A multi-level perspective on the legitimacy of collaborative water governance in Québec

Abstract

Collaborative governance entails a shift in emphasis from government control toward voluntary arrangements directly involving non-state stakeholders in decision-making. However, the sources of legitimacy for novel institutions such as collaborative water governance remain unclear. Three levels of decision-making that are highly relevant to understanding legitimacy in these contexts are identified and used to assess the legitimacy of collaborative water governance in Québec, Canada. Using Beetham’s dimensions of legitimacy – legality, justification and consent – the sources and deficits of legitimacy are identified through 35 in-person interviews with local stakeholders, watershed organization staff and provincial policy-makers. Findings illustrate the diverse sources of legitimacy, but also a disconnect between what constitutes legitimacy according to local stakeholders and the provincial government. Deficits in legitimacy include lack of implementation, misfit between collaborative governance and existing representative government, and lack of consent from the Québec government. However, although collaborative governance may not be appropriate for all contexts, it addresses important social needs that government cannot and therefore has high potential to complement the roles of existing institutions. Providing several novel insights, this multi-level perspective illustrates how legitimacy can be used to understand the challenges of complex sociological phenomena such as collaborative governance.

La gouvernance collaborative comporte une réorientation du contrôle gouvernemental vers des accords volontaires impliquant directement des acteurs non-étatiques dans la prise de décision. Cependant, les sources de légitimité d’institutions novatrices comme la gouvernance concertée demeurent mal définies. Trois niveaux de prise de décision particulièrement pertinents à la compréhension de la légitimité ont été identifié, et appliqué ceux-ci à une évaluation de la légitimité de la gouvernance collaborative de l’eau au Québec (Canada). Mettant en application les dimensions de légitimité énoncées par Beetham – légalité, justification et consentement – les sources de légitimité furent identifiées par l’entremise de 35 entrevues de individus comprenant des acteurs locaux, le personnel d’organismes de bassin versant et des décideurs provinciaux. Les résultats illustrent les sources diverses de légitimité, mais aussi déconnexion entre ce qui constitue selon légitimité aux acteurs locaux et le gouvernement provincial. Les lacunes de légitimité comprennent un manque de mise en œuvre, une inadéquation entre les institutions représentatives et collaboratives, et l'absence du consentement du gouvernement du Québec. Cependant, bien que la gouvernance collaborative n’est pas approprié à tous les contextes, la collaboration remplit les fonctions sociales que le gouvernement ne peut remplir, et montre donc un fort potentiel pour complémenter les institutions préexistantes. Fournissant plusieurs nouvelles aperçus, cette perspective multi-niveau illustrent comment la légitimité peut être utilisée pour comprendre les enjeux de phénomènes sociologiques complexes tels que la gouvernance collaborative.

Introduction

Water issues are some of the most pressing environmental challenges of the twenty-first century, and have the potential to significantly affect all societies (Intergovernmental Panel on Climate Change [IPCC] 2013). Developing governance solutions that enable societies to address these challenges is paramount, especially since there is increasing consensus that many water issues are rooted in mismanagement and poor governance rather than lack of scientific understanding (de Loë and Kreutzwiser 2007). State-led approaches have often proven inadequate because governments alone do not possess the capacities and resources required to address these problems (Holling and Meffe 1996; Stoker 1998). Rather, multiple actors hold power. The loss of faith in government as a custodian of nature has spurred the recent popularity of environmental governance, which recognizes the inclusion of non-state stakeholders as well as governments in decision-making (Lemos and Agrawal 2006).

Collaborative governance is one such approach that has recently gained favour (Lemos and Agrawal 2006; Ansell and Gash 2008; Prager 2010). Collaborative governance refers to a formal arrangement where a variety of state and non-state stakeholders are involved, voluntarily, in decision-making on public issues (Ansell and Gash 2008; Prager 2010). Legitimacy is important for novel institutions that define relationships differently than traditional models of the state and citizens, but especially for collaborative governance (Baird et al. 2014). When governance does not follow well-known conventions, such as relying on the decision-making authority of elected representatives, its sources of legitimacy may be unclear (Stoker 1998). Without legitimacy, institutions may not be supported, resulting in inadequate strategies for achieving common goals (Berger and Luckmann 1966). Moreover, changing what is legitimate may take longer than creating policies such that even innovative policies may lack legitimacy (Gearey and Jeffrey 2006). But it is especially important to examine the legitimacy of collaborative governance because its reliance on voluntary compliance rather than coercion by government is premised on its legitimacy (Connelly 2011). Rather than using regulations or incentives to ensure compliance, voluntary arrangements rely on the moral obligations of stakeholders (Beetham 1991). If collaborative governance is perceived to be legitimate by stakeholders, the assumption is that they will feel morally obliged to take part in decision-making, plan development and implementation (Wallington and Lawrence 2008).

Of particular concern for the legitimacy of collaborative governance are different perspectives between multiple levels of decision-making. Actors at many levels and interactions between levels influence the success of these arrangements, and collaborative governance in particular (Lemos and Agrawal 2006; Bakker and Cook 2011). But lack of fit or conflicts between perspectives at multiple levels can detract from the legitimacy of governance arrangements (Moss and Newig 2010; Quack 2010). For example, there exist legitimacy trade-offs between citizen participation and effectiveness for multi-level water governance that cannot solely be addressed at the local level (Moss and Newig 2010). In addition, lack of both consent to decisions and accountability for implementation may detract from legitimacy in multi-level governance arrangements (Newig and Koontz 2014). Others have called for improved communication and coordination through mediating structures such as bridging organizations to resolve disconnect between levels (Prager 2010). Despite these concerns, the grounds for the legitimacy of collaborative governance have only rarely been questioned or empirically assessed, and not yet from multiple levels of decision-making (Connelly 2011; Wallington et al. 2008). Thus, it is critical to understand different perspectives of legitimacy at different levels of decision-making related to collaborative governance.

Collaborative water governance in Québec presents a pertinent example for assessing legitimacy at multiple levels. Water governance in Canada is experiencing a rapid shift toward incorporating watershed initiatives, often involving local stakeholder participation and collaboration (Hill et al. 2008). Multi-level governance is inherent to such arrangements within Canada’s federal political system (Bakker and Cook 2011). However, many watershed initiatives lack regulations and bylaws that grant watershed organizations the authority to govern water issues (Furlong and Bakker 2011). Even so, water governance in Québec is a unique and progressive example of voluntary collaborative governance in Canada (Hill et al. 2008). The Québec Water Policy (QWP) articulates a novel approach to governance that mandates the formation of watershed management organizations to engage local stakeholders in decision-making, plan development and implementation as a central component (Québec 2002). Its legitimacy is important because it relies on the collective responsibility of and voluntary implementation by actors at multiple levels to improve water quality and protect the environment.

This research assesses the legitimacy of collaborative water governance in Québec from different perspectives at multiple levels using a sociological interpretation of legitimacy. The following section discusses legitimacy as a sociological concept, identifying Beetham’s (1991) dimensions as an appropriate framework for assessing legitimacy. Three levels of decision-making critical to understanding the legitimacy of collaborative governance are identified, and current knowledge on the legitimacy of collaborative governance is reviewed. The next section provides a brief overview of collaborative water governance in Québec, followed by the research methods. Next, the legitimacy of collaborative water governance in Québec is assessed at each level, and the implications of these results are discussed.

Legitimacy: A sociological concept

Defined as shared strategies - rules, norms and conventions used by society, - institutions provide collective solutions for governing environmental problems (Vatn 2005). However, it is far from guaranteed that an institution such as collaborative governance will be adopted, perpetuated or otherwise supported by society. Compliance to institutions may occur in two ways: through coercion or voluntarily. Individuals may be coerced through mechanisms such as incentives, regulation and the use of force, or they may comply voluntarily with the moral power of legitimate institutions. In this way, legitimacy embodies the moral dimension of compliance, reflecting the moral grounds or reasons for cooperation and obedience (Beetham 1991; Peter 2010). In the absence of legitimacy, institutions may persist with the aid of coercion in the form of sanctions and incentives (Beetham 1991). However, coercion is costly and requires the use of force to maintain. Without legitimacy, the degree or quality of cooperation may erode, diminishing the ability to achieve objectives (Beetham 1991). Thus, legitimacy is essential to institutions such as collaborative governance.

Legitimacy has both normative and sociological interpretations, although the two are closely related (Buchanan and Keohane 2006). Normative interpretations of legitimacy refer to justification of institutions according to external normative criteria (Quack 2010). Thus, normative assessments evaluate whether or not institutions meet certain previously decided criteria, telling us whether institutions adhere to some predefined notion of the good. But whether institutions adhere to predefined norms does not help understand legitimacy in practice because it assumes that certain beliefs and norms are held within society. To understand the legitimacy of institutions in practice, they must be assessed against norms, values and standards that exist within the society in question rather than external criteria of the right or good (Beetham 1991).

In contrast, sociological approaches attempt to understand how legitimacy is defined and constructed in practice (Connelly et al. 2006). This interpretation of legitimacy means that institutions can be justified in terms of shared beliefs and norms held within society (Beetham 1991). Legitimacy thus interpreted as a socially constructed characteristic of institutions means that institutions are morally justifiable, but also indicates whether they are accepted in practice (Simmons 2001; Peter 2010). This entails empirically assessing how and whether institutions are appropriate and accepted within society according to people’s perceptions.

Several approaches have been used to assess legitimacy according to its sociological interpretation. These methods include: a cultural studies framework to examine the sources of legitimacy of sustainable urban governance (Häikiö 2007, 2012); discourse analysis to assess the legitimacy of an environmental governance process (Connelly et al. 2006); and ethical triangulation to explore a range of legitimate perspectives related to forestry and fisheries issues (Fennell et al. 2008). However, none of these methods captures the underlying multidimensional nature of legitimacy. Beetham (1991) developed a multidimensional theoretical framework based on a sociological interpretation of legitimacy that is appropriate for assessing the legitimacy of collaborative governance from a sociological perspective.

Beetham’s framework for legitimacy is based on three dimensions: legality, justifiability and consent (see Table 1). Legality refers to whether behaviour adheres to established rules, norms and conventions (Beetham 1991). These “rules of the game” are secondary to the other dimensions because the rules themselves require justification and consent. Institutions also require justification in terms of shared beliefs (Beetham 1991). Justification is divided according to source norms and content norms. Source norms refer to the rightful source of authoritative decisions, which can include science, tradition and the interests of people in society (Beetham 1991). Content norms refer to the legitimate way of making decisions and carrying them out, such as the appropriate means of dividing decision-makers from followers as well as their appropriate roles and responsibilities (Beetham 1991). Beetham divides content norms according to two principles. The principle of differentiation refers to justification for who decides and according to what rationale they are chosen. The differentiation between decision-makers and followers rests upon justification that those who make decisions possess qualities that make them more qualified than those who follow (Beetham 1991). More generally, this is interpreted to include larger questions of procedure alluded to by content norms. The principle of common interest refers to the social necessity of institutional arrangements (Beetham 1991). Decisions must meet the substantive goals of society and achieve normatively justifiable and desirable ends (Parkinson 2003). Consent is “the act of deliberately and suitably communicating one’s intentions to undertake obligations toward another” (Simmons 2001, 166). Without consent, justification alone cannot secure moral obligation. Consent can be granted through explicit means such as voting, expressed implicitly by support through party membership, or through continued participation (Beetham 1991). Express consent is an action that is direct and explicit so that it expresses acceptance or agreement (Simmons 2001). For example, swearing an oath of allegiance, public acclamation or expressions of approval are explicit and public acts that demonstrate consent. In contrast, taking part in consultations, negotiations or otherwise participating signify tacit consent (Beetham 1991). Conversely, lack of participation, boycotting or withdrawal from participation indicates lack of consent.

Table 1. Beetham’s dimensions of legitimacy, definitions and examples (Beetham 1991).

Dimension of legitimacy	Definition	Examples
Legality	Process follows established rules and conventions	Adherence to laws, established procedures, stated goals or directives
Justification
Source norms	Rightful sources or bases for decisions	Science, interests of people, natural law, religion
Content norms (procedure)	Content and process of decisions	Who decides and how, what decisions are made, roles and responsibilities
Content norms (common interest)	Substantive and normative goals of society	Justifiable and desirable ends
Consent	Act of communicating one’s commitment confers moral obligation	Participation, voting, public approval, recognition or accolade, providing support

Three levels of analysis

A critical question for assessing the legitimacy of novel governing arrangements is: According to whom is it legitimate? Beetham’s interpretation of legitimacy is based on assumptions of dominant-subordinate power relationships (1991). However, contemporary forms of governance often do not rely as clearly on dominant-subordinate power relationships (Stoker 1998). In other words, who the dominant and subordinate parties are in contemporary governance arrangements may not be obvious, especially for voluntary collaboration. Then according to whose perspectives should an arrangement be justified? And who can or should grant consent, and how should this be done? Since assessing all perspectives presents an overwhelming task, one alternative is to identify those perspectives most pertinent to understanding and improving legitimacy. Identifying the sources and deficits of legitimacy from those perspectives provides a way to pursue enhanced legitimacy (Parkinson 2003). The question becomes, which perspectives are most important to understand? While it is important to consider differences between stakeholder groups or contexts, every case is unique, with diverse biophysical and social characteristics (Milot and Lepage 2010). Thus, results from cross-case analyses may have little relevance and generalizability beyond specific cases. For collaborative governance, which relies on local stakeholders but is often initiated at the policy level, an alternative and perhaps more useful approach is to focus on differences between levels of decision-making.

To assess the legitimacy of collaborative water governance in Québec from multiple perspectives at different levels, three levels of decision-making were identified for which individuals' perspectives are highly relevant: local stakeholders, organization staff and policy-level decision-makers. First, the perspectives of local stakeholders involved in collaborative governance are essential to understanding its legitimacy. The term “stakeholder” refers to actors and interest groups that can affect or are affected by the issues of concern (Reed et al. 2009). In voluntary arrangements, if local stakeholders do not view the process as legitimate, implementation will likely not be effective because stakeholders will not feel morally obliged to take part in decision-making, plan development or implementation (Papadopoulos 2003). Lack of implementation, support and buy-in may undermine environmental policies and initiatives (Wallington et al. 2008). Second, organization staff have a large influence on how and whether their organization is perceived as legitimate or not. They facilitate and coordinate collaborative processes and may have a large influence on the ideology and ethos of their organization (Connelly 2011). They may also interpret government policy and mandates, translating goals into collective narratives and actions (Connelly 2011). Furthermore, organization staff define processes to engage stakeholders at the watershed level and shape involvement over time (Booher 2004). Ultimately, the intermediary role organization staff play in shaping the legitimacy of collaborative governance within and between institutional levels has yet to be fully explored (Prager 2010). Third, perspectives at the policy level such as those of government decision-makers are important, but have remained relatively unexplored when assessing the legitimacy of water governance (Wallington et al. 2008). But despite recent changes in modes of governing, governments remain important actors in environmental and natural resource governance arrangements, and need to be considered when analyzing governance. Governments often initiate and oversee novel arrangements for water governance defining how they operate, setting agendas and steering outcomes, as well as allocating resources and authority (Newig and Koontz 2014). States also fill the critical role of defining the rules for collaborative initiatives, shaping policies, laws, regulations, jurisdictions and capacities (Bell and Park 2006). Given this reality, local and regional collaborators are often limited in their ability to directly make decisions (Ribot 2002). Hence, the perspectives of policy-level decision-makers are important to understand collaborative governance and its legitimacy.

Assessments of legitimacy according to Beetham's dimensions at multiple levels

While much research on collaborative governance refers to legitimacy, only two studies have used Beetham’s dimensions (Wallington et al. 2008; Connelly 2011). Both assessed novel institutions for governing environmental problems, but no studies have yet assessed legitimacy at multiple levels; one evaluated perspectives at the local stakeholder level, while the other did so at the organization level (Wallington et al. 2008; Connelly 2011). This section reviews findings of legitimacy for collaborative governance in the water sector, and in the environmental field more generally, drawing out findings relevant to multiple levels, organized according to Beetham’s dimensions: legality, justification and consent (see Table 1).

Legality has been challenging to assess and has yet to be explored at multiple levels. Since institutions take a long time to become established, and individuals must learn the accepted operating procedure, the legality of collaborative governance has been difficult to assess (Leach et al. 2002). Difficulty assessing legality may be further exacerbated because rules and procedures may be developed and changed over time. One study found that organization representatives sought to maintain legitimacy by developing rules and procedures that were justifiable (Connelly 2011). Referring to how these organizations approached legitimacy: “it was not taken for granted and strenuous efforts were made to establish and maintain it” (Connelly 2011, 936). These rules and procedures were changed in pursuit of increased legitimacy and in response to perceived legitimacy deficits. This meant that whether rules and procedures were followed was often unclear, but also that legality and justifiability were difficult to distinguish in a rapidly evolving institutional context.

Collaborative governance may face problems of legality because it relies on voluntary compliance with potentially unclear roles and responsibilities (Stoker 1998). Unclear or evolving rules and procedures make it difficult for actors to follow comprehensible patterns of behaviour (Booher 2004). One result is that participants may adopt roles and responsibilities without clear justification for how they represent the ideas or people they claim to (Hendriks 2009). For example, Roberge and colleagues (2011) examined the perceptions of local - both civil society and industry – stakeholders concerning roles in collaborative governance through forest stewardship programs in Québec. They concluded that stakeholder roles are often unclear, resulting in unrealistic intentions, unfulfilled expectations, dissatisfaction and, ultimately, deficits in legitimacy. This is not restricted to a single level. Another study observed that managers and committees deliberately chose not to follow rules and agreed-upon management regulations (Fennell et al. 2008).

Assessments of the legitimacy of collaborative governance most often focus on justification. Diverse norms are used to justify collaborative governance, and often a combination of norms is most legitimate. But while researchers have identified the existence of multiple legitimacy discourses (Häikiö 2007; Wallington et al. 2008), some have also noted the absence of a single overriding discourse (Connelly et al. 2006). Moreover, whether certain rationales constitute sources or deficits to legitimacy often depends on perspective. For example, active participants may perceive whether norms of inclusion are justified quite differently than those who do not participate at all (Wallington et al. 2008). Appropriate roles can also depend on perspective. For example, civil society and forestry industry representatives in Québec disagreed on appropriate roles related to forest stewardship (Roberg et al. 2011). Hence, diverse perspectives and justifications give legitimacy a hybrid nature but mean that judgement of legitimacy against any single norm may be weak.

Collaborative governance may lack legitimacy simply by not fulfilling norms and expectations – old or new. Collaboration is costly and difficult to achieve in practice, but it also suffers from lack of democratic accountability and limits related to its voluntary nature (Huitema and Meijerink 2009). Often, the most important factor that detracts from legitimacy is that arrangements do not meet the common interest (Wallington et al. 2008; Baird et al. 2014). A major concern is that collaborative governance will not achieve substantive results, including improvements to environmental quality (Wallington et al. 2008). Such concerns are compounded because environmental outcomes of collaborative governance are often not adequately evaluated (Ansell and Gash 2008; Hogl et al. 2012). These observations at the local and organization level are echoed at the policy level. From a policy perspective, collaborative governance may lack legitimacy compared to other policy tools since voluntary arrangements provide no guarantee plans will be implemented (Furlong and Bakker 2011).

Perspectives of legitimacy can also incorporate elements from multiple, at times even contradictory, sources, embedding a mix of often-competing rationales (Connelly et al. 2006; Wallington et al. 2008; Connelly 2011). For example, while effectiveness represents a primary justification and potential deficit for collaborative governance, research at the local stakeholder level has demonstrated that justifiable process and outcomes are both required for legitimacy; collaborative arrangements must balance procedural norms against substantive results while addressing both (Connelly 2011). This balance between procedural and substantive components remains largely unexplored in the environment and water fields (Connelly 2011; Baird et al. 2014). Thus, in addition to conflicting perspectives, individual perspectives may also be problematic, reflecting tensions or trade-offs. These challenges may be exacerbated because established norms are often fiercely held on to. Several studies have observed that established norms such as those of representative democracy and accountability dominate even for novel governance institutions (Connelly et al. 2006; Häikiö 2007).

A related consideration is the fit between collaborative and established governance. Different perspectives may be rooted in established or novel ways of governing (Wallington et al. 2008; Connelly 2011). Collaborative watershed governance often coexists with, and is initiated by, traditional governments (Baird et al. 2014). But others have observed that while collaborative norms exist in parallel with those of representative democracy, they may also conflict (Newman et al. 2004). When governing arrangements are set up, tensions may emerge between established models of elected representation and newer non-electoral models. In one instance, elected officials were openly hostile towards participatory processes (Connelly 2011). Hence, some have argued that a clearer role of collaborative institutions within larger democratic contexts is required because their sources of legitimacy are vastly different than democratic norms (O’Neill 2001). For these reasons, understanding collaborative watershed governance as situated within and in relation to its larger institutional context of representative democracy means that a central question is whether and how new and established institutions fit best together.

A few have noted different perspectives on the legitimacy of collaborative governance at different levels. For example, one study observed that while accountability to state or regional governments was not a problem, lack of accountability to the community and representation of local communities were deficits to legitimacy (Wallington et al. 2008). A different study emphasized that state and non-state actors had different perceptions about whether collaborative governance processes and outcomes were legitimate, with the state being more likely to view it as legitimate (Baird et al. 2014). Still others observed that organization staff negotiate and shape organization procedures and goals based on their and others’ perceptions of legitimacy (Connelly 2011). While such factors may be important to connecting and scaling up from local to regional levels, ways of fostering collaboration between different levels have yet to be fully explored (Prager 2010).

Consent is perhaps the least empirically visible dimension of legitimacy (Connelly 2011). Collaborative governance lacks the traditional means of consent granted to democratically elected governments through voting, leaving its sources of consent unclear (Stoker 1998; Hogl et al. 2012). While some argue that consent must be actively granted for collaborative governance to be legitimate (Bernstein and Cashore 2007), it is unclear the extent to which express consent is necessary in principle or present in practice (Connelly 2011). And express consent may be less important than the absence of dissent. Stakeholder participation can be assumed to reflect tacit consent; the assumption is that continued participation means that no stakeholder strongly opposes decisions or implementation of plans (Wallington et al. 2008). Whether consent is lost or gained as a result of collaborative governance remains an unanswered question.

Individuals at different levels may grant consent in different ways. At the local stakeholder level, one study showed that collaborative processes often lacked consent: positions went uncontested or remain unfilled, and there was low voter turnout (Connelly 2011). Another study identified other deficits, including lack of local stakeholder support for decisions, lack of implementation, and non-cooperation (Wallington et al. 2008). However, local stakeholders expressed support in other ways, such as participating and providing funding (Connelly 2011). At the policy level, research on metagovernance has emphasized the unique role and mandate of states and other governments, including distinct means of granting consent (Bell and Park 2006). While scant attention has been paid to consent at the policy level, some have observed that states are often reluctant to grant collaborative bodies power (Furlong and Bakker 2011). While consent has not been an explicit topic of research at the organizational level, organization staff typically support collaborative governance. In one instance, paid organization staff played a major role in shaping community representation, contacting, convincing, encouraging and enabling individuals to get involved who might otherwise not be (Connelly 2011). In these ways, consent is granted and withdrawn in diverse ways, varying among individuals in different roles at different levels.

Collaborative water governance in Québec

Adopted in 2002, the Québec Water Policy (QWP) articulated a collaborative approach to governing water. The policy proposed regulatory mechanisms in conjunction with integrated watershed management, emphasizing voluntary implementation (Québec 2002). Specifically, it mandated the formation of 33 watershed management organizations as key to integrated water management based on a collaborative approach (Québec 2002). In 2009, the number of watershed organizations was increased to 40, and their territories were expanded to cover all of southern Québec (Ministère du Développement durable, de l’Environnement, de la Faune et des Parcs [MDDEFP] 2011).

Watershed organizations are the policy’s main mechanism for water governance and to improve environmental conditions (Québec 2009). Watershed organizations are formal, decentralized, participatory organizations, maintained and supervised by the government, and have no separate legal authority (Baril et al. 2006). However, they have clear decision-making roles: they are to act as planning and consultation tables, enabling stakeholders to develop and implement water master plans, while government representatives provide scientific and technical support (Baril et al. 2006). Watershed organizations are intended to equally represent elected officials, economic interests and civil society, including environmental and other social organizations (MDDEFP 2012). The process also incorporates multiple public hearings to encourage participation by and input from all members of the community (Milot and Lepage 2010). Furthermore, representatives from various provincial ministries are designated to act as resource persons to provide support and expertise. Thus, the QWP is designed to involve both governmental and non-governmental stakeholders at multiple levels – from local to provincial – in decision-making.

A clear example of collaborative governance, integrated watershed management in Québec is designed to be non-regulatory and voluntary (Baril et al. 2006). Watershed organizations are directly responsible for implementing integrated management by preparing a water master plan. Plans are to be updated every five years, but are not legally binding. In 2009, Québec adopted the Act to affirm the collective nature of water resources and provide for increased water resource protection, which specifies that government bodies at different levels must “take into consideration” the water master plans, although they have no legal requirement to implement them (Québec 2009). Although watershed organizations possess no regulatory authority, the government stakeholders involved do possess the authority and obligations of their mandates (Baril et al. 2006). Therefore, watershed organizations must rely on public consultation, stakeholder participation, local and regional expertise, and the mandated responsibilities of municipalities, regional county municipalities (RCMs), and the provincial government (Baril et al. 2006). Stakeholders, including governments, private companies, non-governmental organizations and citizens, have the “voluntary opportunity” to implement the plan (Baril et al. 2006). Ultimately, watershed organizations rely on voluntary collaboration and implementation by these stakeholders.

Collaborative water governance has evolved rapidly in Québec, and has been increasingly studied (Milot and Lepage 2010). Watershed organizations deal with a diverse range of issues and interests within their territories, and organization staff employ diverse approaches to engaging local stakeholders and carrying out their mandates (Milot and Lepage 2010). The diversity and flexibility of watershed organizations is viewed as one of their strengths because they can adapt to address issues in their local context. Even so, legitimacy at multiple levels remains a critical issue for collaborative water governance in Québec. Some have questioned the effectiveness of watershed organizations, argued that they may lack legitimacy, and alluded to their ongoing need for legitimation, referring to them as “an organization that initially has to survive” (Milot and Lepage 2010, 152). Despite this need to understand water governance and its legitimacy, research in Québec has focused on specific environmental issues such as water quality in Lake Champlain rather than on institutions (for example see Smeltzer et al. 2012). The legitimacy of watershed organizations was studied from an organizational perspective in their initial stages of becoming established (Beaulieu 2008). Findings showed that multiple rationales confer legitimacy on watershed organizations: the existence of proven examples implemented elsewhere, logic based on an understanding of water issues, and participatory norms of inclusion. But an evaluation of the QWP 5 years after its adoption articulated the difficulty of assessing its effectiveness because institutional structures take time to develop and actions must first be implemented; longer term evaluations are needed (MDDEFP 2011). Collaborative water governance in Québec has existed for over 10 years, and watershed organizations completed their water master plans in 2014 (Regroupement des Organismes de Bassins Versants du Québec [ROBVQ] 2014). But these plans have yet to be implemented to a significant degree (Vérificateur général du Québec 2013). For these reasons, the Québec model provides a timely and highly relevant example to assess the legitimacy of collaborative water governance from a multi-level perspective.

Methods

To understand the legitimacy of collaborative water governance in Québec at different levels of decision-making (local stakeholder, watershed organization staff, and policy levels), multiple perspectives were assessed at each level. Several sources of qualitative evidence were used, including document review, participant observation and 35 in-person semi-structured interviews (Yin 2003). A total of 39 individuals were interviewed, as several interviews were conducted with two or more interviewees in the same session. Since watershed organizations in Québec are diverse, ranging in their size of territory, dominant water issues and durations of existence, perspectives on legitimacy were also likely to be diverse (Leach 2006; Milot and Lepage 2010; Schmidt and Morrison 2012). For interviewee sampling purposes, three out of the 40 watershed organizations were selected to include a range of issues, geographic sizes and durations of existence, within practical limitations (Table 2). Agence de bassin versant des 7 (ABV des 7) was selected to explore the challenges of large geographic size and many administrative jurisdictions, but low population density. In contrast, Conseil des bassins versants des Mille-Îles (COBAMIL) was selected to capture challenges associated with a smaller territory but a much higher population density, such as urban pollution issues. Finally, Société de conservation et d'aménagement du bassin de la rivière Chateauguay (SCABRIC) was selected primarily because it has a long history and relationships within its territory compared to the other two watershed organizations.

Table 2. Watershed organizations and their characteristics.^a

	Watershed organization
	ABV des 7^b	SCABRIC^c	COBAMIL^d
Area (km^2)	40254	2543	1053
Population	320 000	297 000	435 401
Population density (/km^2)	8	117	414
# Municipalities	50	39	25
# RCMs^b	9	5	4
Age	2004	2002 (1993)^c	2010

^a Sources: Brodeur et al. (2012); COBAMIL (2013); (ABV des 7 (2014); Audet et al. (2014); COBAVER-VS (2014); COGESAF (2014).

^b Agence Bassin Verssant des 7.

^c Société de Conservation et d'aménagement du Bassin de la Riviere Chateauguay

^d Conseil des bassins versants des Mille-Îles

^eA regional county municipality (RCM) or municipalité régionale de comté is a regional administrative division in Québec.

^fWatershed organization existed previous to Québec Water Policy.

Interviews were conducted with a total of six staff members from among the three watershed organizations (Table 3). Interviews were also conducted with 8–10 local stakeholders in each of the three watersheds. These individuals were identified in consultation with organization staff and recruited by the researchers based on their participation as stakeholders in watershed organizations. Where possible, a balance of interviewees from all sectors identified in the QWP (municipal, economic and social and environmental) was included (Québec 2002; MDDEFP 2012). Finally, six decisions-makers with extensive experience at the Québec policy level were also identified through snowball sampling methods and interviewed. These individuals, all of whom influence decisions at the level of the Québec government, included an elected representative from the Government of Québec, current and former public administrators from the Ministry of the Environment (MDDELCC) and the Ministry of Municipal Affairs and Land Use (MAMOT), and representatives from ROBVQ, the organization representing watershed organizations at the provincial level. Snowball sampling methods were used to identify additional local stakeholders as well as provincial level interviewees. Snowball sampling involves identifying subsequent individuals from previous interviews, beginning by identifying initial interviewees using selection criteria such as participation in watershed organizations (Reed et al. 2009). The interviews continue until the data becomes saturated, meaning that no new information is revealed from subsequent interviews, no new interviewees are identified, or practical constraints such as time or resources are reached (Small 2009). Snowball sampling increases the number of respondents since people are more likely to respond to a researcher when the researcher has been in contact with someone they are familiar. One potential weakness of this method is that sampling may be biased by the social networks of the first person in the snowball sample (Small 2009). Although saturation was reached, this method may still omit marginalized and less-visible groups. However, rather than exhaustive sampling, our aim was to identify dominant narratives and perspectives to provide a multi-level analysis. Moreover, omission of marginalized groups was minimized since initial contact was made with the coordinators of watershed organizations; those directly involved in the process are expected to have contact with all relevant parties. One important exception was that no First Nations representatives were interviewed, both because First Nations issues often involve the national government and thus were beyond the scope of this study, and also because none of these groups were directly engaged with the selected watershed organizations. Thus, although the implications First Nations’ lack of participation have for the legitimacy of watershed organizations are discussed, this research does not specifically assess whether collaborative governance is suited to effectively engaging First Nations, or whether alternatives are needed.

Table 3. Numbers of interviewees according to role and watershed organization.

Level of decision-making	Interviewee role	Watershed organization
		ABV des 7^a	SCABRIC^b	COBAMIL^c	Québec	Total
Local stakeholder	Economic	1	0	3	0	4
	Social	3	2	1	0	6
	Local government	5	6	2	0	13
	Regional government	1	1	2	0	4
Organization staff		1	2	3	0	6
Québec policy		0	0	0	6	6
Total		11	11	11	6	39

^a Agence Bassin Verssant des 7

^b Société de Conservation et d'aménagement du Bassin de la Riviere Chateauguay

^c Conseil des bassins versants des Mille-Îles

Interviews explored themes of collaborative governance identified in the literature as relevant to legitimacy according to Beetham’s dimensions. Themes included organization history, institutional structure, mandate, roles and responsibilities, participation, fit and appropriate function (Newman et al. 2004; Leach 2006; Ansell and Gash 2008; Wallington et al. 2008; Connelly 2011). Interviews were tape-recorded and transcribed, and responses were coded for themes previously identified as well as emergent themes. Most interviews were conducted in English, although several were conducted in French with the assistance of a translator. Interviews were transcribed in the language they were conducted in and then translated if necessary. Quotations from interviews were translated directly from French to English. No significant issues were encountered when moving between languages. Responses were identified and analyzed using directed content analysis, the preferred method when sufficient theoretical frameworks are available (Hsieh and Shannon 2005). Dominant narratives and perspectives that emerged were verified against those of other interviewees, as well as through document review to triangulate results. Results were further verified by key informants, including organization staff. Discursive construction was used to understand and present a coherent picture of the dominant sources and deficits of legitimacy of collaborative water governance in Québec at different levels, according to Beetham’s dimensions (Holstein and Gubrium 2011).

Results

Results are presented in the form of dominant narratives that illustrate the legitimacy of collaborative water governance in Québec. Although watersheds were not homogenous, others have emphasized their diversity (Milot and Lepage 2010). Thus, while significant differences between watersheds and stakeholder groups are noted in discussing local stakeholder level perspectives, the main focus is to understand and contrast perspectives at different levels rather than to provide a cross-case comparison. Focusing on differences in perspectives between the three levels, important issues are organized and discussed according to Beetham’s dimensions of legitimacy: legality, justification and consent.

Legality

Legality refers to whether practices conform to established rules, norms and conventions (Beetham 1991). In Québec, determining whether actors such as organization staff conformed to established rules and procedures was challenging. At its inception in 2002, collaborative water governance in Québec lacked clearly defined rules that worked well in practice. For example, interviewees who had helped officially create SCABRIC, the oldest watershed organization in this study, described its original mandate as too broad and unclear. The process was trial and error, such that guidelines and deadlines from the government were changed several times; rules presented moving targets. This result confirms those of other studies: the existence or stability of rules and practices cannot be assumed in contemporary forms of governance, which are often evolving and less formal than traditional governments (Connelly 2011).

However, the process has evolved and matured rapidly since the QWP was adopted and watershed organizations were set up. Procedures and expectations were clarified and better adapted to individual watersheds. Watershed organizations worked closely with MDDELCC to improve the process with the assistance of ROBVQ, the provincial organization that represents them. For example, although the policy originally articulated that local stakeholders from economic, governmental and social sectors were to define the content of water master plans, technical committees were created when the need for them was identified. Moreover, guidelines, training opportunities and tools were developed to enable watershed organization staff to better achieve their mandates (ROBVQ 2014). As a result of this coevolution, adherence to the established rules and procedures set out in the QWP has been unproblematic.

However, this multi-level perspective provided new insight not captured in studies that limit their scope to the local level. Interviewees at all levels expressed concerns that local actors did not adhere to laws and regulations that, while not directly related to collaborative governance, still strongly influenced its effectiveness. In many instances, local and regional governments did not possess the resources to enforce local bylaws or to monitor water quality, which impeded watershed organizations’ missions. This result was seen in all three watersheds. In ABV des 7, the largest watershed, half of all interviewees expressed concerns that local governments did not adequately enforce regulations such as riparian property bylaws, agricultural regulations and measures to prevent the spread of invasive species. A mayor in the SCABRIC watershed described how municipalities did not have adequate funding to monitor and enforce regulations on septic tanks. In COBAMIL, interviewees representing the agricultural sector emphasized strong enforcement of agricultural regulations, even while several other interviewees noted a widespread inability of local governments to meet requirements set by the province. In all three watersheds, the inability to enforce regulations was a barrier for watershed organizations to achieve their mandates. This inability detracted from the legitimacy of the collaborative process even though these laws and regulations were not the jurisdiction of watershed organizations. Likewise, failure of provincial ministries to enforce regulations and implement goals stated in the QWP also detracted from the legitimacy of the process. Without provincial mechanisms such as the user- and polluter-pays principles proposed in the QWP, action on water issues remains of little concern for many stakeholders who have a large impact on water.

Justification

Justifications that ground legitimacy may be according to shared beliefs based on a commonly justified understanding (source norms) or the way things are done in practice (content norms) (Beetham 1991). First source norms, then content norms are discussed.

Source norms

Justification according to the sources of decisions for collaborative water governance was critical to legitimacy at all three levels. The QWP makes three claims regarding the legitimate sources of decisions: they should reflect the interests of citizens, they should be locally informed and they should be based on science (Québec 2002). While the policy affirms that decisions should be based on scientific knowledge, it also appeals to norms of both participatory inclusion and local decision-making, referring to the responsibilities of “all members of society” and reliance on “local and regional expertise” (Québec 2002, 19). However, these positions were neither universally accepted nor uncontroversial. Tellingly, over half of all interviewees mentioned the sources of decisions as a concern, expressing a diversity of positions.

These tensions were most apparent among local stakeholders at the watershed level, although different concerns dominated in each watershed. In ABV des 7, a common theme was that decisions should reflect the common interests of citizens, rather than only select groups or individuals. For example, a local business owner voiced this perspective: “The purpose of an area should be a community thing. There is no industry or single landowner that can decide for everybody; it requires a collective effort.” Echoing this, the leader of an environmental group asserted that ABV des 7’s decisions should be informed by grassroots issues and driven by citizens’ concerns. Although these individuals approved that watershed organizations addressed interests of citizens, in all three watersheds even more local stakeholders and organization staff expressed that it was problematic to base decisions solely on public interests. For example, a scientist on COBAMIL’s technical committee was concerned that the public was inadequately informed so that their decisions did not reflect the importance and urgency of water issues: “We need to cope with the perception of the population. It's an issue of communication. It's an issue of priority.” This was an issue in all three watersheds studied: priorities from public consultations reflected short time scales and omitted important water issues such as flooding, which later had to be incorporated into plans. Thus, decisions that were not adequately informed by scientific understanding or expertise lacked legitimacy. Similarly, in SCABRIC several interviewees were concerned that the organization was not neutral or unbiased because they felt that decisions and plans made by watershed organizations needed better scientific grounding. For example, an urban planner representing their RCM provided watershed organization staff with data from their city, connecting them to sources of information. Each of these individuals participated in meetings to ensure that decisions were informed by their scientific expertise or local experience; they were intimately involved with water issues and felt it important to share their knowledge.

But, good information was not solely scientific information. Rather, local experience and knowledge were important to account for both the physical and social characteristics of watersheds. This view was dominant among interviewees who participated on technical committees, especially in COBAMIL and SCABRIC. In SCABRIC, a representative of the agricultural producer’s union (UPA) participated to communicate farmers’ perspectives: “There are limits to what we can do and I think that being involved in that kind of project is a way to better communicate the reality of farmers.” Similarly, a ministry representative became involved in COBAMIL because they wanted to communicate public health concerns. In these ways, diverse source norms combined to ground legitimacy for local stakeholders, albeit in different ways depending on the watershed and context.

Organization staff ensured that all positions were taken into account, balancing them such that a combination of these sources was the most legitimate. One organization staff member described that, in the event that these norms had conflicted, “It would have been difficult to resolve which one is more important. It is really hard to say.” A staff member from a different watershed organization noted that the basis for decisions of initial public consultations were “totally irrational,” and that it was important to bring together a technical committee to complement public perspectives. Staff in all three watershed organizations attempted to ground decisions in science, but possessed limited resources to do so. Watershed plans require information about many complex issues. Despite preparing extensive, multi-stage plans that identify issues, prioritize them and propose actions, watershed organizations face limitations regarding the amount of information they can integrate.

At the Québec level, the policy both articulates a shift from the government issuing directives toward more locally informed decisions, and emphasizes the importance of science (Québec 2002). Rather than top-down decision-making, decisions were intended to incorporate local knowledge, practical experience and scientific expertise, enabling the province to better complement and support watershed organizations. Accordingly, provincial ministries would actively participate in and engage with watershed organizations. However, uptake of these norms has not occurred within government administration where traditional norms of government decision-making persist. Several interviewees at the Québec level emphasized the importance of science-based decision-making, but also the benefits of locally informed decisions. In addition to scientific evidence, politicians and public administrators also expressed that decisions needed to consider socio-political information. One public administrator who had been intimately involved in developing and implementing the QWP described the intended approach - that of government listening to and learning from experts in watersheds: “We have a lot to learn from people in the watershed. They know some things that we cannot necessarily understand by looking at the data.” But this mentality remained an ideal rather than a reality within MDDELCC. While a majority of those interviewed at the Québec level maintained that decisions should be informed by local understanding and expertise, these interviewees emphasized that uptake of these norms had not occurred throughout government administration.

A disconnect exists between source norms relied on by decision-makers at the local watershed and Québec levels. While processes are justified by local stakeholders using collaborative and participatory rationales, the idea that elected officials make decisions on behalf of citizens remains dominant within the provincial government. Although some policy-makers advocate for fundamental changes to governing, advocating locally informed decisions and increased engagement between the provincial and watershed levels, norms at the Québec level continue to hinder collaborative governance and realization of the ideals of the QWP.

Content norms

Justifications that ground legitimacy may also be according to the content of decisions - the way things are done and whether common interests are addressed (Beetham 1991). These issues are discussed, organized according to four topics: inclusion, voluntary collaboration, effectiveness and complementary functions.

Inclusion

The QWP and subsequent provincial guidelines emphasize inclusion as important: equal representation from municipal, economic, and social and environmental sectors on watershed organizations’ administrative councils, as well as the inclusion of First Nations, is required (Québec 2002; MDDEFP 2012). Responses related to inclusion varied widely, many indicating that it either significantly contributed to or detracted from the legitimacy of the collaborative process. These differences were most apparent when comparing between levels.

Most local stakeholders interviewed did not view inclusion as problematic: it was either not an issue or they approved of watershed organizations based on organizations’ perceived inclusion of all members of the community. However, while those individuals had not encountered insurmountable barriers to participation, they did describe several such barriers that had prevented others from participating. For example, in ABV des 7 and SCABRIC, the sizes of the territories made participation a challenge. Often, local stakeholders did not have the time or money to participate when meetings were far from their homes or workplaces. Thus, from the perspectives of local stakeholders interviewed, inclusion remains an ongoing challenge, although not a critical legitimacy deficit.

In contrast, the majority of watershed organization staff and policy-level interviewees thought that inclusion was important and required attention. For one staff member, inclusion was not a problem: “Everybody’s invited. It’s easy to be at the table.” A provincial member of parliament supported this ideal, but recognized its limitations: “Of course not everyone can be involved, but everyone who wants to be involved should be able to be.” But from the perspectives of organization staff and policy-makers, inclusion was problematic in practice. Some actors or groups possessed influence, capacities and resources that enabled them to participate in more meaningful ways than others. In SCABRIC, where agriculture was the dominant land use, farmers were thought to be some of the most important stakeholders to include; water issues were affected most by agriculture. In the other two watersheds, municipal water use was a predominant concern, such that cities were perceived to possess the majority of responsibilities and capacities to address water issues. Organization staff in all three watersheds viewed local governments as some of the most important stakeholders to include since they often possessed the resources to act and the democratic mandates to make decisions. In COBAMIL, the most urban watershed, one staff member expressed that municipalities hold so much responsibility for managing water that they absolutely must be included. This reality caused tension with norms of equality because watershed organizations were not supposed to favour any group over others.

Another concern of both watershed organization staff and Québec policy-makers was whether the collaborative process was appropriate for all contexts and issues. In some contexts, such as for addressing diffuse problems, the collaborative approach has been highly effective. For example, it has been successful at addressing non-point source agricultural pollution. The director of one watershed organization explained that if the causes of water pollution are diffuse, individual contributions are easy to ignore. The watershed organization overcame this tendency by fostering collective responsibility: “If we sit together we can’t say I’m not responsible.” Bringing everyone to the table meant that those responsible would be present; no one could be used as a scapegoat. In this way, the inclusiveness of collaborative governance may be well suited to addressing diffuse problems. But in other contexts, inclusion remains a significant barrier. For example, including First Nations as simply a regular stakeholder was not viewed as appropriate because the approach may not fit the norms established for interacting with this group. First Nations often work nation-to-nation, addressing issues with the Canadian government as a result of complex history and political issues.

Voluntary collaboration

The voluntary nature of collaborative governance in Québec was highly controversial, in some instances detracting from legitimacy, while in others bolstering legitimacy. Interviewees at all three levels were divided on whether the process was appropriate or not and whether plans should be voluntary. The voluntary collaborative approach outlined by the QWP was intended to be legitimate to local stakeholders. A policy-maker who helped write the policy explained that the process was made voluntary so that watershed organizations would be neutral, increasing their capacity to solve conflicts between stakeholders. The idea was “to give the watershed organizations the moral authority to coordinate actions on a watershed.” Watershed organizations were designed to have moral rather than legal power, and therefore rely on legitimacy, the moral dimension of power. The interviewee continued, “They have no legal power, but they have the responsibility to bring all stakeholders together to write the master plan, and after that to have stakeholders implement the master plan.”

To a certain degree, this vision has been effective. Some local stakeholders in each watershed expressed that this voluntary approach was more legitimate than forcing people to act. Stakeholders such as farmers who have been forced to change practices numerous times were more willing to work with watershed organizations once they realized that the process was voluntary and that they would not be forced to do anything. One farmer explained, “If it’s forced, I don’t think it would be better. I think that it would be worse.” Similarly, a watershed organization employee described this approach: “That's why we think people will agree to work with us: because it's not something imposed on them.” Since their roles were not to force stakeholders to act, watershed organizations were viewed as neutral and unbiased, which added to their legitimacy.

Despite these justifications, the voluntary process and reliance on moral compliance were also predominant concerns among interviewees in all watersheds and, stakeholder groups, and at all levels. In the absence of regulation or enforcement, collective responsibility was not adequate for realizing substantive results. As one organization staff member explained, reliance on the morality of citizens to act has limits:

There is a collective responsibility in saying that the environment is important, in saying that preserving the water is an important issue. But it's a long way [from acting]. On a certain level, you need to have a way to constrain people to do certain things.

Similar concerns were explicitly voiced by nearly half of all interviewees. A former mayor succinctly summarized this sentiment: “Everybody wants to go to heaven but nobody wants to die.” They explained that regulation is needed since a certain proportion of people will always break the law; values and education have limits. Tellingly, the individual who was intimately involved in developing the policy explained that the voluntary approach was a political compromise and an imperfect solution. They called it the “Achilles heel of the Québec Water Policy.”

There appears to be no easy solution because participation and implementation could not simply be made mandatory; this was not the appropriate role of watershed organizations and interviewees at all levels expressed that the collaborative process should not be mandatory or enforced. One farmer voiced this view: “It's not their role. Their role is to put people around the table.” Rather, the majority of interviewees perceived the voluntary role of watershed organizations as appropriate in relation to established institutions of representative democracy. That is, elected representatives should make decisions. A regional government representative from MAMOT explained their reasoning for why collaborative governance should be voluntary: “We cannot have two sets of regulations.” They continued, “Regulation is the object of governments.”

In the absence of enforcement, it is unclear what can ensure implementation of watershed plans. One agricultural union representative expressed that, since the process was voluntary, mechanisms to hold actors to account were needed:

It needs some kind of coercion to make it more effective, but coercion for every group. There's a need to have some kind of accountability for every group, so if you commit your organization [to actions] and fail to do them, there should be some kind of sanction. If there's no sanction it's everybody's problem and nobody's problem.

How the voluntary collaborative process and existing democratic institutions can fit best together remains unclear. Many stakeholders, especially provincial actors such as MDDELCC and ROBVQ, were actively searching for ways to improve this fit. Policy-makers were concerned that mandating implementation of watershed plans that were produced by voluntary participants would be problematic, but also that the legal status of watershed organizations was not ideal. The ongoing debate about whether RCMs should be mandated to incorporate watershed plans into their regional land-use plans illustrates these tensions. Since participation in watershed organizations and development of water master plans was voluntary, it was viewed as highly problematic to require elected RCM representatives to abide by those plans.

Nevertheless, public administrators, politicians and others at the Québec level recognized that it was important for collaboration to fit well with pre-existing strategies. One public administrator in a regional MDDELCC office explained that voluntary initiatives are one element, to be complemented by regulation, incentives, education and other measures:

It's all an equilibrium between regulation and a [voluntary] watershed group approach. But when you commit to plans and actions, you turn back and say, “you [the government has] the tools, regulations and education to change some things.” Some issues in environment will be changed by regulation, some by programs or initiatives, some by education. So it's a big puzzle to work on environment.

These attitudes hold promise that the government will support and improve how collaborative water governance in Québec fits with existing institutions.

Effectiveness

Concerns about the ability of collaborative water governance in Québec to address a common interest were paramount. Interviewees at all levels most commonly measured success against whether the process was effective and able to achieve tangible results. A majority of interviewees supported and agreed with watershed organizations’ roles and actions on the basis that they addressed water problems. Local stakeholders frequently justified watershed organizations by referring to projects the organizations had either implemented or were instrumental in making happen. They gave diverse examples including the revitalization of an urban creek; implementation of water monitoring, invasive species control and education programs; and coordination of a Green Corridor project in an agricultural region. One employee of an RCM who sat on a watershed organization’s board of directors expressed how the watershed organization had gained legitimacy by improving riparian zones: “It's not everybody who wants to go out and get the shovel themselves and do 3 km of shoreline, but ABV des 7 will do that.” Thus, implementation of projects with tangible environmental impacts was a prime contributor to the legitimacy of collaborative water governance at the watershed level.

Provincial actors and decision-makers also recognized that watershed organizations needed to demonstrate their effectiveness to gain legitimacy. One provincial policy-maker described how important tangible actions were to legitimacy: “To be on the field is the best way for them [watershed organizations] to be recognized. And for the population to know that they exist, that's the best way.” This role was justified despite organizations’ mandates only to facilitate the implementation of plans rather than to implement them (Québec 2002). Other interviewees from MDDELCC and representing watershed organizations at the Québec level agreed that, while implementing projects was not organizations’ primary role, it was critical to their legitimacy.

However, the future effectiveness of the QWP remains uncertain. This represents the policy’s largest potential deficit to legitimacy because its effective scale and impact are limited. The director of one watershed organization expressed frustration that watershed organizations are limited by the scale of projects they are able to implement: “Is this the best solution? I don't think so. We are doing small activities that in a small way show that we exist.” Watershed organizations do not have even a fraction of the capacities or resources they require to tackle many of the issues they are attempting to address. All interviewees questioned whether actors other than watershed organizations would implement plans, indicating that in the absence of results the process would lack legitimacy. If plans were not voluntarily implemented once MDDELCC approved them, they would have limited value: while plans and strategic visions were important, they would only be meaningful if they helped improve environmental conditions. A public administrator emphasized this concern: “it is probably the biggest question.” Concern at the Québec level over the long-term sustainability of the Québec model of collaborative governance reflects the need to implement actions rather than merely produce plans.

Complementary functions

Collaborative water governance was also justified in meeting the common interest because it performed functions complementary to the roles of existing actors such as governments, addressing social needs. These functions included producing a common vision, conducting environmental education, developing knowledge and shifting values within society. Interviewees also described watershed organizations’ policy, planning and coordination roles as important. Interviewees at all levels, especially a large majority of local stakeholders from all stakeholder groups in all three watersheds, shared this view that collaborative governance filled a social need beyond implementation of tangible actions. For example, the representative of an environmental organization in ABV des 7 expressed their approval of the collaborative process, which they referred to as “the production and promotion of a common vision and way of relating to the land.” This complementary role was further legitimized because interviewees viewed elected representatives as unable to address certain social needs. Economic, social and governmental interviewees all felt that watershed organizations were justified because they performed functions that governments could not; the production of a common good was beyond the scope and mandate of governments. Even those working in government saw collaborative water governance as complementary, in part because it was seen as desirable that citizens themselves act. A former mayor in ABV des 7 described the need for alternative models of collective action in addition to government: “Government fulfills some needs of citizens, but hopefully not all their needs.” They continued, “Citizens must themselves take the lead.” One RCM employee in that same watershed explained that watershed organizations were more able to present certain information and act in the public interest than elected representatives because they were at arms length from the government, whereas elected representatives were more politically restricted. A mayor in the SCABRIC watershed described the limits of government and the need for alternative ways to meet social needs: “I know government represents people, but we're just representing, we're not making somebody care about something. And I think that there's an element of taking care of something that is missing in all of this.” Thus, collaborative water governance may play a legitimate role complementary to that of governments.

Consent

The degree and way in which individuals demonstrated consent to collaborative water governance varied widely among watersheds, groups and levels. Local stakeholders expressed consent through diverse means including participation in meetings and consultations, collaboration on projects, implementation and incorporating watershed plans into regional plans. There was consistent participation on the boards of directors for each of the three watershed organizations; a large majority of the seats were filled and attendance was regular. Other key local stakeholders also expressed substantial consent. In all three watersheds, at least a few individuals strongly supported their watershed organization, advocating on its behalf. In COBAMIL, several well-known and respected community leaders and politicians associated themselves with and advocated on behalf of the watershed organization, enabling it to gain rapid and widespread support.

But beyond these core stakeholders in each watershed, the degree of consent varied widely between stakeholder groups. Municipal and RCM representatives were commonly and closely engaged, most participating regularly. Many actively partnered with watershed organizations, sharing resources, implementing projects and organizing community events, although a few in each watershed did not, often because of personal opinions. Social and environmental groups also supported each of the three watershed organizations, although their degree of consent often reflected how well their vision aligned with that of watershed organizations. The director of a Comité de zone d'intervention prioritaire (comité ZIP), an organization with a mandate similar to that of watershed organizations, approved of the watershed organization they worked with and had partnered with it on multiple projects. In contrast, the director of an environmental organization in ABV des 7 thought that the watershed organization should be more politically active and thus typically ignored or avoided the watershed organization. While a few economic stakeholders recognized watershed organizations, in most instances their lack of participation or recognition of the process reflected a clear legitimacy deficit. In agricultural territories, farmers regularly supported watershed organization initiatives, although the fraction of individuals participating was low. However, in all three watersheds the rest of the economic sector was notable for its lack of participation, or in some instances its dissent through withdrawal from participation. Industries such as forestry participated rarely if at all. Finally, consent was varied and lacking from citizens in general. While public consultations or events involved up to 200 participants, participation was typically much less, representing a small fraction of the population of a watershed. These examples demonstrate the varying degree and diverse means of local stakeholder consent.

Although consent may vary over time, the Québec government clearly demonstrated its consent to collaborative water governance in the past. Watershed organizations were set up with strong involvement and leadership from the Ministry of the Environment, which provided funding, guidance and expertise (Québec 2002; MDDEFP 2011). Again in 2009, the Québec government demonstrated further consent, doubling funding for watershed organizations to approximately CAD $125,000 per year, increasing the sizes of existing watershed territories, and creating seven new watershed organizations (MDDEFP 2011). That year, Québec also adopted a law to legally recognize water as a common resource for the people of Québec. This law identified the province of Québec as the guardian of water in the public interest through the Ministry of Environment (Cumyn 2009). Referring to water master plans, article 15 of this law requires that all governmental stakeholders, including municipalities and RCMs, must “take the plan into consideration when exercising their powers and duties under the law” (Québec 2009, 9). But the force of this law is limited since governmental stakeholders have no legal recourse for not implementing the plan. One government policy-maker explained that this law was more to demonstrate consent than a legal tool, recognizing of the importance of the water policy and watershed organizations.

However, in recent years the Québec government has limited, if not partially withdrawn, its consent for collaborative governance, reducing expertise, limiting recognition of watershed plans and maintaining limited funding. To follow and implement government policies, watershed organizations need regular access to and interaction with regional ministry representatives who understand local contexts but also government procedures. Moreover, decentralization requires support, expertise and funding that are often not available at the local level in smaller or more rural areas. Policy-makers and organization staff alike emphasized that watershed organizations did not have adequately close relationships with and support from MDDELCC to fulfill their mandates and ensure continued participation of stakeholders. One government administrator noted, “for real collaborative water governance all stakeholders have to be there and government is a stakeholder.” Pervasive reduction in government staff meant that all ministries that were interviewed were increasingly limited in the expertise and support they were able to provide. Participation of ministries on the board tasked with integrating different ministry responsibilities for monitoring and actions related to the QWP had also declined (Vérificateur général du Québec 2013). The Québec government has limited the funding and resources it provides to watershed organizations and for implementation, reflecting further lack of consent. Substantial implementation of watershed plans will require increased commitment and support from the Québec government (Vérificateur général du Québec 2013). Finally, MDDELCC’s recognition and use of watershed plans was also lacking. One organization staff member described how government consent to their plan was limited: “They recognize it. We have the ministry's seal. They approve it. But they are not using it at all.” A staff member from a different watershed organization expressed similar concerns: “The voluntary part in my opinion is pretty important. But that is not working because we send the water master plan to the government and every ministry reads it but they don't do anything with it.” Notwithstanding concerns voiced by some Québec-level interviewees that plans may be too long and inconsistent, MDDELCC’s failure to recognize or use watershed plans represents a significant lack of consent. In these ways, the Québec government has limited, or even withdrawn, its consent.

Discussion and conclusions

This research identified diverse sources and deficits of legitimacy for collaborative water governance in Québec, focusing on perspectives from multiple levels of decision-making. This multi-level perspective also revealed several novel insights, demonstrating how legitimacy can be used to understand and evaluate novel governance arrangements.

This multi-level perspective identified the major sources and deficits of legitimacy for collaborative governance in Québec, some of which were prominent across all levels. The voluntary nature of collaboration, its effectiveness and its fit with established institutions were closely related issues and, considered together, represented the greatest challenge to legitimacy. Voluntary norms coexisted uneasily alongside established democratic norms. The voluntary nature was a source of legitimacy but was at the same time a major deficit: the voluntary process did not conflict with established representative norms, but lacked mechanisms of enforcement and ways to hold stakeholders to account. This result in itself was not surprising since the lack of mechanisms for enforcement and accountability is a common deficit of voluntary processes (Huitema and Meijerink 2009; Cohen and Davidson 2011). However, concerns about voluntary collaboration and effectiveness are linked, since transferring responsibility from elected representatives who can be held accountable for their decisions to voluntary participants has often resulted in weak implementation and inaction (Pellizzoni 2004). This was problematic because substantive results - effectiveness - were critical to legitimacy, even while lack of implementation of watershed plans presented the greatest potential deficit. This result supports previous findings that collaborative initiatives need to be perceived as serving the common interest and “not just a ‘talking shop’” (Connelly 2011, 937). Others have similarly observed that the largest outstanding concern related to the legitimacy of collaborative governance may be the extent to which collaboration leads to substantive results (Baird et al. 2014).

Assessing legitimacy at different levels also helped identify important challenges and ways to pursue enhanced legitimacy. Different challenges were visible from perspectives at different levels. For instance, while local stakeholder perspectives did not indicate that inclusion was a major issue, perspectives on inclusion at the organization and policy levels revealed that collaboration was more suited to some issues and contexts than others. Thus, policy-makers and organization staff may need to consider collaborative approaches as one of many for addressing environmental problems. Collaborative governance and its fit with established institutions might be further improved by identifying for what issues and contexts collaboration is most and least effective. In particular, several perspectives, such as First Nations and large industries, were not included in this research. These groups did not participate in the watersheds studied, and more exhaustive sampling was beyond the scope of this research. Understanding and accounting for these perspectives will be critical to determining how appropriate collaborative governance is for different issues and contexts, as well as how other solutions can best complement and support collaborative governance. This research provided a tentative first step toward achieving this end.

Assessing legitimacy at different levels also revealed potential solutions. One novel finding was that watershed organizations gained legitimacy by fulfilling social functions that government itself was unable to, acting as trustees of the common good. This role has the potential to improve their complementarity because voluntary collaboration sidesteps longstanding debates about whether elected representatives should act simply as delegates for their constituents’ interests or as trustees of the common good (Urbinati and Warren 2008). While modern democracies rely heavily on the delegate model, favouring aggregate individual self-interests rather than the common good, collaborative governance may address larger social needs, helping to rebalance collective interests with individual self-interests. Further enabling watershed organizations to play complementary roles to those of governments will improve their legitimacy.

While tensions between new voluntary collaboration and established representative institutions remain unresolved, improving their fit through complementarity holds potential. Representative government and collaborative governance are not mutually exclusive (Urbinati and Warren 2008). Rather, as Booher (2004) argues, “collaborative governance practice is emerging as an augmentation to existing government, not a replacement” (41, emphasis added). Improving complementary methods of governing such as policies, regulation, legislation, incentives and education can contribute to improving the legitimacy and long-term effectiveness of collaborative governance. But responsibilities for these instruments are not adequately held at the watershed level. Rather, it is governments that possess the capacities and appropriate mandates to support and complement the strengths and weaknesses of collaborative governance in these ways. This consideration is important because voluntary approaches are often ineffective without regulatory intervention by governments (Layzer 2008). Indeed, the QWP states that watershed organizations are intended to complement government regulatory measures such as the user-pays and polluter-pays principles (Québec 2002). These measures have not been implemented to a significant degree, emphasizing that barriers between levels remain to making collaborative and representative governance truly complementary.

Moreover, perspectives on legality also revealed the importance of attending to the way collaborative governance is embedded within its larger institutional context. Legality was largely unproblematic for watershed organizations since they coevolved along with expectations from MDDELCC, but lack of regulation and enforcement related to established government institutions detracted from the legitimacy of watershed organizations. This suggests that if the role and purpose of collaborative governance is not explicit, it may be associated with existing practices, correctly or incorrectly, positively or negatively. The implication is that the legitimacy of collaborative governance is not separate from its institutional context and cannot be considered in isolation.

Understanding consent from a multi-level perspective revealed key insights and perhaps warrants the most attention. First, consent revealed a disconnect between perspectives at the local stakeholder and policy levels. Consent at the local stakeholder level was diverse but not a predominant concern, especially since a core group of individuals demonstrated significant consent in each watershed. But a lack of more widespread participation reflected limited knowledge and support of collaborative water governance, implying the need for both increased education by and visibility of watershed organizations. A lack of consent from the policy level not only exacerbated challenges at the watershed level but also inhibited improvements across Québec. A similar lack of government consent has been observed to hinder legitimacy and effectiveness elsewhere (Ribot 2002; Bell and Park 2006). Improving the fit and complementarity between collaborative governance and established institutions will require ongoing government support, while substantial implementation of watershed plans will also require increased commitment and support from the Québec government (Vérificateur général du Québec 2013). However, appropriate ways of expressing but also maintaining both public and government consent are lacking. It remains unclear how individuals or groups at different levels can or ought to demonstrate consent.

Source norms similarly reflected this disconnect between the local stakeholder- and policy-level perspectives, but indicated that key individuals are integral to establishing and improving legitimacy. Norms concerning the justifiable sources of decisions at the Québec level conflicted with norms at the watershed level as well as those articulated in the QWP. Organization staff attempted to balance source norms and improve inclusion. These results confirm previous findings that differing perspectives on what constitutes legitimacy, especially between local actors and provincial decision-makers, can create challenges for organization staff (Booher 2004). These intermediary coordinators may be stuck trying to resolve conflicting perspectives and justifications from policy-makers “above” and collaborators “below” (Newman et al. 2004). Further research is needed to more fully understand how individuals such as organization staff shape institutions and negotiate their legitimacy.

This analysis also suggests several questions important to understanding and improving the legitimacy of collaborative governance:

(1)	For what contexts and issues is collaborative governance most appropriate?
(2)	How can fit between established representative and novel collaborative institutions be improved?
(3)	What are the most appropriate ways for individuals and groups at different levels to demonstrate and maintain consent?
(4)	How do and can key actors (local leaders, organization staff and policy-makers) shape and improve legitimacy?

This approach can be used to establish and pursue enhanced legitimacy for novel governance arrangements, providing a way to identify and evaluate complex sociological barriers and opportunities that define the justification, support and, ultimately, long-term success of these arrangements. Attending to legitimacy at multiple levels can make the difference between solutions that are simply desirable and those that are politically viable. This has the potential to not only improve the legitimacy of existing solutions but also dramatically broaden society’s range of available options to address environmental problems. For these reasons, understanding and enhancing the legitimacy of institutions that improve the human–Earth relationship should be a priority.

Acknowledgements

This research would not have been possible without the generous participation of numerous interviewees. Peter Brown also provided critical feedback throughout this research. The authors would like to thank three anonymous reviewers for their comments that helped improve this article. Financial support was provided in part by a SSHRC Grant held by Jan Adamowski.