Who's in and who's out? Inclusion and exclusion in Canterbury's freshwater governance


  • Sylvia Nissen

    Corresponding author
    1. School of Geography & Environmental Science, Monash University, Melbourne, Australia
    • Note about authors: Sylvia Nissen holds a masters degree from the School of Geography and Environmental Science, Monash University. She is currently a PhD candidate in the Department of Political Science at Canterbury University.

      E-mail: sylvia.nissen@pg.canterbury.ac.nz

    Search for more papers by this author


The question of who is included, or who is excluded, from environmental governance arrangements is at the heart of debates of institutional legitimacy. The recent redefinition of representation as a claim to ‘speak for’ another provides an opportunity to re-examine the inclusiveness of environmental governance arrangements. This article builds on these developments by empirically evaluating the inclusiveness of three representative claims embedded in freshwater governance in Canterbury since 2002. The contribution argues that accounts of environmental governance must consider implicit forms of representation that can exclude or marginalise and raises some caveats relating to expansion of network governance.

The question of who is included, or who is excluded, from environmental governance arrangements is at the heart of debates of institutional legitimacy. The inclusion of representative members in a governance body, and the claims they make to act on behalf of wider populations, shapes both the process and outcomes of decision making (Fraser 2007; Hendriks 2008). Yet too often the implicit and explicit representative claims embedded in environmental governance arrangements are overlooked in both theory and policy practice (O'Neill 2001; Weale 2009).

To help extend discussions of institutional legitimacy in the context of environmental governance, this article empirically evaluates patterns of inclusion and exclusion in the mechanisms governing freshwater in Canterbury, New Zealand, since 2002. With respect to the concept of representation, the highly contested freshwater context in Canterbury provides a valuable means of evaluating the dilemmas of representation in practice. For the practice of freshwater governance in New Zealand, this paper seeks to provide a gentle caution regarding the recent widespread interest and adoption of collaborative governance by drawing attention to patterns of inclusion and exclusion in the (sometimes implicit) representative claims of members of the governance process.

This contribution builds on the recent redefinition of representation as a claim to ‘speak for’ or ‘act on behalf of’ another (Saward 2006). The enriched definition of representation helps to identify sources and mechanisms of representation in electoral and non-electoral contexts that can contribute to, or detract from, the legitimacy of institutional arrangements governing the environment (Junker et al. 2007; Lövbrand et al. 2009; Biermann & Gupta 2011; Seymour et al. 2011). Distinguishing facets of representation from other aspects of governance arrangements additionally offers a unique perspective of the role different types of representation could play in the future of Canterbury's freshwater governance.

This article proceeds in three steps. First, the role of inclusiveness in shaping institutional legitimacy is reviewed, followed by an overview of the redefinition of representation and its relationship to inclusion. The second section then evaluates the legitimacy of practices of inclusion in the three types of representative claims in Canterbury (the councillors, the commissioners and the collaborative frameworks). Building off this appraisal, the third section then evaluates the strengths and weakness of the inclusiveness of the representative claims, and practical suggestions are made for strengthening the legitimacy of Canterbury's water governance.

The analysis in this article is based on three types of sources: interviews, participant observation and document analysis. Semi-structured, hour-long interviews conducted in mid 2012 to early 2013 with four key Environment Canterbury staff members, three former councillors, three stakeholders involved in the collaborative process and a national politician. Pseudonyms are used throughout this article to ensure the participants' confidentiality. The interviews provided an opportunity for the participant to freely express ideas and were particularly adept at examining the dynamics of delicate issues. The second source was participant observation. Four zone committee meetings were also observed from November to December 2012, which provided rich insights for this research into the social relations and processes of the collaborative frameworks.1 Third, a wide array of over 270 relevant documentary sources was also drawn on to help verify findings.

Legitimacy, inclusion and representation: Three concepts intertwined

The legitimacy of systems of environmental governance has been of increasing concern in both scholarly debate and political practice as legitimacy provides the ‘moral basis of political authority’ in modern political systems (Lövbrand et al. 2009). It is essentially concerned with why people should consent, agree and obey decisions of authority (Bäckstrand 2006; Biermann & Gupta 2011).2

The legitimacy of environmental governance arrangements is shaped by two types of inclusiveness. First, inclusion influences ‘input legitimacy’ – the justice and fairness of the procedural characteristics of decision making – by affecting the openness and accessibility of political arrangements to those affected by decisions (Young 2000; Bäckstrand 2006; Fraser 2007; Hendriks 2008; Biermann & Gupta 2011). At the core of this understanding is an acute awareness of the potential for marginalisation, disempowerment and disaffection through the exclusion of affected voices from decision making (Phillips 1995; Williams 1998; Young 2000). This exclusion is recognised to occur externally through restrictions on membership or lack of economic and social resources. It can also come about internally through self-exclusion based on concerns of co-option, membership or potential outcomes (Young 2000; Hendriks 2008).

Second, inclusiveness also affects the ‘output legitimacy’, or (perceived) effectiveness, of a policy process by shaping the dynamics and interactions of a governance body (Bäckstrand 2006; Saward 2006; Biermann & Gupta 2011). From this perspective, inclusion relates to policy outcomes. What is important from this stance is the ability of members of the governance process to solve the problem at hand in a manner that is inclusive for those concerned (Hendriks 2008; Reed 2008). In other words, these theorists argue that it is not necessary to include all affected in decision processes, so long as policy outcomes incorporate their interests. This interpretation of inclusion stems from concern that inclusive processes do not necessarily engender effective, efficient and durable outcomes (Delmas & Young 2009; Innes & Booher 2010).

The concept of representation provides a versatile analytical frame through which to examine patterns of inclusiveness, both in relation to input and output legitimacy. Representation has gone through a ‘renaissance’ over the past decade within democratic theory, which has broadened and enriched the concept (Hendriks 2009; Maia 2012). Representation has been redefined as a dynamic process of claim making and of agency on the part of both the representative and the represented (Saward 2006). At the same time, the redefinition of representation has extended the concept to processes that occur both within and outside of electoral systems, which has significantly widened the lens through which representation can be examined (Saward 2009; Hendriks 2009; Maia 2012). Representative claims – intentional or otherwise – matter because it determines the perspectives, world views and experiences that are brought into the decision making process (Young 2000; Hendriks 2008). The redefinition of representation has made the concept significantly more versatile and a relevant framework for analysing overlapping claims made by a variety of actors in diverse forums and institutional frameworks (Alonso et al. 2011).

Thus far, however, the literatures on representation and environmental governance have tended to talk past, rather than engage with, one another. ‘Green’ theorists have often rejected or ‘demoted’ representation as too flawed or problematic in favour of direct participatory modes of governance (Barry 1999; Eckersley 2000; Baber & Bartlett 2005). This separation of literatures has been at the expense of meaningful and explicit analysis of the inclusiveness of representative claims embedded in governance process, including of ‘new’ forms of governance within and beyond the state (O'Neill 2001; Weale 2009). Using the case study of water governance in Canterbury, this article distinguishes representative claims from other aspects of the governing frameworks to identify sources and mechanisms of representation that can contribute to enhanced legitimacy of environmental governance arrangements.

Inclusion in practice: Governing water in Canterbury

Issues of inclusion, representation and legitimacy have become tangled in Canterbury's ‘water woes’ (Robb 2000; Bidwell 2006; Jenkins 2007; Taylor 2010). Canterbury's surface and groundwater resources have been strained by land-use change and intensification of farming practices (ECan 2010, 2011). The region has experienced an almost 10-fold increase in the number of dairy cows since the early 1990s (Pangborn & Woodford 2011, p. 83). Farming has also intensified; Canterbury has the largest number of cows per hectare of land in New Zealand (LIC & DairyNZ 2011, p. 14). Increased competition for water coupled with rising levels of water contamination has led to growing demand for improved governance mechanisms to sustainably govern Canterbury's water.

The last decade has seen three types of governance for Canterbury's water (Fig. 1). The first was provided by a democratically elected council within a regional governing body, Environment Canterbury (ECan). ECan is the regional council of the Canterbury region and holds responsibility for local natural and physical resources, including freshwater (Bush 1992; Palmer 2012). From 1989 until 2010, up to 14 councillors were democratically elected by wards within the region every 3 years (ECan 2009; Palmer 2012).

Figure 1.

The three major changes to the structures governing Canterbury's water in the period 1989 to2012, with reference to the councillors, commissioners and the four stages of the collaborative frameworks.

Alongside the council, collaborative governance frameworks developed in four stages from 2004, which directly brought members of key stakeholder, indigenous and activist groups among others into the governance process. The frameworks reflect an international shift towards network governance involving both public actors (politicians and administrators) and non-public actors of differing natures (stakeholders, interest representatives, experts, firms and so on) (Delmas & Young 2009; Innes & Booher 2010; Papadopoulos 2010). Following a central government document published in 2002, which later became known as Stage 1 of the Canterbury Strategic Water Study (CSWS) (Morgan et al. 2002), local leaders commissioned Stage 2 of the CSWS to find and identify water storage projects and their feasibility. A Steering Group of local leaders and key stakeholders commenced in 2004 to examine whether environmental needs and potential water demands could be met through the use of water storage systems (Dark et al. 2008).A third stage of the Canterbury Water Management Strategy (CWMS) was launched in 2006, which evaluated the storage options advanced by the Steering Group(Whitehouse et al. 2008). Building from the success of the CSWS, the collaborative frameworks were significantly expanded in 2009 to encompass a wider range of environmental and social values, under the title of the CWMS (Canterbury Water 2009; Lomax et al. 2010).

A third form of governance was introduced in 2010 with the replacement of the elected council of ECan with commissioners appointed by central government. At the same time as the collaborative frameworks were being developed, central government was increasingly concerned with the governance of freshwater in Canterbury, specifically the performance of ECan (Ministry for the Environment 2008, 2009). In April 2010, central government ministers replaced ECan's 14 elected councillors with seven appointed commissioners (DIA 2010; Minister for the Environment 2010). The announcement in September 2012 that the commissioners' term would be extended to 2016 with a review of Canterbury's governing structures in 2014 has left the form of the structures governing Canterbury's freshwater after 2016 unclear (Carter & Adams 2012).

Representative claims in Canterbury

Members of each form of governance presented representative claims that encompassed different groups and perspectives within the Canterbury region. These patterns of inclusion are summarised in Figure 2, with thicker lines indicating a greater number of members that made claims to speak for a particular group.

Figure 2.

Inclusion of groups in the representative claims of the councillors, collaborative frameworks and commissioners, with thicker lines indicating a greater number of members who made claims to speak for a particular group

The broadest claims of representation were presented by the democratic councillors who claimed to act on behalf of communities within the region. Despite common electoral processes, the councillors' claims of inclusion frequently went beyond the boundaries of their electoral wards. Some councillors claimed to stand for the region as a whole; Councillor Jane Demeter (2010) proffered that she stood ‘in the best interests of the Canterbury region – not one interest, not one area’ (also Kane 2010; Councillor 2, interview, 2012). Other councillors emphasised the descriptive quality of the council based on interests or perspectives (Councillor 1, interview, 2012; Councillor 3, interview, 2013).The breadth of these claims reinforces that representation is a representative's idea of a constituency, rather than the constituency itself (Castree & MacMillan 2004; Saward 2006).

The commissioners' representative claims also broadly encompassed citizens within the region and were very similar in expansiveness to the claims of the councillors. The commissioners claimed to act on behalf of the public, despite not being directly authorised by or accountable to the community. David Caygill, for instance, described the role of a commissioner as ‘a very different way of serving the public’, while a staff member reflected that the commissioners sought to ‘act for the benefit of all Cantabrians to the best of their abilities’ (Public Sector 2010, p. 20; SECS4, interview, 2012). While the foundation for the councillors' claim was the ballot box, for the commissioners it was extensive consultation with the community (Staff 4, interview, 2012; Staff 2, interview, 2012). Staff members consequently appraised the inclusiveness of the commissioners' ‘benign dictatorship’ as more aligned to a ‘community-based consultation process’ (Staff 2, interview, 2012; McCrone 2012).

In addition to these broad claims, the commissioners were promoted to the Canterbury public based on their descriptive representation – the characteristics of representatives that reflect or mirror affected populations (Phillips 1995; Young 2000; Hendriks 2009). The Minister for the Environment, Dr Nick Smith, asserted that the commissioners had a ‘balance of agricultural, environmental and electricity expertise’ as well as ‘representation from both North and South Canterbury communities’ (The Press 2010a). The view that the commissioners roughly represented a range of interests in Canterbury was reinforced by staff members who noted that the commissioners each brought ‘completely different perspectives’ and a ‘broad skill set’ to the policy process (Staff 2, interview, 2012; Staff 4, interview, 2012).

For both the councillors and commissioners, these encompassing but somewhat ambiguous claims of representation meant that few groups were explicitly excluded from their approach. Yet, at the same time, the encompassing nature of the councillors and commissioners' claims fostered scepticism among some policy actors of their actual inclusiveness in practice. Staff members identified the councillors as ‘good people’ with ‘good intentions’ but narrowed their inclusiveness to at best ‘somewhat of a reflection of the areas that elected them’ (Staff 3, interview, 2012; Staff 2, interview, 2012). These accounts of the councillors resonate with critiques in green theory of democratic representation that argue that broad claims of elected representatives are too general, with the potential to marginalise those with a stake in a policy issue (Reed 2008; Bengtsson & Klintman 2010).

This scepticism of the councillors' inclusiveness was reinforced by the ‘volatile’ relationships within the council, which undermined perceptions of inclusiveness in outcomes (Gorman 2009; Stakeholder 1, interview, 2012). ‘Bizarre’ alliances formed between council members that did not necessarily follow ideological or interest boundaries, which led some councillors to become ‘disenfranchised’, ‘frustrated’ or ‘alienated’ by the process (Staff 1, interview, 2012; Councillor 1, interview, 2012; Staff 3, interview, 2012).While these interactions did not necessarily prevent the councillors from achieving ‘reasonable decisions’, they nevertheless shed doubt on the inclusiveness and effectiveness of the outcomes (Staff 1, interview, 2012; also Staff 4, interview, 2012; Stakeholder 3, interview, 2012).

Some interviewees were also dubious of the inclusiveness of the commissioners' claims. Particularly in the wake of the ‘sacking’ of the council, the commissioners were described as ‘dictatorial’ and ‘there to achieve a pre-determined agenda’ (Councillor 2, interview, 2012; Councillor 1, interview, 2012; also Councillor 3, interview, 2013). Certainly, the commissioners brought a sense of ‘urgency’ and timeliness to ECan and publically displayed stable interactions with high levels of internal trust (Staff 2, interview, 2012; Staff 4, interview, 2012; Stakeholder 2, interview, 2012). These processes contributed to perceptions of the commissioners' effectiveness – but also reinforced notions amongst some policy actors that they were overriding or marginalising particular interests and perspectives (Staff 3, interview, 2012). As one stakeholder wryly commented, ‘they do what they think is right, which is certainly not what everyone wants’ (Stakeholder 3, interview, 2012). Taken together with criticisms of the councillors, a common critique between the two claims of representation was of exclusion as a result of the breadth of their representative claim.

Building from this criticism, the collaborative frameworks took an opposite approach and directly included key stakeholders, experts, traditional claimants and activists among others. Inclusion in the collaborative frameworks was based on the requirement to bring the ‘right’ individuals into governance arrangements to develop collaboration and lead to some consensus (Staff 1, interview, 2012; Staff 4, interview, 2012; Hendriks 2008). High levels of influence, status and recognition within the community were valued characteristics of members, with one stakeholder remarking that he was included by ‘just making a lot of noise at the time’ (Stakeholder 1, interview, 2012; also Stakeholder 3, interview, 2012; Staff 1, interview, 2012).

In theory, the collaborative frameworks were explicitly non-representative. The facilitators of the frameworks asserted that those individuals brought into the decision making process were required to not represent a specific group or interest in decision making. Instead, members were asked to act in an individual capacity (Whitehouse et al. 2008; Staff 1, interview, 2012). In practice, however, and like other forms of collaborative governance elsewhere, members made a variety of claims to act on behalf of wider populations. Often these claims went beyond the boundaries of their organisation or locality, such as for ‘farmers’, ‘the environment’, ‘local communities’, ‘Māori’ or ‘businesses’ (Stakeholder 3, interview, 2012; Stakeholder 1, interview, 2012; Stakeholder 2, interview, 2012; Field notes 2012).

Direct inclusion of key stakeholders provided a strong base for legitimacy in the eyes of those involved. Like the commissioners, the architects of the frameworks emphasised the descriptive representative elements of members. For instance, Bede O'Malley (2008), chairman of the Stage 2 Steering Group, asserted that the Group had ‘a wide base of stakeholders who represent the varied opinions on Canterbury's water both locally and regionally’ (also Staff 1, interview, 2012; Staff 4, interview, 2012). Almost all policy actors interviewed held an impression that the collaborative frameworks were inclusive with ‘broad membership’ (Stakeholder 2, interview, 2012; also Stakeholder 3, interview, 2012; Staff 4, interview, 2012). The direct involvement of stakeholder interests also fostered perceptions of the frameworks as highly effective and inclusive in their outcomes (Stakeholder 2, interview, 2012; Staff 1, interview, 2012; Staff 3, interview, 2012; Stakeholder 3, interview 2012). Although there remained some contestation over the forcefulness of policy outputs, stakeholders commented that, while interactions were initially quite volatile, mutual understanding developed and in some cases brought a ‘shifting towards the middle’ (Stakeholder 1, interview, 2012; Stakeholder 3, interview, 2012).

Comparison of the frameworks with democratic ideals reveals embedded patterns of exclusion relating to input legitimacy. The intent to bring key interests and perspectives to decision making significantly narrowed the inclusiveness of membership, which is a critique common to network governance more generally (Sørensen 2005; Hendriks 2008). One concern was of the relative inclusiveness of interests within the membership of the frameworks, shown in Figure 3. Specifically, the collaborative frameworks relied heavily on voluntary time and energy. Several interviewees consequently argued that membership became skewed towards the ‘economic and farming interests’ (Stakeholder 3, interview, 2012; Councillor 3, interview, 2013; Lomax et al. 2010). Self-exclusion underwrote membership, with some actors choosing to work outside the frameworks as a result of resource limitations (Stakeholder 1, interview, 2012; Staff 1, interview, 2012; Staff 4, interview, 2012).

Figure 3.

(a) The changing membership of Stage 2, Stage 3 and Stage 4 of the collaborative frameworks by sector in the period 2004–2012. (b) The membership of Stage 2, Stage 3 and Stage 4 of the collaborative frameworks by gender in the period 2004–2012.

Less clearly articulated or developed perspectives were also not present in the frameworks. As one stakeholder contended, ‘who speaks for your average man or woman on the street that isn't a farmer or an environmentalist or whatever?’ (Stakeholder 2, interview, 2012; also Stakeholder 1, interview, 2012; Councillor 2, interview, 2012; Lomax et al. 2010). Tied into this perspective were concerns about the under-representation of Christchurch residents in all zone committees aside from the Christchurch-West Melton committee, as the emphasis on local governance prioritised ‘user’ interests (Councillor 3, interview 2013). The less defined perspectives of women, children and future generations were also identified as under-represented by some interviewees (Fig. 3) (Councillor 1, interview, 2012; Councillor 3, interview, 2013; Stakeholder 2, interview, 2012).

Sources and mechanisms of legitimacy

From the analysis so far, it is clear that there were great similarities in the inclusiveness of the claims of the councillors and the commissioners. Both presented remarkably comparable claims to represent an interpretation of the ‘best’ or ‘public’ interest of the region either collectively or individually. It is worth noting, however, that these common patterns of inclusiveness do not make the legitimacy of the representative claims identical. The legitimacy of representation is grounded not only in inclusiveness, but also in meaningful practices of authorisation and accountability to verify and validate claims (Phillips 1995; Castree & MacMillan 2004; Maia 2012). Authorisation and accountability formed the primary point of difference between the councillors and the commissioners, not claims of inclusion. Although it is outside of the scope of this article, a full account of legitimacy must include both of these facets.

Inclusion nevertheless played a significant role in the development of legitimacy in Canterbury. Of particular impact was the development of multiple sites and types of inclusion in institutional arrangements (Hendriks 2008; Maia 2012). Broad claims to represent entire communities or the region as a whole were brought into the governance process by the councillors and commissioners. At the same time, key stakeholders were given direct voice in the collaborative frameworks, which fostered perceptions of inclusiveness of more specific interests and perspectives. Consultation also accompanied all three types of representation to varying extents to broaden the scope of inclusiveness (Staff 1, interview, 2012; Staff 4, interview, 2012; Staff 2, interview, 2012). This dual mode of governance provided forums for different types and styles of representative claim to be heard.

Although the councillors presented highly encompassing claims of representation, they were nevertheless challenged for their exclusiveness (Stakeholder 2, interview, 2012; Staff 1, interview, 2012). The broad claims of the councillors reflect a tendency of representative claimants to invoke groups of people as if they were units or homogenous aggregates (Papadopoulos 2010; Maia 2012). Perhaps not surprisingly, stakeholders criticised the representative claims as too general and therefore exclusive of specific – and arguably the most relevant – interests (Stakeholder 1, interview, 2012; Staff 4, interview, 2012; Reed 2008). A particular theme was a perception of disconnect between the council as representatives and what was happening ‘on the ground’ (Stakeholder 2, interview, 2012).

The commissioners made similarly broad representative claims as the councillors but did not attract the same levels criticism from stakeholders when they replaced the councillors. The primary reason for this, advanced by several interviewees, was that the collaborative process ‘muted’ or ‘offset’ some of the criticism of the commissioners by providing a forum for expression of specific interests (Staff 3, interview, 2012; Staff 4, interview, 2012; also Councillor 1, interview, 2012; Stakeholder 2, interview, 2012). The engagement of the commissioners shortly after the significant extension of the collaborative frameworks under the CWMS meant that many of the ‘most vocal’ were ‘kept busy in the committees’ (Staff 4, interview, 2012).

Together with the perceived effectiveness of the collaborative frameworks, the reduction of criticism suggests that multiple forums and levels of governance in institutional design may hold the potential to strengthen both input and output legitimacy. For Canterbury, this assertion reaffirms the value of the region's dual model of governance, as both broad and more specific claims to representation are given forums in which to be heard. However, this source of legitimacy comes with some significant caveats. These caveats relate to: (i) distance in decision making and (ii) ‘masking’ exclusion and ‘muting’ opposition. They provide points of caution in the adoption of collaborative governance and boundaries for their advancement within Canterbury's institutional arrangements.

Distance in decision making

Multiple sites and types of inclusion raise the issue of ‘distance’ in decision making (Dryzek & Stevenson 2011). Although direct inclusion of stakeholders was perceived as highly inclusive and effective, it can be problematic in relation to input legitimacy. The merging of strong and highly resourced stakeholder interests and empowered space – where authoritative collective decisions get produced – can result in the exclusion of alternate and less powerful actors and perspectives that are not able to organise themselves sufficiently in civil society organisations (Papadopoulos 2010; Biermann & Gupta 2011; Dryzek & Stevenson 2011, p. 1867). The deliberate inclusion of key stakeholders (often the most well-resourced and most-vocal) in the collaborative frameworks (the empowered space) marginalised less articulated perspectives and policy actors with limited resources (Stakeholder 2, interview, 2012; Stakeholder 1, interview, 2012; Councillor 2, interview, 2012). In other words, it further empowered those already most empowered.

These concerns suggest that ‘critical distance’ between stakeholder interests and empowered space is an important facet of institutional design (Dryzek & Stevenson 2011). This idea posits that strong stakeholders need to remain separate from direct involvement in empowered decision-making space in order to allow for less well-articulated voices and concerns to be heard. It is commonly suggested in democratic scholarship on network governance that a forum be set up distinct from the empowered decision-making space for stakeholder interests to be heard but that final decisions are made by elected politicians to ‘democratically anchor’ the frameworks via ‘coastal patrol’ (Sørensen 2005; Hendriks 2008, p. 1022). Both these proposals seek to ensure that less defined or powerful voices and perspectives do not become crowded out by stronger and more specific stakeholder interests.

In Canterbury, critical distance was initially maintained between stakeholders and empowered space. The collaborative frameworks provided a forum for stakeholders to express perspectives, working relationships to develop and solutions to be discussed (Staff 1, interview, 2012; Stakeholder 1, interview, 2012). However, these forums were not given direct legislative power, as the collaborative committees presented recommendations to the councillors or commissioners who then took them into account in final decisions (Councillor 3, interview, 2013). This model kept open the possibility for less organised interests to be represented in the more encompassing claims of the councillors or commissioners (Sørensen 2005; Hendriks 2008).

The heralded success of the collaborative frameworks and the subsequent push for their expansion both in Canterbury and other regions has not come with an equal recognition of their exclusionary potential. The boundary between stakeholders and empowered space was narrowed by the ECan Act 2010, which gave the collaborative frameworks under the Canterbury Water Management Strategy formal legal status (Minister for the Environment 2010). As one former councillor commented:

The scope of the zone committees has expanded. When they were set up it was never intended that they develop the regional plans and policies on water allocation and water quality limits and that the committees would sign off on those provisions. That was the role of the Council. (Councillor 3, interview, 2013)

Additionally, the removal of the councillors in 2010 made the exclusionary facets of the collaborative frameworks more critical as there is little to link members with systems of democracy and less of a ‘check and balance’ on the committees (Councillor 3, interview, 2013). Interest in collaborative governance has also gained momentum for other local councils in New Zealand (Staff 2, interview, 2012; National politician 1, interview, 2012).

Masking exclusion and muting opposition: Descriptive representation with a twist

Exclusion was an inherent part, to varying extents, of all governance mechanisms in Canterbury (Fraser 2007; Hendriks 2008). To strengthen input legitimacy and draw attention away from patterns of exclusion, advocates of the collaborative frameworks and commissioners, in particular, stressed the descriptive elements of the membership. The geographic location, occupation, organisation affiliation and indigenous affiliation of members were emphasised when ‘selling’ the modes of governance to the wider public (O'Malley 2008; The Press 2010b). This form of inclusion provided a means to indicate to those excluded from the governance process that those included shared some characteristics and affiliations (Staff 1, interview, 2012; Whitehouse et al. 2008).

This interpretation of descriptive representation is particularly shallow. Several interviewees drew attention to the limits of symbolic descriptive representation in Canterbury. In relation to the collaborative frameworks, one stakeholder reflected that:

I bring [to the process] my own single set of views and experiences and networks … sure, they're valid and all that, but they're just my view. (Stakeholder 3, interview, 2012)

The stakeholder went on to comment that despite ‘a whole lot of info fed at you’, decision making was ‘not really a two way thing’ (Stakeholder 3, interview, 2012). The narrow ties between symbolic representatives and constituents were also recognised by the commissioners as a key restraint of their governance, as they did not have the same connection to the community as the councillors (Staff 3, interview, 2012).

The veneer of inclusiveness additionally led to a paradox in roles within the collaborative frameworks. Members were brought into the groups and committees in part due to their interests and characteristics but were then asked to leave those interests and characteristics at the door (Staff 1, interview, 2012; Whitehouse et al. 2008). The dual messages appear to have developed multiple interpretations of the role members undertook in the governance process. Some interviewees clearly acted as representative of their interests, with their links to an organisation ‘never a closed door’ (Stakeholder 1, interview, 2012). Others, particularly if they had no formal ties to an organisation, acted on ‘whatever I thought sounded about right’ (Stakeholder 3, interview, 2012). An awkward middle ground also existed between those two positions, with tension between a member's role as a representative of an interest or organisation and their own individual judgement (Stakeholder 2, interview, 2012; Staff 3, interview, 2012).

More critically, symbolic descriptive representation was framed as a means to minimise external opposition and limit controversy. Particularly in relation to the collaborative frameworks, interviewees noted a number of practical benefits of descriptive representation in terms of its ability to ‘mute’ or ‘offset’ stakeholder criticism of governance arrangements (Staff 4, interview, 2012; Staff 2, interview, 2012). Some interviewees directly endorsed the frameworks for making the final outcomes of decision making ‘less surprising and controversial’ for key stakeholders and therefore ‘more likely to be accepted’ (Staff 2, interview, 2012; Staff 4, interview, 2012).

These motivations regarding stakeholder involvement are a far cry from the role of descriptive representation envisioned by democratic theorists as a means to bring substantive contributions and perspectives into the debate and to develop equality and equity (Phillips 1995; Williams 1998; Young 2000). Indeed, Carolyn Hendriks (2009, p. 709) has described the practice as ‘dangerously manipulative’ in that it can signal that decisions have been inclusive, when the reverse may be true. Stakeholders are recognised for their potential to blackmail through threat of ‘voice’, and ‘muting’ stakeholders via inclusion in policy forums could distract policy actors from serious breaches of input and output legitimacy at a broader scale (Papadopoulos 2010). In this light, the provision to keep the most vocal voices ‘busy’ through direct involvement in alternative forums could be attractive in institutional design for superficial or even subversive reasons of minimising conflict and opposition.

Fostering legitimacy in Canterbury

There are a range of possibilities to counter the caveats raised in this analysis. Foremost is for the collaborative frameworks to be accepted and addressed, at least partially, as a representative body for stakeholders and not the region as a whole (Sørensen 2005). This understanding acknowledges the strengths of the frameworks as a forum for direct involvement of key stakeholders. However, it also recognises the limitations of their inclusiveness and reinforces the value of ensuring ‘critical distance’ is maintained between the frameworks and empowered space.

The limitations with shallow descriptive representation suggest that a key challenge for the institutional design of network governance is to develop mechanisms to deliberately bring the frameworks closer to the public realm. One possibility would be to reinstate the councillors to provide democratic oversight for the frameworks (Sørensen 2005; Hendriks 2008). Changes could also be made to the design of the collaborative frameworks. In Canterbury, the public meetings and minutes of the committees are commendable, coupled with the significant consultative processes prior to decision making. These mechanisms could be extended by (Sørensen 2005; Hendriks 2008; Montanaro 2008):

  • developing more transparent and democratic selection processes, including granting stakeholders the opportunity to influence the selection of members;
  • acknowledging and respecting ties of members of the frameworks to organisations and interests, alongside the role of members to act for the ‘greater good’ (Stakeholder 3, interview, 2012);
  • furthering public dialogue with the collaborative frameworks via more direct forms of deliberative citizen engagement, such as citizen fora or juries; and
  • encouraging responsiveness on the part of members of the frameworks to stakeholder criticism and public dialogue processes.

The potential for the inclusiveness of the collaborative frameworks to become superficial also suggests that multiple sites of representation need to be adopted in a manner that meaningfully seeks to strengthen inclusiveness, not just of those with greatest ‘voice’. Processes are already in place in Canterbury to improve inclusiveness but could additionally include (Phillips 1995; Sørensen 2005; Hendriks 2008):

  • supporting affirmative action to include under-represented or marginalised groups, such as women, youth, Christchurch residents or less defined interests;
  • incorporating historians and social scientists into membership along with scientific experts; and
  • assisting marginalised perspectives with resources and political competencies.

These actions would serve to help reframe policy debate, which is recognised as a ‘powerful but subtle’ means to influence the way members of the collaborative frameworks construct their interests and their relationships to the broader community (Hendriks 2008, p. 1025).


Choosing who's in and who's out

The question of ‘who's in’ and ‘who's out’ is at the heart of debates over institutional legitimacy for environmental governance. This article examined the inclusiveness of representative claims as a site of legitimacy, which shapes both the input and output legitimacy of governance arrangements. Too often these explicit and implicit processes are overlooked in theory and policy practice, which leaves sites of exclusion and marginalisation under-scrutinised.

Underlying this article has been a question of institutional design: how can we create mechanisms to govern our environment in a way that engenders meaningful outcomes and is also just and fair? The findings of the case study of water governance in Canterbury suggest that multiple sites and type of inclusion hold the potential to improve both input and output legitimacy, so long as inclusion of stakeholders is adopted with healthy caution. In particular, three caveats were raised in relation to distance in decision making, the ‘masking’ of exclusion and the ‘muting’ of opposition, which provide boundaries for their advancement within Canterbury's institutional arrangements.

Stemming from these sources of legitimacy, this article provided suggestions for strengthening legitimacy in Canterbury's water governance in light of ongoing governance reforms. These suggestions also speak to ongoing changes to freshwater governance in New Zealand. While the dual model of governance should be applauded, it is pertinent that clear separation be maintained between frameworks and empowered space grounded in recognition of their exclusionary aspects. The democratic councillors could also be reinstated to allow Cantabrians to sanction false claims of the ‘public interest’ and to democratically anchor the collaborative frameworks. The democratic credentials of the collaborative frameworks could also be strengthened by supporting affirmative action to extend inclusiveness and by deliberately connecting members of the frameworks to stakeholder groups and broader public dialogue processes.


  1. 1

    Christchurch-West Melton, Waimakariri, Banks Peninsula and Orari-Opihi zone committees.

  2. 2

    Legitimacy can be both prescriptive and descriptive. Prescriptive theories of legitimacy set out general criteria derived from political theories against which modes of governance can be appraised (Bäckstrand 2006; Biermann & Gupta 2011). By contrast, descriptive accounts of legitimacy refer to how well institutions correspond to cultural and historical belief systems (Bäckstrand 2006; Lövbrand et al. 2009; Maia 2012). Legitimacy is approached from both a prescriptive and descriptive perspective in this article not only to overcome difficulties in measuring tangible outcomes, but also to surmount limitations in the subjective perspectives of policy actors (Lövbrand et al. 2009; Biermann & Gupta 2011).