direct commercial interest in the outcome of the project toward which they wouldbecontributing...forthisreason,the Organization does not accept funding from the food and beverage manufacturers for work on NCD prevention and control. 61 As has been the case for Quebec in the 20th century,itislikelythatsuchglobalfactorsmayinteract with other variables at multiple levels to ultimately influence the engagement of private sector in policy. Diagnostic analysis of interactions across broader contexts and microsituations Diagnostic policy analysis is more suitable for understanding variability and dynamism in polycentric systems, 8 and can inform adaptive learning about the complex workings of governance of policy processes. Interactions between variables across levels also suggest the utility of diagnostic analysis for adaptive learning. For example, from a diagnostic perspective, policy outcomes emerge from the interactions of broad political and economic considerations (e.g., the political events and financing of Quebec s policies to promote healthy lifestyles) and microsituational factors (e.g., decreased feasibility of repeated interactions between MSP participants). In the Quebec case, despite the emergence over time of shared goals to address physical activity and healthy eating, there are conflicts in the priorities of the multiple organizations that constitute local partnership groups, as well as priorities at higher provincial, or global levels. 62 Whereas prescriptive analysis of such an observation is characterized by the provision of recommendations for achieving policy coherence, 63,64 diagnostic analysis develops a deeper understanding of how convergence in the human well-being and economic goals of the various stakeholders especially between community groups and private sector actors emerges through the building of trust between actors at all centers of decision making, as they interact with each other. Our analysis shows that the majority of community partnerships do not engage with commercial actors, whose profit seeking goals may be perceived to be in conflict with those of the communities. From the diagnostic perspective, it becomes clearer that solutions to policymaking challenges emerge from interactions between broader contextual and microsituational factors. For example, political and economic factors related to greater private sector engagement may interact with microsituational factors such as reputation of actors during repeated interactions in community groups to promote or discourage partnerships with business actors. As part of an ongoing study by Addy et al., 12 interviews of local partnership groups in Quebec suggest that consistent with the framework used in this study, 31 an initial ambivalence towards the Chagnon Foundation and Quebec en Forme, as private funding sources, has improved, given positive experiences that some communities have experienced through repeated interactions with these funding sources over years. Specifically, within the broad context of a welfare state in Quebec, cooperation that was initiated by the private foundation and government in repeated situations as part of MSPs around public health has given these actors a reputation of being trustworthy, with others in community groups reciprocating, further increasing the levels of cooperation, and with benefits accruing to MSP members, and subsequently reinforcing levels of trust and reciprocity in the provision and use of public health. These lessons may be useful in engaging other private sector actors in Quebec s MSPs. Similar to how trust and shared goals have been built among stakeholders in the local partnership groups over a long time horizon, participation of individuals from forprofits in the local partnership groups may enable further face-to-face communication and building of trust over time, as well as convergence in their goals. This remains a hypothesis to be tested, bringing us to outlining a research agenda. Ann. N.Y. Acad. Sci (2014) C 2014 New York Academy of Sciences. 225

11 Developing an agenda for diagnostic policy analysis The complexity and dynamic nature of polycentric governance in the WoS approach requires that, moving forward, there is an extensive research agenda to inform adaptive learning in multiple contexts, and assessing linkages between processes and outcomes, for ultimately simultaneously achieving economic and human well-being outcomes. For analysts to develop a better understanding of polycentric systems is a challenge that can be addressed from empirical research, as well as studies employing computational modeling. Our analysis of the Quebec case has provided some initial insights, although it is limited, in not adequately capturing the microsituations out of which the broader policy context emerged, nor in addressing the linkages between the processes and policy outcomes. Addressing such limitations requires research that, for example, draws upon longitudinal data. Archival data on the processes and outcomes of policymaking over the time periods noted in our case-study can yield a deeper, more formalized analysis, for example, with metrics developed from the variables highlighted in this casestudy to measure the effectiveness of governance, and linking with metrics that are being developed in public health, including on interventions. 65 Additionally, the development of metrics facilitates the modeling of such complex processes, using computational simulations, as employed in nonhealth domains. 66 Future research efforts could also shed deeper insights into the interactions between other systems that may interact with health, including education, finance, technology, or food, 67 and how such interactions shaped the broader contexts, microsituations, and policy outcomes that emerge. For example, indicative of the interactions between health and education systems, current policymakers and practitioners in public health were largely trained in education systems promoting positivist perspectives and approaches of the 20th century, with divides between disciplines (e.g., public policy vs. medical vs. business), which did not facilitate repeated interactions between actors in different domains. Shifting from prescriptive to diagnostic policy analysis requires rethinking training of individuals working in public health, enabling repeated interactions to develop across their training and practice. Thus, we propose an agenda that bridges learning and practice in public health. Similar to the decadeslong, extensive empirical research efforts of the 2009 Nobel laureate Elinor Ostrom and other researchers in policymaking around environment and public safety, we propose developing analytical case studies, social science lab and field experiments, and meta-analyses of studies from multiple contexts as part of the training of government, private, and civil society actors currently or prospectively working in public health, as researchers, policymakers, and practitioners (Table 4). This Quebec case exemplifies such empirical exercises, and is a good starting point to extend adaptive learning for addressing partnerships between government, civil society, and private sector actors engaged in WoS in the specific contexts being Table 4. Types of studies for diagnostic policy analysis Type of study Case studies Lab experiments Field experiments Meta-analysis Example Conducting comparative studies of broader contextual factors and microsituational dynamics in public health governance, involving students and practitioners. Studying factors that enable the development of trust and cooperation between students across sectors (e.g., public policy, social work, medical, and business programs). Studying factors that enable the development of trust and cooperation between individuals from multiple sectors working together in MSPs. Developing interdisciplinary research projects for analyzing case studies, and lab and field experiments cited above, and also developing metrics and computational simulation models for assessing linkages between governance and policy outcomes. 226 Ann. N.Y. Acad. Sci (2014) C 2014 New York Academy of Sciences.

12 analyzed. Multilevel research of the processes in Quebec and other regions will inform our understanding of how stakeholders self-organize, building trust among themselves to develop policy in addressing public health challenges, within contexts where they are faced with factors such as unanticipated political cycles, financial challenges, and conflicts of priorities. Such empirically grounded research will contribute to answering questions that this paper does not address about the linkages between governance and successful policy implementation, and how to assess such linkages, as is being done in nonhealth fields. 8 Further, it will complement knowledge that is being developed about health governance at field and organizational levels Conclusion Analyzing the emergence of Quebec s public health policies in its WoS effort has highlighted the interactions between individual and collective actors, their processes, and the manner in which polycentric governance functions in a given context. Policies emerged over time through the actions of leaders from various sectors at the multiple centers of decision making that they inhabited. These policies also provided the contexts within which provincial, regional, and local action plans were developed and implemented by various stakeholders. The leadership of some stakeholders, including government officials, public health experts, philanthropic organization leaders, and community activists has contributed to intersectoral activities to address physical activity and healthy eating, all towards improving public health outcomes. However, engendering trust and convergence in the goals and approaches of the multiple stakeholders, particularly between private sector and community groups remains a challenge. There is the need for continued government leadership in partnership with civil society and private sector for the WoS approach; as well as leadership by research institutions in enabling experimentation, research, and evaluation to further develop polycentric governance. For moving forward we have proposed a diagnostic policy analysis agenda, focusing on understanding how policy emerges from interactions between broader contextual and microsituational variables, to support adaptive learning for polycentric governance. The agenda will allow for a deeper understanding of polycentric governance for developing policies that simultaneously address desired economic and human well-being outcomes as part of a WoS approach in the 21st century. Acknowledgments We gratefully acknowledge support by the Fonds Québécois de la recherche sur la sociétéetlaculture (FQRSC), the Social Sciences and Humanities Research Council (SSHRC), and the Rockefeller Foundation in studies that informed this paper. Conflicts of interest The authors declare no conflicts of interest. References 1. Nelson, J., D. Prescott & S. Held Partnering for success: business perspectives on multistakeholder partnerships. The World Economic Forum, in collaboration with the International Business Leaders Forum, and the Kennedy School of Government, Harvard. 2. Cox, M., G. Arnold & S.V. Tomás A review of design principles for community-based natural resource management. Ecol. Soc. 15: Ahmed,S.,T.Evans,H.Standing,et al Harnessing pluralism for better health in Bangladesh. Lancet. 382: Bentes,M.et al Health care systems in transition: Portugal. WHO Regional Office for Europe on behalf of the European Observatory on Health Systems and Policies, Copenhagen. 5. WHO Whole-of-Society Pandemic Readiness: WHO guidelines for pandemic preparedness and response in the non-health sector. World Health Organization, Geneva. 6. Ostrom, V., C.M. Tiebout & R. Warren The organization of government in metropolitan areas: a theoretical inquiry. Am.Polit.Sci.Rev.55: Ostrom, V Polycentricity (Part 1). In Polycentricity and Local Public Economies: Readings from the Workshop in Political Theory and Policy Analysis. M.D. McGinnis, Ed. Ann Arbor: University of Michigan Press. 8. Ostrom, E Beyond markets and states: polycentric governance of complex economic systems. Am. Econ. Rev. 100: McGinnis, M.D. & E. Ostrom Reflections on Vincent Ostrom, public administration, and polycentricity. Public Admin. Rev. 72: Dubé, L., P. Pingali & P. Webb Paths of convergence for agriculture, health, and wealth. Proc. Natl. Acad. Sci. USA 109: Addy,N.&L.Dubé Philanthropy and a whole of society approach to innovation: a report to the Rockefeller Foundation. McGill Centre for the Convergence of Health and Economics (MCCHE). 12. Addy, N. et al A study for supporting decisionmaking processes in the creation of community portraits. Ann. N.Y. Acad. Sci (2014) C 2014 New York Academy of Sciences. 227