Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Filling the void : struggles over implementing freshwater policy in Aotearoa New Zealand Tadaki, Marc Yoshio 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2019_february_tadaki_marc.pdf [ 3.25MB ]
JSON: 24-1.0373468.json
JSON-LD: 24-1.0373468-ld.json
RDF/XML (Pretty): 24-1.0373468-rdf.xml
RDF/JSON: 24-1.0373468-rdf.json
Turtle: 24-1.0373468-turtle.txt
N-Triples: 24-1.0373468-rdf-ntriples.txt
Original Record: 24-1.0373468-source.json
Full Text

Full Text

FILLING THE VOID: STRUGGLES OVER IMPLEMENTING FRESHWATER POLICY IN AOTEAROA NEW ZEALAND  by  MARC YOSHIO TADAKI B.Sc.(Hons.), The University of Auckland, 2011 M.Sc., The University of Auckland, 2012  A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Geography)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  November 2018  © Marc Yoshio Tadaki, 2018ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled: Filling the void: struggles over implementing freshwater policy in Aotearoa New Zealand  submitted by Marc Yoshio Tadaki  in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Geography  Examining Committee: Trevor Barnes, Geography Co-supervisor Eugene McCann, Geography, Simon Fraser University Supervisory Committee Member Jamie Peck, Geography Supervisory Committee Member Leila Harris, Resources, Environment and Sustainability University Examiner Natasha Affolder, Law University Examiner   iii  Additional Supervisory Committee Members: Kai Chan, Resources, Environment and Sustainability Co-supervisor    iv  Abstract This dissertation analyzes how the implementation of environmental policy is shaped by struggles over interpretation. Policy implementation is not a linear or mechanistic process, but is influenced by the subjective interpretation of policy concepts and how those relate to implementation practices. At the same time, however, not all interpretations are equal, and some are more influential than others. While there may be an initial ‘institutional void’ in which many interpretations are considered legitimate, political actors will attempt to structure this void by narrowing down the range of acceptable policy interpretations. The following chapters investigate how actors attempt to structure the institutional void of policy implementation. Chapter 2 situates the meaning of freshwater policy within historical debates about freshwater and its connection to colonial, biophysical, economic, and regulatory problems. Chapters 3-5 explore how three sets of actors engage in interpretive channelization by crafting and circulating their preferred interpretations of policy concepts in the implementation process. The state uses a diverse menu of regulatory and non-regulatory mechanisms to require and guide local authorities to implement policy in a certain way. Local governments face unique circumstances and must narrate the policy requirements into alignment with the interests of local organizations. Non-state experts such as intermediaries can augment local government interpretations by providing research and technical advice, but can also engage antagonistically by challenging interpretations in court or by using spatial inter-referencing to weaken the power of specific interpretations.  By treating the implementation of policy as amenable to political struggle, a clearer political diagnosis of opportunities and constraints can be conducted. Within each of the channelizing processes described here, both constraint and opportunity exist: even a neoliberal government must respond to public aspirations for water; politicians influence but cannot completely control bureaucrats within the state apparatus; local governments experience v  progressive as well as conservative contextual forces; and expert intermediaries can be champions of decolonization and not just advocates of cost-minimization. While the contours of opportunity and constraint will be unique for each policy context, it is crucial to attend to both together if we want to realize environmental justice in practice.   vi  Lay Summary Across the world, many nations have enacted policies to protect freshwater ecosystems from pollution, over-extraction, and material transformation. Often the implementation of such policies is delegated to the local level, purportedly to reflect local needs and conditions. But does local implementation always secure the intended outcomes of policy? This research follows the implementation of a water policy in Aotearoa New Zealand to see how different actors – the state, local government, and non-state experts – attempt to inject their interests into a flexible policy implementation process. Drawing on 76 interviews with water policy professionals, I show how implementation flexibility is actively struggled over by public and private actors to secure very different environmental, economic, and cultural objectives. By making visible these political struggles, in-depth social research can illuminate how local policy implementation does, does not, or might yet realize the aspirations of environmental justice and sustainability.    vii  Preface This research project was designed and undertaken by the author, Marc Tadaki. The resulting dissertation is an original intellectual work by the author. The fieldwork for this project received approval from the UBC Behavioural Research Ethics Board, under the project title: ‘Structuring the field of environmental governance: tracing the 'collaborative turn' in New Zealand's freshwater management’ (certificate number: H14-01509).   viii  Table of Contents Abstract…………………………………………………………………………………………………………………………iv Lay Summary…………………………………………………………………………………………………………………vi Preface………………………………………………………………………………………………………………………….vii Table of Contents………………………………………………………………………………………………………….viii List of Tables…………………………………………………………………………………………………………….…….x List of Figures………………………………………………………………………………………………………….…….xii List of Abbreviations……………………………………………………………………………………………………..xiii Acknowledgements..………………………………………………………………………………………..…………….xv Chapter 1 – Introduction…………………………………………………………………………………………………..1 Chapter 2 – From public concerns to state policies: a history of New Zealand freshwater policy……………………………………………………………………………………………………………………………38  Chapter 3 – State action for water policy implementation: coherent rollout of an ideological vision? ………………………………………………….……………………………………………………………………..97 Chapter 4 - Spatial differences in water policy implementation: evils or opportunities?.......147 Chapter 5 - Shaping policy implementation from outside the state? Spatial strategies of translocal intermediaries……………………………………………………………………………………………...199 Chapter 6 - Conclusion…………………………………………………………………………………………………250 Post-script: New Zealand freshwater policy post-2017…………………………………………………….279 References…………………………………………………………………………………………………………………..285 ix  Appendices…………………………………………………………………………………………………………………..315  x  List of Tables  Table 1. Trends for water quality variables in NIWA’s National River Water Quality Network 1989-2013 (77 sites). Source: MfE and Stats NZ (2015)………………………………………………..…..49 Table 2. A sample of commentaries, documentaries, and in-depth investigative articles on New Zealand’s freshwater crisis (from 2016-2018)..…………………………………………………………57 Table 3. A sample of media articles about freshwater (2016-2018), organized into five themes………………………………………………………………………………………………………………………….59 Table 4. Part II, Section 5 of the Resource Management Act 1991, as of 18 October 2017. Available at Table 5. Comparing the high-level objectives of a Regional Policy Statement with the detailed rules of a regional plan. Source: Environment Canterbury (2017, 2018).……………………………65 Table 6. Compulsory and non-compulsory national values in NPSFM 2014 (New Zealand Government, 2014)………………………………………………………………………………………………………..83 Table 7. A summary of three versions of the National Policy Statement for Freshwater Management…………………………………………………………………………………………………………………89 Table 8. A snapshot of key projects in the economics capability-building workstream…….…111 Table 9. Water Directorate team descriptions and some Illustrative projects, as of February 2016……………………………………………………………………………………………………………………………120 Table 10. Five regional councils’ responses to MfE’s questionnaire regarding Policy CA1: ‘The council has identified freshwater management units (FMU) that include all freshwater bodies within its region.’ Source: personal copy of MfE questionnaire………………………………………..158 xi  Table 11. Five regional councils’ responses to MfE’s questionnaire regarding Policy CA2(a): ‘In each FMU, the council has considered how all the national values, listed in Appendix 1, apply to local and regional circumstances.’ Reproduced with permission from MfE. Source: personal copy of MfE questionnaire…………………………………………………………………………………………….160 Table 12. Comparative evaluations of regional councils by thematic criteria (from MfE, 2017e). Only the six themes listed here contained comparative evaluations in the MfE report. I have bolded the regions that are examined in this chapter………………………………………………162 Table 13. Area, land use, and demographic variables for New Zealand’s regions…………..…..166 Table 14. Author’s synthesis of conclusions from MfE’s assessment of five regional councils’ NPSFM implementation programs (drawn from MfE, 2017c, 2017d, 2017f, 2017g, 2017h)…186 Table 15. A selection of headlines on stories about Overseer® published since 2013………...216    xii  List of Figures Figure 1. Infographic of statistics from Our Fresh Water 2017. Reproduced with permission. Source: MfE (2017j)…………………………………………………………………………………………………...….50 Figure 2. Dairy cattle by region 1994, 2002, 2015. Reproduced under a Creative Commons Attribution 4.0 International Licence. Source: Statistics New Zealand (2017a)……………………54 Figure 3. Map of New Zealand’s 16 regional council boundaries. Reproduced with permission. Source: Figure 4. Planning with community diagram. Reproduced with permission. Source: MfE (2013a)…………………………………………………………………………………………………………………………82 Figure 5. Snapshot of part of NPSFM implementation review questionnaire issued to regional councils. Reproduced with permission from MfE. Source: personal copy….………………………158 Figure 6. New Zealand’s 16 regional councils, with five study regions shaded. Map credit: Kati Doehring……………………………………………………………………………………………………………………..164 Figure 7. The five Whaitua for the Greater Wellington area. Reproduced under a Creative Commons Attribution 4.0 International Licence. Source: Greater Wellington Regional Council     xiii  List of Abbreviations  AC   Auckland Council ECan   Environment Canterbury, also known as Canterbury Regional Council GWRC   Greater Wellington Regional Council HBRC   Hawkes Bay Regional Council Horizons Horizons Regional Council, also known as Manawatū-Wanganui Regional Council Iwi/hapū Iwi are Māori tribal groups who share a common ancestor. Hapū are collections of families which make up an iwi. National Objectives  Introduced in the NPSFM 2014, the Framework requires regional councils to Framework set quantitative limits for nine biophysical attributes for waterbodies at or above the national bottom line.  NPSFM  National Policy Statement for Freshwater Management - gazetted in 2011, revised substantially in 2014, and revised with minor amendments in 2017. NPSFM 2011 National Policy Statement – Freshwater Management 2011. Requires councils to set water quality and quantity limits for all waterbodies, based on local and national values. NPSFM 2014 National Policy Statement for Freshwater Management 2014. Revised version of the NPSFM added the National Objectives Framework and Te Mana o te Wai.  xiv  NPSFM 2017  National Policy Statement for Freshwater Management 2014 (Amended 2017). These 2017 amendments required regional councils to measure, report upon swimmability of waterbodies, and create plans to improve swimmability over time. It also required reporting on a new ecological index. NRC   Northland Regional Council RMA  Resource Management Act 1991 - New Zealand’s principal environmental statute. Tangata whenua  ‘People of the land’. Usually refers to the Māori of a geographical territory who identify with that place. Te Mana o te Wai ‘The mana of the water’. Te Mana o te Wai is a statement included in the NPSFM 2014 emphasizing that water is important to all New Zealanders. In the amended 2017 NPSFM, Te Mana o te Wai emphasizes that the purpose of the policy is to ensure the integrated and holistic well-being of freshwater bodies, which includes the values of tangata whenua. WRC   Waikato Regional Council    xv  Acknowledgements  It takes a village to realize a PhD, and this dissertation has been supported by the generosity and kindness of many people. Thanks first and foremost to the participants of this research, for sharing with me your time, your experiences, and your wisdom.  Jim Sinner has been the best mentor an emerging environmental researcher could have. Since enrolling me into the Freshwater Values, Monitoring and Outcomes project (2010-2016), Jim has challenged me to think hard about the practical problems of governing, while being profoundly open-minded and encouraging me to pursue critical lines of inquiry.  Suzie Greenhalgh and the Freshwater Values, Monitoring and Outcomes research team took a leap a faith with me, and they have been supportive and nurturing mentors. In addition to Suzie, I’d like to thank Andrew Fenemor, Garth Harmsworth, Margaret Kilvington, Shaun Awatere, and Will Allen for helping me learn the ropes over the years. The University of Auckland has been my second academic home during fieldwork visits. Gary Brierley, Jenny Salmond, Nick Lewis, Brendon Blue, Karen Fisher, and Richard Le Heron have continued to be invaluable mentors, friends and colleagues. And new collaborative projects with Anne Salmond, Charlotte Šunde, Dan Hikuroa, and Ian Fuller have been enriching and rewarding experiences. In Vancouver I have benefitted from the camaraderie and inspiration of many people. The Critical Geographies of Policy reading group (2014-18) organized by Eugene McCann has been a valuable space of intellectual inspiration, and CHANSLAB at UBC have kept me grounded in the ecological and the practical. At UBC Geography, fellow grad students have made the PhD experience fun and nourishing: Joe, Colin, Emily H, James, Paige, Leigh, Rachel Bok, Julian, Fernanda, and others have been encouraging and supportive friends. A fateful conversation at the 2015 Geography Christmas party with Leonora King led to several ambitious attempts to xvi  ‘bridge the divide’ in geography, with co-conspirators Lucy MacKenzie, Dave Reid, and others. One of these was a geoscience project that also included Sara Cannon, Kiely McFarlane, and Michele Koppes, which, among many things, demonstrated to us the value of having and being caring colleagues. Olav Slaymaker has been a generous mentor in the department, and I will forever cherish our disciplinary musings. Beyond Vancouver, I am grateful for Rebecca Lave and Noel Castree for involving me in international disciplinary initiatives along the way. Trevor Barnes and Kai Chan have been thoughtful, caring, and committed supervisors. Trevor’s close readings of my thesis from initial mess to final draft helped teach me how to write with economy, clarity, and coherence, while letting me figure out what I want to say. Kai’s strategic advice has helped me set my ideas within broader frames of reference, especially within the logics of environmental management. My committee members Eugene McCann and Jamie Peck have been diligent advisors, encouraging me to think harder about the conceptual and methodological frontiers I want to engage.  Thanks to my family in Aotearoa New Zealand and Hawai’i. Mom, Chris, Dad, Kat, and Anthony: thank you for encouraging me, for feeding and housing me occasionally, and bearing with me through it all. To Gram and Papa, the McDs, Sandi, Gary, Lynn, Steve, Patti, Kayla and Grace – thank you for your love and support. Boo (2001-2017): I wish you could have made it till the end. You wouldn’t have known what a PhD is, but I’m sure you would have sat on it. To the animals of British Columbia, thank you for reminding me how and why the more-than-human world needs our attention.  Finally, Kiely McFarlane has provided intellectual and emotional support at every step of the way, without a break, and while doing her own PhD as well. I couldn’t have made it without you. 1  Chapter 1 – Introduction Prologue May, 2011. The New Zealand government gazettes the National Policy Statement on Freshwater Management. The policy requires New Zealand’s 16 regional councils to set water quality and quantity limits to protect waterbodies from pollution and over-extraction. The limit-setting process must reflect local and national values. September, 2012. I sit in a workshop in the North Island city of Napier, on the east coast. We are in a conference room of Hawkes Bay Regional Council. I am part of a research partnership with regional council staff, tasked with providing analysis and guidance to support the improvement of water management. We have just had a presentation by a council planner from Canterbury Regional Council (also known as ECan), and the room is abuzz with excitement. ECan, despite having their elected regional council sacked by central government, have proceeded with an ambitious approach to implementing the National Policy Statement. Ten ‘Zone Committees’ composed of selected community members are charged with setting objectives for waterways, and recommending regulations to council. ECan is the first region to attempt such an ambitious approach to community engagement, and staff from other councils are very interested in how it all played out. By the end of the day, we were no longer talking about ECan and its experiences; we were talking about ‘the ECan model’. Despite the fact that the National Policy Statement does not require councils to engage communities in the way that ECan has, the ECan model became a de facto benchmark for implementing the policy.  February, 2016. As a member of the Ministry for the Environment’s water policy Implementation team, I attend a Ministry-held workshop about prospective revisions to the National Policy Statement. The purpose is to discuss how ecological ‘limits’ for freshwater bodies should be defined. The opening presenter flashes a slide on screen. It read ‘[Um29]’ in 2  the style of a chemical in the periodic table, with the subtitle, ‘the element of confusion’. He introduces the motivation for the workshop: New Zealand’s 16 regional councils were interpreting freshwater ‘limits’ differently, and this posed a problem because not all interpretations of ‘limits’ ensure a basic level of ecosystem functionality.  1.1 Interpreting environmental policy Using public policies to protect ecosystems is not a simple task. Bound by the constraints of language, policies are open to multiple types of interpretation (Yanow, 1996; Freeman, 2012; Hupe et al., 2014). As the vignettes above indicate, sometimes a new policy term or concept can be interpreted in ways that exceed or subvert the intention of the policy itself. In the case of defining limits, the observed variation in implementation styles was decidedly a bad thing, according to the government officials. In the workshop discussing the ‘ECan model’ of community engagement, I was fascinated by how excited other council staff were about the idea of collaboration, and how willing they were to promote an explicitly collaborative (as opposed to simply consultative) interpretation of community involvement in their own regions.  It is easy to see how interpretive flexibility in policy implementation can be seen as a bad thing. If environmental policies made in the public interest are interpreted flexibly at the moment of implementation, the policy may not achieve what is intended. For example, if permits for polluting freshwater are ‘flexible’ and if each new permit application presents a valid argument for use, this flexibility may unwittingly lead to the issuance of more pollution permits than a freshwater ecosystem can sustain. Many political scientists and policy scholars see the creation of government policy as the pre-eminent method of stating and realizing the public interest, as it is made by elected representatives who are accountable to the voting polity (see e.g. Pressman and Wildavsky, 1984; Hill and Hupe, 2014). In this view, interpretive flexibility can undermine the achievement of the public interest as contained 3  within the policy itself, especially when local interpretations are shaped by unelected actors such as experts or industry lobbyists. Against a rigid view of policy interpretation, however, many scholars of environmental governance and policy studies have argued that interpretive flexibility can actually make policy more legitimate and effective (e.g. Bulkeley and Mol, 2003; see also Hill and Hupe, 2014). Where a national environmental policy might be perceived as arbitrary, ideological, and unrelated to the interests and livelihoods of local communities (e.g. Scott, 1998), having the ability to interpret and tailor implementation programs to local settings can strengthen the local acceptance and abidance of the policy. It is argued that interpretive flexibility can allow for the incorporation of local knowledge at the point of implementation, better local buy-in, lower costs, and better environmental outcomes (e.g. Margerum, 2011; Holley et al., 2012). Classic studies of policy implementation have even argued that because policies are written at such a high-level, interpretation and judgement are required for a policy to realize its own objectives (Lipsky, [1980] 2010). The debate about the appropriate level of interpretive flexibility in policy is ongoing. Should local governments and implementing actors be allowed to tailor their implementation approaches based on local conditions, livelihoods, and interests? If standardization is required, what form should it take, and what are the costs? Will local flexibility always lead to economic and political elites co-opting the process to secure their interests? My suspicion is that, like many things, the answer to such questions is ‘it depends’. Whether interpretive flexibility is good or bad depends upon the specifics of policy and its intent, it depends upon the specific implementation pathways in question, and it depends upon how that implementation pathway relates to the place in which it is being undertaken.  When the National Policy Statement for Freshwater Management was gazetted in 2011, I was intrigued by both its aspiration and its vagueness. On the one hand, the policy was radical in that it required all councils to create and enforce quantitative limits to protect freshwater 4  ecosystems. This could be a world-leading approach to environmental management. On the other hand, the policy required regional councils to set these limits in ways that they could determine themselves, informed by local values, environments, and needs. I was interested to know: would this interpretive flexibility lead to co-option by local elites and continuing environmental degradation? How would the state respond to this interpretive flexibility to ensure that its favored interpretations were privileged? And finally, how would non-state actors – such as experts and industry lobbyists – shape the way that local governments interpret the policy? In response to the interpretive flexibility of the National Policy Statement, I wondered whether and how interpretative flexibility might be mobilized toward progressive objectives. For example, can environmental policies be interpreted in ways that strengthen ecological protection, enhance local democracy, and empower indigenous groups? This is not the same as asking how public policies can be designed to secure progressive outcomes, which assumes that current politicians in government want these outcomes. My questions are not about policy design, but of implementation. As the government officials who ran the ‘limits’ workshop in the prologue would surely attest, interpretations can mean the difference between ecosystem sustainability and collapse, or between symbolic and meaningful community involvement in the decision making process. For me, I see the National Policy Statement not as a simple condensation of political-economic interests awaiting mechanical translation into reality, but rather as a text whose meanings and effects are being struggled over in real time. There may be local victories for certain interpretations, there may be state action to control interpretations, and there may be non-state experts who attempt to shape which interpretations are seen as valid or sensible.  This dissertation examines the implementation of freshwater policy in Aotearoa New Zealand, with the aim of understanding how different political actors are attempting to shape the terrain of policy interpretation. Creating and sustaining policy interpretations are tasks that 5  require ongoing work (Mosse, 2005), the outcomes of which are determined by struggle. Recognizing the consequences and possibilities of policy interpretation can open up a wider field of vision in which environmentalists and others can strategize to affect the processes and outcomes of environmental policy implementation. The struggle for healthy ecosystems, indigenous rights, and local democracy does not – and should not – be considered to end with claiming political government and the writing of policy. By looking at how a recent freshwater policy in New Zealand has been interpreted by state officials, by regional councils, and by a cohort of expert non-state actors, I will show how these actors create their own interpretations of policy and how they seek to influence the interpretations of others. Understanding how policy interpretation happens is a first step toward reclaiming these processes for better ends. This introductory chapter outlines my theoretical approach, my empirical case study, my methodology, and the argument I will develop through this dissertation. Section 1.2 outlines my principal theoretical framework and associated analytical priorities. Section 1.3 introduces the New Zealand policy context, situating recent freshwater policy reforms historically and then against my research questions. Section 1.4 outlines my research design and methods of data collection. Section 1.5 outlines the structure of the dissertation and the arguments of the individual chapters. 1.2 Opening up the politics of environmental decentralization  The scholarly fields of environmental governance and interpretive policy analysis provide valuable conceptual resources for designing a study of environmental policy implementation. In this section I outline in broad terms what these approaches contribute to addressing the questions I posed in the introduction, and I identify some needs for the future development of these fields. 6  1.2.1 Environmental governance: a bleak view of institutional transformation? Broadly speaking, the field of environmental governance is concerned with understanding how environmentally impactful human actions are organized, and how they might be regulated or changed (Lemos and Agrawal, 2006; Evans, 2012; Benson and Jordan, 2017). The concept of environmental governance can be distinguished from a narrower concern with environmental policy, where environmental governance addresses the broad relationships between state, private, and civil society actors (Evans, 2012). While the study of environmental policy design and implementation are urgent research priorities, it is important to situate environmental policy as one mechanism within wider relations of legitimacy, authority, and power that enable policy to be effective at changing the world (or not). Environmental governance scholarship in this vein has drawn attention to several important features of the changing contexts of environmental policy making and implementation.  First, across the world, environmental policy making and implementation are increasingly decentralized. Instead of environmental policies being written and implemented at the national scale, policies are handed down to ‘smaller’ scales of government for implementation (Lemos and Agrawal, 2006; Larson and Soto, 2008; Berkes, 2010; Cohen and McCarthy, 2015). In federalist systems such as the U.S. (Doyle, 2012), Canada (Dunn et al., 2014), Australia (Althaus and Morrison, 2015), and Brazil (Libanio, 2018), for example, this means that policies passed by federal government are required to be implemented by state/provincial governments. Decentralization also applies to other sub-national units such as municipalities, regions, or catchments. Decentralized arrangements can grant new powers to local implementing authorities, but also loads them with new responsibilities; this often occurs without providing any new funding or in-kind resources. How local authorities enact their new responsibilities, and what types of outcomes they achieve, constitute important empirical research questions.  7  Second, the use of prescriptive regulation as a mechanism of environmental policy is in general declining in favor of ‘soft’ or voluntary approaches (Cashmore et al., 2015; Mol, 2016; Benson and Jordan, 2017; Perkins, 2017b). Such voluntary approaches involve the production of guidance, industry self-governance, and the provision of incentives for undertaking specific environmental practices, but without the compulsion to adopt those practices. Characterizing the state’s use of regulatory and voluntary mechanisms in environmental policy constitutes a second empirical theme of environmental governance research. Third, as environmental policy is ‘scaled down’ to local authorities, it is also often ‘scaled out’ to involve civil society and private actors (Cohen and McCarthy, 2015). This occurs as new forums and dialogues are created to inform the creation of policy, as well as forums to determine the shape of its implementation (see e.g. Margerum, 2011; Edelenbos et al., 2013). The selective involvement of civil society into the policy making and implementation process constitutes a third empirical theme. How have critical environmental scholars made sense of these changes in the structure of environmental governance? From a political-economic perspective, scholars have explained these changes as related through the ideological project of neoliberalism which seeks to strengthen the power of economic elites. Here decentralization is often understood as the state ‘rolling back’ its responsibilities for environmental protection; the use of voluntary regulation is taken as evidence of a weakened state captured by economic interests; and scaling out mechanisms are interpreted as methods for legitimating and empowering existing economic interests within a democratic framework (see e.g. McCarthy and Prudham, 2004; Himley, 2008; Castree, 2010; Perkins, 2017a). New forums for stakeholder participation in environmental decision making, for example, have been criticized as being malevolently steered by states and co-opted by local elites to pursue narrow economic agendas (Oleson, 2012; Cohen and Bakker, 2014; Brisbois and de Loë, 2016). Swyngedouw (2010) contends that 8  such forums are merely efforts to legitimate existing (and inequitable) social and economic arrangements, and that a radical politics through such forums is impossible. The political story that emerges from these accounts is often quite bleak. Decentralization places burdens on local governments who are either under-resourced to protect the public interest, or who are acting on the needs of economic elites. Novel participatory forums are evacuated of political possibility, not by the process of historical struggle but by design. While these critiques of governance changes are generally rigorous and very important, I feel there is also a need to see new forums as open for contest, struggle, and perhaps, victories (however small). If we only describe the negative outcomes of new governance arrangements, and if these outcomes are always predetermined by powerful political-economic forces (which exist in every society), we cede the battle for policy interpretation before it has even begun. There is a need and opportunity, I suggest, to theorize the changing politics of environmental governance in a way that can identify significant powerful forces shaping governance outcomes, while also treating the outcomes of policy implementation as at least somewhat ‘open’ and amenable to progressive influence through struggle. For this task, I turn to the field of interpretive policy analysis. 1.2.2 Interpretive policy analysis – conceptualizing struggle in environmental governance The field of interpretive policy analysis is concerned with understanding how policy meanings are constructed and contested, and how these meanings change over time through struggle (Fischer and Forester, 1993; Hajer and Wagenaar, 2003a; Fischer and Gottweis, 2012). An environmental policy, for example, can be examined to identify which and whose meanings and languages are contained within it, whose are not, and what values and nature-society ideologies are given precedent (Hajer, 1995; Hajer and Versteeg, 2005; Dryzek, 2013). Interpretive policy analysis pays attention to how and why communities of meaning-making interpret policies in different ways, where different interpretive styles can bear different 9  material implications (see e.g. Yanow, 1996; Hajer, 2003a). Rather than treating environmental policies as a simple distillation of societal interests that are destined to realize specific effects, each actor surrounding a policy is an agent capable of interpreting policy in ways that might differ from others (see e.g. Freeman, 2012).  Governance theorist Maarten Hajer’s work has been unique in developing an interpretive approach to understand institutional restructuring. Hajer and Wagenaar (2003b) observe that as the state has become more permeable and as policy making processes become more open to the involvement of private actors, policy making is increasingly being undertaken within conditions of institutional novelty, where past relationships of authority no longer hold. Hajer (2003b) argues that novel environmental governance arrangements are often malleable in their purpose and structure, and that environmental actors have the opportunity to shape the norms around their interpretation and use. A new participatory planning forum, for example, may not have established objectives, rules, or responsibilities, yet when such rules are decided they can significantly affect environmental and democratic outcomes. To reflect this sense of novelty and the not-yet-determined nature of new governing arrangements, Hajer coins the concept of an ‘institutional void’. An institutional void refers to a situation in which ‘there are no generally accepted rules and norms according to which politics is to be conducted and policy measures are to be agreed upon’ (Hajer 2003b: 175). Within an institutional void,  actors not only deliberate to get to favourable solutions for particular problems but while deliberating they also negotiate new institutional rules, develop new norms of appropriate behavior and devise new conceptions of legitimate political intervention. (Hajer 2003b: 175-176, emphasis in original) The ‘institutional void’ concept highlights the conditions of novelty and ambiguity under which governance takes place within a network society (Hajer and Wagenaar, 2003b). The concept describes a condition surrounding policy making that must be accounted for and 10  explained, rather than providing a complete or predictive theory of policy outcomes. For the purposes of this dissertation, Hajer’s notion of an ‘institutional void’ helps to open up the politics of policy interpretation in two ways. First, the notion of an institutional void provides a sense of political opportunity, in which policy outcomes will be shaped by the type of struggle that unfolds. This emphasizes that the outcomes of new environmental policies or governance arrangements are not necessarily foregone conclusions; there is space for politics. Second, Hajer’s concept of the institutional void also draws attention to a distinctive type of political struggle associated with attempts to structure the rules of the game, including ‘norms of appropriate behavior’. That is, actors struggle not only to ‘win’ in a given dispute but to define the norms that govern how disputes are decided into the future. This is a type of politics that is related to – but distinct from – an interest-based view of policy. There is a difference, for example, between arguing for a policy to have stronger environmental protection (interest-based politics), versus arguing for the use of a participatory forum or decision making criterion with which to decide upon the appropriate level of environmental protection. The first type of politics is about the ends of policy (i.e., environmental protection), whereas the second type of politics is about the means to shape how policy is decided. Hajer’s injunction to explain recognize and explain the institutional void thus invites scholars to consider how environmental actors use discussions about the means of environmental policy to shape their ability to secure their desired ends. This insight highlights the role of technical procedures and norms within the decision making process that are argued to be about the means rather than the ends of policy, even though they are in reality about both. The notion of an institutional void recognizes that new policy objects, institutional configurations, and organizational routines – such as those introduced by an environmental policy – are often malleable and that norms governing these entities will be struggled over by situated political actors. In the environmental governance literature, the institutional void has been used to characterize the malleable character of new policy objects such as biosecurity 11  threats (Enticott and Franklin, 2009), climate change policy (Wejs, 2014), or ecosystem services (Jordan and Russel, 2014), in which scientists and other policy actors struggle to establish norms of interpretation to guide environmental and risk management. Scholars have also drawn on Hajer’s institutional void to characterize the plasticity of emerging participatory governance mechanisms in forestry (Howlett et al., 2009; Kleinschmit et al., 2009), water utility privatization (Beveridge, 2012), and marine governance (van Leeuwen et al., 2012). Such studies emphasize the dynamic unfolding of new governance mechanisms and objects, highlighting how rules of engagement are decided ‘on-the-go’ (Beveridge, 2012: 48). To date, the institutional void has largely been used to describe conditions of institutional novelty surrounding a one-off issue, rather than characterizing the interpretive politics of environmental policy in general. The scope of analysis has consequently been limited to the new policy object or governance mechanisms without regard to wider shifts in the distribution of power and authority in the policy regime over time. In this dissertation, I scale up the idea of an institutional void to conceptualize the nature of struggle within a decentralized environmental policy regime that includes multiple waves of policy over nine years (2008-2017) in Aotearoa New Zealand. The institutional void concept is useful for identifying and characterizing the nature of politics and political struggle over these nine years, which includes several policies and whole-of-government reform of freshwater regulation. I consider how the roles of the central state, local government, and non-state intermediaries (experts) are reconfigured through decentralization and how they exercise their new powers to structure the institutional void of policy interpretation for others. In this way, the institutional void refers to more than conditions of novelty, but a regime of political struggle built around the power to interpret policy and to force one’s interpretations upon others.  I propose to study the contours of this regime of struggle, attending to key actors and how they exercise agency within conditions of constraint and uncertainty. This analysis will help 12  to understand the regime of power that emerges as differently-capable actors seek to structure the field of interpretation surrounding multiple new policy objects. This approach advances the study of environmental decentralization by recognizing the emergent and struggled-over character of norms governing new institutional arrangements. It further contributes to the conceptualization of the institutional void, while still prescient, by providing a more precise specification of actors and their interrelationships that can support comparison of scholarship on environmental governance across policy issues and geographical contexts. The institutional void refers to a condition of policy making, and it was not intended to provide a complete explanation of environmental politics in a network society. The metaphor of a void, for instance, implies nothingness, whereas we know intuitively that concepts will always come from somewhere and someone; there will be existing methods and channels for interpreting new policy concepts across various scales. The term ‘void’ is thus intuitively limited by its inattention to history. While recognizing this logical criticism of the void metaphor, I find the concept of an institutional void valuable for highlighting the opportunity for embedded political actors to shape emerging norms of policy interpretation. The notion of yet-to-be-determined norms embedded within the idea of an institutional void provides an important corrective to those accounts of environmental decentralization in which policy outcomes are simply determined by powerful political-economic forces in society. The metaphor of an institutional void may flatten history, but it at least does not reduce the present to the playing out of the past. Thus, in this dissertation I will build the notion of an institutional void into a more robust framework for theorizing interpretive politics. This requires, firstly, recognizing and addressing its historical blindness. As Peck (2011) observes, 13  policy ideas, innovations, technologies, and models… do not float freely in some unstructured universe, to be picked over selectively by a faceless elite of continuously learning policy-makers. Rather, the field of policy transfer is itself socially and institutionally constructed, being populated by a wide array of actors and institutions; it is sharply contoured and striated, in the form of shifting landscapes of conjunctural openings and preferred channels; it is structured by relatively enduring policy paradigms, which establish intersubjective frames of reference and institutionalized centers of authority; and, perhaps above all, it is saturated by power relations. (p791) While the introduction of a new environmental policy might contribute new concepts, arrangements, and practices to the governance landscape, these are not open to infinite interpretation. Rather, new policy ideas flow across an uneven terrain of existing norms, policy meanings, and power hierarchies, where actors are differently positioned to promote and embed their preferred interpretations into dominance. Further this terrain is ‘a landscape crowded with existing institutional forms, many with different (and often avowedly nonecological) functions and purposes’ (Jordan and Russel, 2014: 193). Environmental politics within an institutional void might be more open than not, but is not completely open; struggle does not play out on an even playing field. This dissertation examines how a new freshwater policy in Aotearoa New Zealand flows across an uneven terrain populated by local authorities, state, and non-state policy actors who attempt to interpret the meaning of policy concepts in ways favourable to (or consistent with) their objectives. I will examine how actors are leveraging the new institutional arrangements and concepts of an environmental policy to structure the field of interpretation around them, while transforming environmental democracy in the process. My objective is to develop an ‘open’ account of environmental politics that is attentive to emerging forums and mechanisms of political struggle, with the intention of supporting environmentalists to develop a broader view of the environmental governance system and how environmental 14  values might be woven throughout its infrastructure. I will operationalize the concept of the institutional void to develop insight into how New Zealand’s freshwater governance architecture is changing with the introduction of new policies, concepts, and mechanisms, and what types of politics these changes are enabling. Put simply, I will analyse how different types of actors struggle to influence the outcomes of a new environmental policy. The institutional void provides a metaphor with which to begin this inquiry, and my two analytical priorities for elaborating the institutional void (as an open opportunity and as a mode of struggle) will provide guidance for this endeavour. In the next section, I describe the New Zealand context of freshwater policy and why I think it constitutes a valuable example through which to deploy and elaborate the institutional void concept. I will then discuss my methodology, and how I develop the institutional void concept through the dissertation.  1.3 An experiment in decentralization: freshwater policy reform in Aotearoa New Zealand Recent developments in environmental policy in Aotearoa New Zealand provide an illustrative setting in which to study how new policy terms are being actively interpreted to advance the objectives of different political actors. In 2011, a major national freshwater policy was enacted that requires implementation by local governments. The ambition and nature of this policy – and the wider reform effort of which it is a part - makes this a valuable case to study how decentralization is changing the structure of environmental politics.  Aotearoa New Zealand (henceforth New Zealand)1 is a settler-colonial island nation in the South Pacific that, like Australia, remains part of the British Commonwealth. New Zealand has several characteristics of a small European country, with a temperate climate, a population of 4.7 million people (similar to Ireland, Costa Rica, and Norway), an area larger than United Kingdom but smaller than Italy, and a gross domestic product per capita                                                  1 Aotearoa is the indigenous Māori name for New Zealand, commonly translated as ‘land of the long white cloud’. 15  comparable to South Korea, Japan, Spain, and Italy (IMF, 2017). Unlike many European countries, however, and like settler-colonial nations such as Canada, Australia, and the United States, New Zealand has a recent history of state violence toward indigenous peoples, including murder and the dispossession of indigenous land (King, 2003; Stevens, 2015). 1.3.1 Myths and their contradictions Globally, outside of its rugby team and America’s Cup endeavors, New Zealand is known for its generally progressive social attitudes and green environmental image. The Lord of the Rings films (2001-3) played a significant role in showcasing New Zealand’s natural landscapes to the world, and the government has continued to brand New Zealand (for tourism and primary production) as ‘clean and green’, and ‘100% Pure New Zealand’2. International companies now trade significantly on New Zealand’s brand in the global marketplace, and the maintenance of this brand has become a significant priority for governments and industry actors (Barkham, 2017).  In recent years, New Zealand has also come into the international spotlight for its approach to restorative justice with Māori (New Zealand’s indigenous people). Since 1975, the New Zealand government has used a tribunal to publicly document and evaluate the colonial state’s historical crimes against Māori, and recommend recompense based on the nature of the grievances (see Sullivan 2016). Recently, this has involved recognition that governance powers over land and water resources should be shared. Several Māori iwi (tribes sharing a common ancestor) have been given shared authority over water management based on principles – including indigenous principles – negotiated between Māori and the government. In 2016, The New York Times published a story titled In New Zealand, lands and rivers can be                                                  2 See, for example, this government-created advertisement from 2017: 16  people (legally speaking) (Rousseau, 2016), and in 2017 The Guardian ran an investigative article titled New Zealand river granted same rights as human being (Roy, 2017). Much popular and scholarly attention has since been paid to understanding the potential environmental and legal implications of granting the Whanganui River the status of legal personhood in 2017 (Boyd, 2017; Kothari et al., 2017; Rodgers, 2017; Smith, 2017; Sutherland et al., 2017; Charpleix, 2018; O'Donnell and Talbot-Jones, 2018).  As with most myths, while there is some truth to these stories, there are contradictions, qualifications, and tales not told. In fact, New Zealand’s freshwater estate has been degrading for decades as a result of agricultural land use intensification and urban development (MfE and Stats NZ, 2017), a reality that Māori and environmentalists have been highlighting for a long time (Knight, 2016). It has taken many scientific studies, several State of the Environment reports, many visible ecological crises, and (perhaps most effectively) continuing negative evaluations by the Organization for Economic Cooperation and Development for freshwater to be placed onto the national policy agenda (Knight, 2018). Prior to 2008, successive governments from the left and right consistently failed to meaningfully confront and address freshwater degradation, despite increasing public concern.  New Zealand’s social and economic fabric is also shaped by historical colonialism and structural inequality. As New Zealand opened its trade markets in the 1980s and 1990s, the infamous fourth Labour government (1984-1990) radically reorganized the government bureaucracy and enacted social, economic, and environmental policies based on neoliberal philosophical principles, most of which were happily embraced and embedded by the following fourth National government (1990-1999). It is widely understood that New Zealand’s experience of privatization and the marketization of (previous) state functions was perhaps one of the most intense experiences of neoliberal reform across the Western world (Goldfinch, 2000; Peet, 2012). Jane Kelsey (1995) evocatively referred to these reforms as 17  ‘The New Zealand experiment’, which became a global model for neoliberal programs and strategies across the world (see e.g. Kingfisher 2013). With respect to Māori, while the 1840 Treaty of Waitangi (Te Tiriti o Waitangi) promised Māori full control over their land and resources, successive laws and policies by left- and right-wing governments have not addressed the economic and environmental rights of Māori. Although Treaty settlement legislation has created new local governance relationships between Māori and central and regional governments, these settlements have not acknowledged – as the Waitangi Tribunal has explicitly clarified – that Māori were guaranteed proprietary rights by the Treaty (see Ruru, 2013; Strang, 2014). The Whanganui River settlement, which creates a new decision making body and novel decision making mechanism, may provide unique local opportunities to improve the health of the river and foster an ethic of kaitiakitanga (guardianship) (Hikuroa et al., in review). However, the settlement does not affect property relations within the catchment – either for land or water – and time will tell how much power the new governance entity actually has to shape catchment land use. Thus, although progressive Treaty settlements such as the Whanganui River settlement may create opportunities for governance steering and strategic activism in the future, as yet Māori remain limited in their ability to direct freshwater governance and improve health outside of a few select catchments. New Zealand’s regional authorities still retain most authority and responsibility for the operational management of freshwater in Aotearoa New Zealand. 1.3.2 A decentralized freshwater policy In response to increasing public and economic concern about New Zealand’s environmental reputation, in 2011 the centre-right fifth National government (2008-2017) enacted the National Policy Statement on Freshwater Management (New Zealand Government, 2011). This policy requires regional authorities to decide upon ecological limits for water pollution and extraction, and to design and implement plans to achieve these limits. Instead of setting 18  national ecological limits to water and land use, the government took a decentralized approach, arguing that local environmental differences and community values needed to inform regionally-tailored implementation. Since the policy was enacted, regional authorities have used a range of approaches to elicit the preferences and values of their communities and operationalize them through planning mechanisms (see MfE, 2017e). Despite central government officially devolving responsibility to regional authorities, central government has actively steered the policy implementation process through a comprehensive program of implementation support, coupled with targeted revisions to the policy in 2014 and 2017. Non-state actors such as agricultural industry lobbyists and environmental organizations have also engaged with the implementation process, employing a range of strategies to champion preferred policy ideas, affect central government priorities, and determine regional environmental planning rules. In these ways, the New Zealand experience offers a uniquely valuable case of a policy being enacted within an institutional void, where norms are not yet clear and hence become the object of political struggle in an uneven terrain.  1.4 Methodology The institutional void refers to a set of conditions under which environmental policy and governance are practiced, and indicates two priorities that can guide analysis of environmental politics (institutional void as an open opportunity, shaping norms as a distinctive mode of struggle). To operationalize these priorities into a conceptual framework with empirical applicability, some structure is needed. The selection of key entities (e.g. political actors, policies) can help to focus and bound inquiry, as will the delimitation of the spatial scale at which emerging norms of interpretation might apply. I have chosen to examine how a specific policy – the 2011 National Policy Statement on Freshwater Management – has enacted new concepts that local authorities across New Zealand must interpret and implement. Since this national policy applies to all regions, all local authorities within New Zealand are potentially subject to the norms that emerge in relation to this policy.  19  My research design focuses on a specific type of actor within the environmental policy system, what Larner and Laurie (2010) call ‘middling technocrats.’ These are actors with significant but constrained (i.e., middling) levels of autonomy, who rely upon and cultivate a sense of expert identity (i.e. technocrat). I am interested in how these actors exercise their autonomy and leverage their expertise to argue for specific interpretations of environmental policy concepts and objectives. The landscape of environmental governance is populated with many middling technocrats; these actors attempt to achieve political objectives not (or at least not only) through claiming to represent some interest in society, but by having technical knowledge that precedes or transcends conventional democratic politics (Prince, 2016). For example, in deciding how to structure a collaborative process with local stakeholders to decide freshwater limits, middling technocrats within a council might make consequential decisions about the criteria for inclusion of participants and the sequencing of deliberations. Here, middling technocrats draw on expert knowledge (e.g. social science, organizational experience, or legal argument) to justify a particular course of action regarding the means of environmental decision making. My interest is in how middling technocrats across New Zealand’s resource management system craft and promote interpretations of environmental policy to secure their objectives.  My approach can be characterized as a multi-sited interpretive analysis of policy implementation. In her book How does a policy mean?, Yanow (1996: 19) argues that the policy process is at its core ‘a struggle for the determination of meanings.’ To study this struggle empirically, one must identify key ‘symbol-sharing communities’ (p20) who interpret policy concepts through a shared lens. I have chosen to focus my analysis on the decisions and actions of three sets of actors: the national state bureaucracy, local governments, and a subset of non-state actors that I call intermediaries. The central state enacted the policy, has vested interests in its implementation in a specific way, and it also has diverse means at its disposal to promote its favored interpretations. Local governments are charged with implementing the policy, and thus must translate policy concepts into the logics 20  of local planning and operational rules. Intermediaries such as researchers, consultants, and lobbyists cultivate and leverage their policy expertise to generate and promote favored interpretations of national policy for implementation by local governments, making them a significant type of actor in the struggle to shape the emerging norms of policy interpretation. Not all members of civil society possess such expertise and can engage in policy debates and court battles over the proper (or most legally defensible) interpretation of policy. I am interested in how intermediaries (expert actors outside the state) engage in debate about the proper interpretation of policy. By tracing the ways these three sets of actors attempt to shape interpretations of freshwater policy concepts, I ‘follow the policy’ (Peck and Theodore, 2012) across its distal sites of meaning-making and reinterpretation. I focus on how these actors understand the policy and how they then attempt to shape its meaning as perceived by others.  1.4.1 Methods and data collection  This dissertation examines three sets of actors as they attempt to shape the meaning of policy concepts through the implementation process. To understand the activities of the state, I spent time within the Ministry for the Environment’s water policy implementation team and interviewed state bureaucrats. To understand local government interpretations of the policy, I selected five regional councils and interviewed key bureaucrats responsible for implementing the policy. To understand the practices and strategies of intermediaries, I sought out veteran resource management professionals who have experience making and evaluating claims for the interpretation of policy in Environment Court and resource management hearings.  My primary data are drawn from 76 semi-structured interviews with 63 embedded policy actors. The interviews lasted between 20 and 90 minutes, averaging around 50 minutes in length. Interviews were triangulated with historical and contemporary policy documents, with other interviews, and participant observation. Interviews provide a way to allow participants – the policy actors – to draw attention to key relations and reasons driving their 21  policy activities. Interviews also provide a way to understand, in retrospect, key causes of outcomes.  To provide an overview of political positions surrounding the National Policy Statement for Freshwater Management, I interviewed three then-opposition members of parliament – David Parker (then-Labour Party spokesperson for water, former and now current Minister for the Environment), Eugenie Sage (then-Green Party spokesperson for resource management, current Minister for Conservation) and Catherine Delahunty (then-Green Party spokesperson for water, now retired). While I approached the office of then-Minister for the Environment Nick Smith for an interview, I was referred to speak to an MfE official instead. To hear the Minister speak about the National Policy Statement I attended two public meetings – in Nelson in 2016 and Wellington in 2017 – at which Smith presented his proposed policy amendments and fielded questions from the public. From these interviews and public meetings, I sought to construct a broad diagnosis of the freshwater problem in New Zealand, how these actors thought the National Policy Statement addressed this problem (or not), and what they thought was needed now and into the future. Interviews and participant observation - state bureaucracy Through investigating the implementation activities of the state’s environmental bureaucracy, my aim is to identify how state bureaucrats interpret the government policy, how bureaucrats attempt to promote these preferred interpretations, and whether these activities can be simply explained by hierarchical translation of the political ideology of the Minister.  To access the state bureaucracy, New Zealand’s Ministry for the Environment (MfE), I leveraged contacts from an earlier research project to contact MfE water policy officials. We negotiated a five-week (25 working days) stay for me within MfE’s water policy Implementation team over February-March 2016. During this period, I produced a desktop 22  literature review on the future of collaborative governance for freshwater. I was inducted through a formal process and introduced to the entire Water Directorate (the department within MfE conducting water policy work). I participated in weekly team meetings, Water Directorate meetings, and MfE-wide meetings. The eight members of the implementation team signed a consent form recognizing my presence and my intention to take anonymized notes on my experiences (see Appendix 1). With MfE management I negotiated a confidentiality agreement giving them the right of review and veto power over confidential material. They have not redacted any material from this dissertation.  I spent another ten days with the implementation team in August 2016, and five days in March 2017. Within these three periods of time in MfE, I solicited interviews with officials in the implementation team, the Water Directorate, and MfE leadership. I ultimately conducted 40 interviews with 28 individuals; some officials I interviewed multiple times to consider changes and developments in policy. Interview questions were tailored to draw on the unique experiences of the interviewee. After discussion the individual’s background and history of professional work (including MfE), I would ask them to describe their current projects, how choices around these projects were prescribed, and through what means. We also discussed the rationale and work of different teams within the Water Directorate, the design and effectiveness of the National Policy Statement, and what they saw as key challenges for implementation. The purpose of these interviews was to compose a broader sense of how MfE and its officials imagine their specific projects and wider work programs, and how they imagine these activities are affecting the implementation of the National Policy Statement.  All interviewees signed individual consent forms permitting me to quote them without identification (see Appendix 2). I also accessed project documentation that helped me understand the development of the Water Directorate and previous (or unfollowed) directions in water policy. However, I am unable to quote from this material because it is confidential. 23  Being physically present within the Water Directorate made it easy for me to meet officials, describe my work to them, and flexibly schedule (and reschedule) interviews. Receiving a keycard and being co-located in the building indicated that MfE managers approved of my presence, which likely make Ministry officials comfortable and willing to speak to me. Most officials were willing to share stories, discuss projects, and praise (and gripe about) different aspects of life in the Ministry. Surprisingly, almost everyone I asked agreed to an interview, including several managers and the chief executive of MfE. Interviews - regional council bureaucrats By characterizing the local implementation activities of five regional councils, my aim is to examine what types of interpretations have been used for concepts within the National Policy Statement, and how these interpretations can be significant for environmental and democratic outcomes.  I solicited interviews with regional council staff from five regional councils: Northland, Auckland, Waikato, Wellington, and Hawkes Bay. As will be explained further in Chapter 4, I sought variation across five dimensions: • Pressures on freshwater: from urban development (Auckland) vs. agricultural intensification (Waikato) • Māori settlements with the Crown (government): complete and strong settlement (Waikato) vs. incomplete settlements and capacity issues (Northland and others) • Water-related controversy: Pre-existing freshwater conflict (Waikato, Hawkes Bay) vs. freshwater as a relatively new public priority (Auckland, Wellington) • Dominant land use: Intensive urban (Auckland, Wellington) vs. largely rural (Northland, Hawkes Bay) • Resourcing and institutional capacity: High (Auckland, Wellington) vs low (Northland) vs. medium (Hawkes Bay, Waikato) 24  All five chosen regions are in the North Island, which presents a limitation to my study. As I discuss in Chapter 4, I decided against studying Canterbury – the most populous South Island region – because of its unique and already much-studied governance arrangements (e.g., Holley et al., 2012; Nissen, 2014; Thomas and Bond, 2016; Kirk et al., 2017; Jenkins, 2018). While I considered studying Tasman District Council from the South Island, I decided to follow the advice of colleagues from MfE and several interviewees, who recommended that the five regions I used as case studies were sufficiently representative that they would provide valuable insight. During August and September 2016, I interviewed four staff from each of Northland, Waikato, Hawkes Bay, and Wellington regional councils, and from Auckland Council I spoke to five staff plus one consultant. In March 2017 I updated my conversation with one participant from Auckland and interviewed an additional Auckland consultant. To ensure some consistency across regional councils, I sought at least one scientist and at least one policy manager from each council. As each council has a different organizational structure, staff roles are not always equivalent, but this at least ensures I have talked to the person ‘in charge’ of implementing the National Policy Statement in each council. For Auckland, Hawkes Bay, and Waikato, I previously had met relevant water policy staff through the Regional Council Forums of the Freshwater Values, Monitoring and Outcomes research program (2010-2016), which supported my masters research in 2011-12 at the University of Auckland. For Northland and Wellington, I pursued contacts through colleagues from MfE’s implementation team, who vouched for my professional integrity and the value of my work.  In addition to the limited representativeness of my selection of North Island regions, my approach has two further limitations. The first is the small number of interviewees from each council (~4). It would have been more robust to speak to 6-10 people from each council, or to interview additional people until I saturated my accounts of freshwater policy implementation within a council. The primary constraint for me was logistical: I wanted to 25  prioritize breadth over depth, to canvas diversity across five regions in some detail rather than understand one or two regions in rich depth. This means that my accounts of these regions’ experiences are not representative of all perspectives on the policy process within a council. Another limitation is that I limited my focus to council bureaucrats rather than local stakeholders, elected councilors, Māori, and other actors. To conduct an interest-based analysis of policy implementation with all major stakeholders for five regions would require many researchers over several years. For my single PhD project, I have focused on the role of bureaucrats as key political actors who structure the nature of local implementation programs in significant ways, based on expert knowledge and technical arguments. This question of how decision-processes are structured would benefit from, but does not require, a comprehensive analysis of all stakeholder perspectives.  Altogether I conducted 24 interviews with 23 regional implementation actors, consisting of 21 council bureaucrats plus two consultants. Interviews with council staff were tailored to the expertise of the interviewee, the region in question, and my relationship with the participant. Overall, they followed the same broad structure: after discussing the interviewee’s background and experience with freshwater policy implementation, I asked how their council had chosen to interpret the National Policy Statement in specific terms and why that interpretation was selected. I asked about the detail of their implementation programs, and invited interviewees to indicate significant place-based challenges and opportunities facing the council and their chosen approach to implementing the policy. Through this approach, I invited participants to describe and explain the unique features of each region’s implementation approach, allowing me to see how much variation exists in implementation. It also allowed me to see how local government actors can exercise agency in the construction of implementation alternatives. 26 Interviews - water policy intermediaries The decentralized implementation of a new environmental policy generates new opportunities for non-state actors to inject their values into the policy process by crafting, circulating, and promoting favored interpretations of concepts and objectives. I explore how a specific type of expert non-state actor – intermediaries – can affect the interpretations and implementation of policy. Intermediaries are involved in translating interests – for instance of an organization, or themselves as individuals, or the environmental public good – into the languages and rationalities of New Zealand’s resource management system. An intermediary could be a lobbyist or lawyer on-staff within an agricultural organization. They could be a public-interest lawyer, a university scientist, a management consultant, an iwi (Māori tribe) resource manager, or a social scientist. These actors are intermediaries because they are attempting to translate – or intermediate – between some interest (whether public or private, individual, organizational, or collective) and public policy (Moss et al., 2009). They might work for a formal stakeholder organization, such as an industry or environmental group, or they might justify their interventions based on their expertise on technical aspects of policy effectiveness or legitimacy (e.g. as social or natural scientists). Put simply, intermediaries are non-state actors who make formal arguments about how the National Policy Statement should be interpreted. While environmental and industry organizations may argue for policy change in the public arena, they can also act as intermediaries by constructing arguments about the appropriate implementation of an existing policy.  I sought a sample of intermediaries to gain a sense of how intermediaries were engaging with these processes to shape implementation outcomes. My first strategy was snowball sampling. Drawing on my existing networks with research colleagues and regional and central government officials, I sought out seasoned veterans of national and regional freshwater politics who could explain how the resource management system worked in local settings, and especially how local policy implementation activities were influenced – both intentionally and unintentionally – by non-state experts. Since environmental organizations were ‘closer’ 27  to my relational networks of environmental research, I made an additional effort to interview industry organizations. I requested interviews from three of the most prominent agricultural organizations in freshwater policy advocacy: Dairy NZ, Federated Farmers, and Fonterra. I secured an interview with a Federated Farmers spokesperson, the Dairy NZ spokesperson was called out to a court hearing on the day of our interview, and Fonterra did not reply to my communications. While interviews with more industrial and other types of intermediaries would have benefitted my study, my sample is sufficient for an exploratory – rather than representative – analysis of intermediaries. Since my aim is to identify some strategies of intermediaries and consider their implications for environmental politics, it is not my purpose to claim representative coverage of all intermediaries. I am concerned with identifying and analyzing some of the significant logics and tactics that intermediaries deploy to alter the terrain of policy interpretation.  I conducted 13 interviews with 11 intermediaries, though this number is somewhat arbitrary: • Consultants (3) • Policy researchers (3)  • Environmental organizations – Fish and Game (1), Forest and Bird (1), Environmental Defence Society (1) • Land and Water Forum secretariat (1) • Industry organization – Federated Farmers New Zealand (1) This number includes the Auckland-based consultants mentioned previously, and it excludes the many central and local government officials I spoke to who have worked as intermediaries previously (or who currently do now). Because intermediaries come from a range of backgrounds and occupy different roles, interview topics were tailored to individuals. For some interviewees, questions addressed the historical development of New Zealand’s resource management system and how that system failed to protect freshwater 28  quality nationally. For other interviewees, questions addressed contemporary projects and how those projects connected to regional implementation programs. Document analysis  Interviews have been triangulated and contextualized through close reading of and reference to historical and contemporary policy documents, as well as a critical analysis of the academic and practitioner literature on New Zealand freshwater policy. For historical context I focused closely on debates about the design and implementation of New Zealand’s Resource Management Act 1991, especially in relation to freshwater and diffuse pollution. To characterize more recent developments, I analyzed contemporary policy documents produced by successive government ministries on freshwater policy, examined State of the Environment reports and scientific studies of freshwater degradation, and the plethora of reports advising the government on freshwater reform since 2008. Since the start of my PhD in 2013, I have also collected online news articles relating to freshwater in New Zealand, and this material informs my analysis of freshwater politics in public and political discourse. Many of these documents are cited throughout this dissertation. 1.4.2 Reflections on limitations My research focuses on the work of bureaucrats in central and local government, and professionals from the environmental, consultancy, industry, and research sectors. My motivation was to identify and analyze the politics of the interpretive work that these policy actors do. As with any research project, I could not study everything, and while I have laid out my reasons for my choices above, it is worth reflecting on the limitations to my research in relation my study design, my sampling, my positionality, and, most broadly, my Western worldview. 29 Study design My research design focuses on the middling technocrats who craft policy interpretations from positions within central and local government as well as from organizations outside the state. My study was not designed to systematically evaluate the five local implementation programs I examined. Such a study would likely require a full PhD focusing on one or two regions to develop saturated accounts of local policy implementation processes, including interviews with more council staff as well Māori and stakeholders participating in those processes. Since I wanted to compare the implementation across more than two regions, in addition to investigating the interpretations promoted by the state bureaucracy and a sample of intermediaries, I sacrificed depth for breadth at the local scale, speaking to at least four officials from each regional council. It is possible that because of this sampling my accounts of the motivations and logics of local policy approaches are skewed in favor of the participants I spoke to. However, through a processes of triangulation with policy documents as well as with MfE officials and other intermediaries, I have confidence that my local accounts as robust. Furthermore, I do not attempt to claim that the overall processes were ‘good’ or ‘bad’ for different interests. I merely seek to investigate how decisions about the structure of policy implementation can be imbued with different substance based on the motivations of these actors. Into the future, in-depth local accounts of water policy implementation would provide a much richer description of the local and historical politics of these processes, enabling a more detailed understanding of the motivations at play and the meanings of local policy languages and practices.  Additionally, it is worth clarifying that my approach to intermediaries is exploratory and not representative. I am interested in how intermediaries distil political objectives (whatever their source) into the languages and logics of policy implementation. My study is not designed to survey or represent all intermediaries in New Zealand. My focus is on understanding how (some) intermediaries work relative to state actors; I do not require complete representation of all intermediaries or all stakeholders to achieve this task. That said, a next step from the 30  analysis developed here would be to catalogue the strategies undertaken by intermediaries according to their named stakeholders or interests. Such an analysis might explicitly investigate specific interests or stakeholder groups, such as local or national Māori collectives, pastoral and other types of agricultural interests, recreational associations, environmental groups, and so on. A fuller characterization of freshwater stakeholders across local and national scales, along with the strategies of their associated intermediaries, would make a valuable contribution to scholarship on New Zealand environmental politics. Sampling  My selection of interviewees was guided by my existing relationships; this affected the nature of the stories I collected. While my physical placement within MfE constrained me to seeking out people who were still working within the building, my institutional situation (e.g. ability to book rooms, MfE email account) allowed me to freely snowball within MfE to search for those who had worked on water policy and implementation to the point where I felt I had reached saturation in relation to the implementation work. Within the regions, I engaged in stratified sampling (as described above) to ensure that I spoke to the relevant policy manager, a relevant scientist, and the key bureaucrat/s driving the implementation work. Thus for both MfE and regional councils I am comfortable that my sampling strategies have generated the most relevant insights possible for my research questions, even though I could have spoken to more people that I did.  For intermediaries, I did not engage in stratified sampling, and the population is not physically concentrated in a building to allow me to easily snowball. My approach here was exploratory, beginning with my existing networks. Drawing on networks from a six-year research project about freshwater management (2010-2016), I first sought out expert intermediaries with >20 years of experience in the environmental sector who could explain the work and strategies of non-state actors such as intermediaries within the resource management system. From this core group, I snowballed out to find other veterans with 31  different geographical and institutional experience of freshwater management, such as experiences with past court cases, for example. I knew I wanted to speak to researchers as well as consultants, and environmental as well as industry intermediaries. Within my existing networks I accessed researchers, consultants, and environmental intermediaries, but I knew few industry intermediaries so I made explicit approaches to invite them for an interview. Since my study of intermediaries is exploratory, stratified or representative sampling was not required. If my sample of intermediaries were biased toward particular interests or perspectives, this does not negate the value of studying how these interests are translated into the policy interpretations. I recognize, however, that my exploratory analysis could be improved upon in the future by seeking out a stratified sample of intermediaries (as described above). Positionality  Finally, I acknowledge that my emphasis on bureaucrats and experts is both theoretical and political. Peck and Theodore (2012) warn that researchers must maintain a critical distance from their research subjects, lest they become ‘dupes’ (p27) of the policy ideas and communities being studied. One of my motivations for undertaking this research was to tell the (largely untold) stories of environmental bureaucrats currently working in councils, the state bureaucracy, and as intermediaries. I wanted to make their politics visible so that they can be recognized and valued by others. While conducting my research and analysis I have been aware that policy workers may act in self-interest, as well as recognizing that policy workers’ actions may be determined by forces beyond their control. In this, I have attempted to scrutinize self-serving claims, while also remaining aware that policy actors are acting within constraints that are not always revisable. It could be argued that I have been ‘duped’ into thinking these actors and their policy ideas matter and that their intents and effects are positive for the environment and democracy. In response to this, I can say three things. First, I have made my motivations and intentions explicit to allow others to see and criticize my reasoning. Second, I have attempted to triangulate the accounts provided to me wherever 32  possible, and I do not equate intention with effect. Third, I have made a systematic effort in every chapter to highlight the constraints within which these actors must work. Western worldview My approach to studying freshwater governance is grounded in a Western social science framework that takes for granted the existence and importance of certain social and material actors and objects in the world, such as the bureaucracy, water chemistry, or ‘scientists’. A perspective grounded in Te Ao Māori (broadly translated, ‘the Māori world’) might have focused on different issues and political arenas, regarding different objects and relationships (such as rivers-as-ancestors), and with a different normative lens (see e.g. Harmsworth et al., 2016; Ruru, 2018). As a non- Māori scholar interested in Western institutions of freshwater governance, I have posed policy-related questions that I feel are important, and which I feel competent to answer. It is important to acknowledge, however, that an environmental politics grounded in Te Ao Māori is urgent and has significant potential for enhancing prospects for environmental justice and ecological sustainability. I hope that my research might align with indigenous freshwater politics by, firstly, offering a way of understanding how interpretations of Māori concepts might be struggled over within Western institutions, and subsequently, illuminating how progressive translations of such concepts might be promoted by Māori from different vantage points across the environmental governance landscape.  1.5 Argument of the dissertation This dissertation provides an empirical exploration of the politics of interpretation within New Zealand’s emerging limits-based system of freshwater governance. I am chiefly concerned with identifying where politics is to be found within this regime, and how it is being conducted. How is the institutional void being engaged by different actors who seek to advance their interests through securing favored interpretations of policy? I focus on the role of state, local government, and intermediaries as they attempt to structure the preferred means of this new regime, to consider how these also privileges certain ends (Hajer, 2003b). I 33  describe what activities these actors undertake to interpret the policy, why they undertake them, and what these activities can tell us about where political agency exists within these new arrangements.  I will argue that the struggle to shape the limits-based regime in New Zealand can be called epistemic politics (Alasuutari and Qadir, 2014). This type of politics involves contest over the definition, meanings, and practices associated with new policy concepts. Epistemic politics constitutes a distinct domain of politics that functions differently to struggles over the content of policy. On the one hand, one might critique epistemic types of politics as elite-centric and technocratic and seek to reject it in all of its forms. On the other hand, ‘Rather than [simply] criticizing these forms of politics for their lack of accountability and their often troublesome relationship to decision making in elected bodies we might also try and figure out how we can build on their success and enhance their democratic credentials’ (Hajer, 2003b: 190). We need to grasp the politics of interpretation if we are to engage it consciously and collectively. By explicating and interrogating these politics of interpretation, I hope to support environmentalists to think more holistically about how the political struggle can be waged even without the levers of political government.  As a theoretical contribution, the chapters of this dissertation revise the institutional void concept to analyse epistemic politics in environmental governance. Hajer’s concept of the institutional void provides some useful starting points but is incomplete as a guide for theoretical and empirical inquiry. Consider, for example, that the metaphor of an institutional void seems to neglect both history and power. We know intuitively that all concepts come from somewhere and someone; there is no such thing as a policy or governance reform that is completely ‘new’ or whose concepts do not have at least initial form in the world. In this way, an initial task for analysis is to identify the historical origins of policy concepts, and to situate these concepts within the political strategies of different groups and interests (Chapter 2). Once we recognize that the institutional void is already somewhat preconfigured, 34  this opens up the prospect of geographical variation. Do different regions encounter slightly different regional institutional voids, where certain actions are easier or more difficult (Chapter 4)? A related task is to consider how differently powerful actors are attempting to structure the institutional void by promoting their own concepts and norms into positions of dominance. Chapters 3 and 5 do this for the state and for intermediaries, respectively.  Within each chapter, I draw upon selected theoretical perspectives on the state, on policy implementation, and experts, to develop a framework for understanding how differently situated actors work in concert and in competition to structure the void of policy meaning. While there are many possible literatures and theoretical traditions I could draw on to understand the actions and intentions of these actors, I focus on those theoretical traditions that help me locate and characterize agency within the policy implementation process. I do not provide a full theorization of the state, for example, from a full range of interpretative and political economic traditions. Instead, I focus on a key axis of debate within the state theory literature regarding the value of insider accounts for problematizing the ‘state’ as a coherent interest or entity. I use that debate to guide my analysis and help me to connect each chapter back to the wider argument of the dissertation.  1.5.1 Structure of the dissertation Chapter 2 places New Zealand’s new policy concepts within historical context to consider their intended meanings and lines of conflict around them. I argue that the concepts contained within freshwater policy – as with any policy – make specific conceptual and linguistic interventions into historical struggles over authority and the categories and logics of governance. Drawing on document analysis and interviews with resource management practitioners, I describe the emergence of four key concepts that crystallize decades of political debate in New Zealand: limits, collaboration, rights and interests, and values. By understanding how the National Policy Statement crystallizes specific interpretations of these 35  concepts, we can understand how the institutional void is prefigured and why dissenting actors might seek to champion alternative interpretations of specific policy concepts. Chapter 3 examines the efforts of the state bureaucracy to structure the institutional void of policy implementation. The state remains one of the most powerful actors to affect policy meanings, so state attempts to steer other actors toward certain interpretations of policy concept merit identification. Drawing on interviews with MfE officials and on my experience within MfE, I describe the projects of the freshwater policy implementation team, the rationale and work of the Water Directorate in relation to the National Policy Statement, and the priorities and projects shaping MfE as an organization. Across these organizational scales, the work of Ministry officials is shaped by multiple logics that are not reducible to the ideology of the Minister or the government of the day. Even those putatively ‘neoliberal’ implementation projects – such as state support of economic valuation – can and should be analyzed for their diversity and polysemy. While the state does engage in structuring the institutional void through a range of programs, the form and purpose of state activities can be affected in significant ways by bureaucrats drawing on organizational, professional, and personal rationalities and values. Rather than rejecting the state as a conduit through which to seek environmental sustainability and social justice, as some have recently advocated (Pulido et al., 2016), I argue that a serious engagement with the practical logics of bureaucratic work can open up space for thinking about where and how environmental and democratic values can be woven through the state apparatus. Chapter 4 explores the geographical constitution of the institutional void by analyzing how five regional councils have implemented the National Policy Statement for Freshwater Management. Not only is the institutional void at least somewhat pre-structured by historical meanings, it is also felt differently by different actors across space. Implementing regions do not all face the same conditions of political possibility; their policy implementation pathways are channeled by local politics and organizational priorities as well as local histories of 36  conflict and environmental change. Drawing primarily on interviews with regional council staff (and a few consultants) from five regions, I identify the local logics of policy implementation within the terms of the regional council actors who assembled them. My aim is to determine whether and how these regional policy actors make decisions that are consequential for the shape of implementation programs. If they do – and I contend that they do – this has implications for understanding who has what kind of power in a decentralized environmental governance regime. By describing five implementation programs and explaining their origins and logics, I hope to explicate how local government bureaucrats can and do act as environmental democratic change agents in this decentralized regime.  Chapter 5 considers some ways in which intermediaries are attempting to structure the institutional void. By developing and promoting certain policy concepts and frameworks, intermediaries are shaping the landscape of freshwater policy implementation. Drawing on interviews with intermediaries and supported by a review of related technical, scientific, and policy documents, I describe several mechanisms through which intermediaries are attempting to structure regional councils’ implementation plans. First, I describe how a scientific model – Overseer® - has acquired preferred status for implementing the concept of ‘limits’, even though many industry and government actors are now trying to rein in its influence. Second, I show how the notion of Good Management Practice offers a regulatory alternative to Overseer® which privileges certainty of economic cost over certainty of environmental protection. Both approaches involve the concept of ‘limits’ which is specified in the National Policy Statement, yet they produce different distributive outcomes. Third, I consider how social and policy researchers are developing and promoting best practices in relation to Māori ‘rights and interests’ and ‘collaboration’. Fourth, I describe how intermediaries re-compose their spatial knowledge and experiences of different settings to advocate for preferred policy elements within a place. I argue that through all of these spatial strategies, intermediaries are attempting to structure the void of policy implementation, and as non-state actors their influence deserves further and ongoing analysis.  37  Chapter 6 summarizes the theoretical contributions of the dissertation and distils what might be learned for New Zealand and international environmental policy communities. After revisiting the contributions of each chapter to a broader theorization of environmental decentralization as a structured terrain of interpretation, I offer some learnings for other decentralizing states, and I suggest some future research directions. A discussion of the recent New Zealand election and freshwater policy developments post-2017 is included in a postscript. Through this dissertation, I revise the notion of the institutional void into a conceptual framework that can be used to characterize and analyze how key actors are engaging in epistemic politics. While the metaphor of a void usefully draws attention to the somewhat open nature of political outcomes through the process of struggle to shape interpretations, it does not systematically grasp how interpretive politics are structured by history, by the state, by local context, and by non-state actors such as intermediaries. The chapters of this dissertation demonstrate how the institutional void alluded to by Hajer is not without structure and power dynamics. Indeed, a more accurate metaphor for understanding the dynamics of epistemic governance might be ‘terrain’ which draws attention to the historically channelized, yet constantly weathered and changing landscape of meaning in environmental politics.  38  Chapter 2 – From public concerns to state policies: a history of New Zealand freshwater policy  2.1 Introduction At a basic level, an environmental policy can be understood as the state’s response to historical public concerns. Public concern might arise over degradation of a particular ecosystem or in relation to particular economic practice, for example, and the state responds by creating a policy to address that problem and secure the state’s legitimacy to govern (Whitehead, 2017). However, state policies are not neutral responses to public concerns. (Fischer and Forester, 1993; Hajer, 1995; Fischer and Gottweis, 2012). Only some public concerns will be selected and formally defined as a state problem, for which policies will be proposed as a solution (Jessop, 2016). Further, the way in which state policy constructs environmental problems is an ideological activity, as environmental policies can privilege certain ways of relating to the environment – for example as a hazard, a resource, a commons, or an ancestor (Hajer, 1995; Robbins, 2012; Dryzek, 2013; Yates et al., 2017). To begin my enquiry into the meaning of freshwater policy, then, it is pertinent to ask what public concerns have emerged in relation to freshwater, and how state policies have framed and responded to these problems.  Policies are not enacted upon terra nullius in which all interpretations of policy are possible or equally likely. Rather, the specific language and concepts used in policies have histories and specific meanings attached to them (Bacchi and Goodwin, 2016). Understanding these histories can tell us who has made claims about a policy issue like freshwater, what those claims include, how governments attempt to strategically incorporate, frame, and address those concerns through policy, and what types of interpretations exist around a given policy. Developing a genealogy of the policy problem can allow us to understand what is at stake with how new policy concepts might be interpreted. 39  This chapter examines how concepts introduced in New Zealand’s National Policy Statement for Freshwater Management are pre-figured by historical debates and public concerns. Specifically, the chapter argues that the interpretation of the National Policy Statement is shaped by historical debates about indigenous rights and sovereignty, diagnoses of the causes of environmental degradation, decisions about national identity and economic development, and perceived ‘gaps’ in New Zealand’s environmental regulatory framework. These public concerns have been amalgamated by successive governments into a version of the freshwater ‘problem’ that the National Policy Statement is intended to address. The National Policy Statement codifies elements of these public concerns into keywords, which act to delimit the problem and frame it in a way that the policy can be said to have succeeded in addressing. In this way, a major aspect of interpretive conflict is about whether the keywords used by the state adequately reflect the deeper public concerns that the policy purports to resolve. To provide some history to the institutional void, it is necessary to interpret these keywords within the wider array of public concerns from which they were drawn. Tracing the historical roots of contemporary policy concepts serves two functions. First, as a historical foundation for the chapters that follow, this chapter provides an orientation to important social, regulatory, and environmental issues related to freshwater policy in New Zealand. Terms like ‘limits’, ‘collaboration’ and ‘rights and interests’, for example, have specific meanings in New Zealand that need to be understood to grasp what is at stake in recent policy developments. Second, historical analysis reveals how the concepts introduced by a new freshwater policy are likely to be interpreted, and how state and non-state actors have struggled to shape these interpretations. If a public concern is left out of policy or inappropriately codified in a keyword, non-state actors may promote their own keywords or interpretations, which the state may then respond to with a policy revision, and so on. Understanding the historical cycles of incorporation, contestation, and revision of policy concepts is a first step toward understanding how the interpretive politics within an institutional void is structured by a range of actors.  40  This chapter considers how freshwater policy codifies selected public concerns, why, and with what effects. In New Zealand, claims about freshwater – and state responsibilities for it – are related to issues of colonial history, biophysical understandings of environmental change, economic development, and regulatory arrangements. In the sections that follow I examine each of these issues, tracing the historical antecedents of public concerns that manifest today, and against which current policies derive meaning. I highlight how each aspect of New Zealand’s freshwater problem has been codified in selective ways over time, and which have been further embedded through recent freshwater policy reform. My analysis in this chapter draws on government and policy documents; a review of the academic literature on New Zealand’s environmental governance; semi-structured interviews with veteran intermediaries of New Zealand’s resource management system, three sitting members of parliament, central and local government officials; and notes from two public meetings with the Minister for the Environment in 2016-2017 (see Chapter 1).  Section 2.2 outlines how freshwater is entangled with the problem of ongoing colonialism and indigenous rights. Section 2.3 considers freshwater as a biophysical problem, describing the evolution of scientific knowledge of New Zealand’s freshwater estate, and its associated diagnoses of cause and effect. Section 2.4 shows how freshwater has become a political-economic problem linked with specific industries and urban development, and reviews public perceptions of freshwater over the 2009-2018 period of major national freshwater reform. Section 2.5 describes freshwater as a regulatory problem, caused by the failed implementation of New Zealand’s central piece of environmental legislation, the Resource Management Act 1991. Section 2.6 discusses how successive waves of governmental policy since 1999 have attempted to resolve these freshwater problems through the mechanisms of national policy. These policies do more than simply respond to the freshwater problem; they actively integrate and perform a new concept of the freshwater problem. Finally, in Section 2.7 I summarise the historical context of five keywords that have emerged to centre-stage in discussions of national freshwater policy since 2008. These keywords can be understood as 41  contemporary ways of codifying and responding to the diverse public concerns in relation to fresh water outlined in Sections 2.2-2.5.  2.2 A problem of colonization: freshwater and indigenous rights  Contemporary freshwater policy needs to be understood in the context of colonialism and the historical and ongoing dispossession of Māori (New Zealand’s indigenous people). Ongoing indigenous struggles for recognition, authority, and sovereignty are being pursued against the historical fact that these were promised to Māori by the colonial British Crown. While the Treaty of Waitangi (Te Tiriti o Waitangi) 1840 promised Māori the undisturbed possession of their lands, taonga (treasures), and resources, the Crown failed to fulfil these obligations and instead committed grievous acts of murder, theft, and displacement (see e.g. King 2003). Recognition of the Crown’s failures led to the creation of the Waitangi Tribunal and a legal process of settlement between iwi (tribes) and the Crown since 1975. This process of legal settlement is one, partial form of restitution, which has included the creation of specific environmental governance powers along with other forms of redress (Section 2.2.1). The Treaty has also been acknowledged by various environmental and governance laws in New Zealand, notably the Resource Management Act 1991, which has created expectations for regional councils to involve iwi in environmental decision making (Section 2.2.2). A third form of restitution is the quantitative allocation of renewable environmental resources such as fisheries, and perhaps eventually freshwater (Section 2.2.3). 2.2.1 The Treaty of Waitangi 1840 and the settlement of historical grievances New Zealand is a settler colonial society. After early contact between the indigenous Māori and sailors from Holland and then England, European colonists began to settle in New Zealand from the early 1800s (King 2003). While New Zealand has no written constitution or Bill of Rights that supersedes the laws of parliament, the Treaty of Waitangi signed in 1840 stands as its founding and most significant document, and it is through the Treaty that the 42  authority to govern New Zealand rests (Mulholland, 2015). The Treaty contains three articles, the meanings of which have been the subject of intense and ongoing debate. Broadly, Article 1 ceded to the British Crown the ability to govern New Zealand, Article 2 guaranteed Māori undisturbed possession over their lands, taonga (treasures), and resources, and Article 3 guaranteed Māori the same rights as Crown subjects (Ministry for Culture and Heritage, 2017). Because the Treaty was written and signed in both Māori and in English, differences of meaning between versions have been hugely consequential. For instance, while Māori ceded an unfamiliar concept of governorship to the Crown, they did not agree to giving up authority to self-govern (see King, 2003). These interpretive differences – as well as blatant disregard by the Crown for its own obligations – have led to continued demands from Māori for their promised rights to land, sovereignty, and equality under the law (see Mulholland, 2015; Sullivan, 2016; Charpleix, 2018).  The actions of the Crown toward Māori after the signing of the Treaty were often violent and reprehensible, involving murder, the confiscation of land, massive displacement, and forced cultural assimilation (King, 2003; Stevens, 2015). In recognition of these injustices, in 1975 the New Zealand government set up the independent Waitangi Tribunal to hear claims of breaches of the Treaty and to make recommendations to government on matters of law and policy. In 1995 the government set up the Office of Treaty Settlements to forge formal legal settlements with ‘large natural groups’ (LNGs) consisting of Māori who share an identity around a common grievance, whether geographical or tribal (New Zealand Government, 2018). These legal settlements can involve land transfers, governance powers, and financial payments, among other things. For iwi (tribes) across New Zealand, the settlement process represents a significant step to reclaiming their authority in a symbolic and material sense. When the settlements are agreed between the government of the day and the LNGs, the settlements are written into law and passed, becoming enshrined in legislation. Settlements include recognitions that designate specific geographical features and an area over which a given iwi (a tribe sharing a common ancestor) has specific legal status, and designated iwi 43  must be consulted by councils for any activities affecting these designations. Iwi with this status are referred to within the Resource Management Act 1991 as mana whenua (designated authorities). In the global context of indigenous and environmental law, New Zealand’s Waitangi Tribunal and settlement processes grant significant statutory authority and (sometimes considerable) financial recompense to indigenous peoples, especially when contrasted with the settler state governments of Australia, Canada, or the United States. For example, in a world-famous settlement passed into law in 2017, the Whanganui iwi in the North Island have settled with the Crown to grant the Whanganui River the same legal status as a human being (Boyd, 2017; Roy, 2017; Charpleix, 2018). This is being operationalized through a specific co-governance structure, and allows the river’s representatives (one Crown appointee and one iwi appointee) to assert the river’s rights to dignity and autonomy in court and other planning processes. This settlement included $80M payment in financial redress, as well as other governance powers. The Whanganui settlement has only been the most recent of high-profile settlements between the Crown and iwi; Waikato-Tainui’s 1995 settlement with the Crown included $170M in land, payments, and financial holdings, and Crown’s settlement with Ngāi Tahu (1998) included $170M in financial redress, among other forms of cultural redress and governance roles. In 2014, the Tuhoe-Crown settlement vested the area of Te Urewera National Park back from the Crown, transforming Te Urewera into ‘a legal entity’ with ‘all the rights, powers, duties, and liabilities of a legal person’ (section 11(1)), that would be governed by a Tuhoe-Crown appointed board for the public interest (see Ruru, 2014). 2.2.2 Māori participation in environmental management  While the Tribunal and settlement processes address past grievances relating to the Treaty of Waitangi, there is also an expectation that any new laws and policies should give effect to the Treaty, or else risk further claims for redress through the Tribunal. As Māori have unique relationships to New Zealand’s environment that are guaranteed protection under Article 2 of 44  the Treaty, government policies relating to the environment must explicitly address Treaty principles (Robb et al., 2015). If Māori consider that new policy or legislation contravenes the Treaty, they can file a claim in the Waitangi Tribunal to seek redress.  Over the years, many laws and policies enacted by local and national governments have attempted to give effect to Treaty obligations. Although a rigorous evaluation of the authenticity and effectiveness of such attempts is beyond the scope of this chapter, it is worth noting that provisions for Māori participation in decision making within major pieces of legislation such as the Resource Management Act 1991 (RMA) and Local Government Act 2002 were generally vague and minimal. Section 8 of the RMA, for example, instructs all persons exercising powers within the RMA to ‘take into account the principles of the Treaty of Waitangi’. Although Section 35A of the RMA instructs councils to keep records of iwi and hapū (subtribes) within their regions, expectations for Māori involvement in planning beyond this were underspecified, and as such progress toward shared decision making has been limited (Williams, 2007; Robb et al., 2015).  While the RMA and other laws have highlighted the significance of the relationships between Māori and the natural environment (e.g. identifying culturally meaningful sites and ecosystems), in practice the involvement of Māori in freshwater planning has been heterogeneous and generally lacking, although some regional councils have made considerably more effort than others (Harmsworth et al., 2016). Increasingly, for example, regional councils are creating co-governance boards to oversee environmental planning and decision making; these can range from purely advisory (e.g. Northland) to having voting and veto rights (e.g. Greater Wellington, see Chapter 4). While such co-governance boards are applauded by some Māori researchers (e.g. Harmsworth et al., 2016), in practice the composition of these boards can involve intense political struggle and their legitimacy can be contested by Māori. Some iwi have had their genealogy and grievances formally acknowledged by the Crown through the settlement process and thus have acquired mana 45  whenua status, whereas other iwi have not. This adds complexity and conflict when seeking formal representation in the planning process (e.g. see Sinner and Harmsworth, 2015). Judge Joe Williams (2007) observes that it has been Treaty settlement processes – rather than vague RMA provisions – that have materially enabled Māori to engage meaningfully with resource management processes. Crawford (2007), Robb et al. (2015), and Ruru (2018) also contend that regional councils’ engagement with iwi through the RMA has been variable and unsatisfactory. They cite a need to train council staff in the role of the Treaty as well a need to compel councils explicitly to engage with iwi through defined parameters. As we will see in Section 2.6 (national policy developments), the National Policy Statement on Freshwater Management has included additional provisions to specify expectations for the involvement of iwi in freshwater management.  2.2.3 Māori ownership of water While Article 2 of the Treaty guaranteed Māori the ‘undisturbed possession’ of their land, taonga, and resources, the issue of ownership of water was not directly addressed. In the context of increasing resource scarcity and commercial use of water, the issue of ownership has risen to become a major political topic. Stated first in the 1960s and again more forcefully in the late 2000s, the Crown has asserted the common law position that ‘no one owns water,’ reserving the sole right to manage water for the public interest (Strang, 2014; Salmond, 2017). Crown law – such as the RMA – created a framework to allow private interests to extract water, which could then be sold for a profit, but this was not the same as allocating ownership (according to the argument). This legal doctrine was thrown up for contest in 2011 when the fifth National government proposed to sell off major public assets including 49% of the state-owned enterprise Mighty River Power, a hydropower company. The Māori Council, a statutory body representing all Māori, argued to the Waitangi Tribunal that by effectively selling the rights to manage water, the Crown is asserting ownership, and if ownership exists then Māori must have prior right to the resource. The Tribunal agreed that Māori ownership 46  of water was indeed equivalent to the Crown’s, and it recommended delaying the sale of the assets until such a time as Māori rights and interests in water could be worked out (see Strang, 2014; Salmond, 2017). Despite this recommendation, and despite a public-initiated referendum overwhelmingly against the asset sales (67%), in 2013 John Key’s National government sold off 49% of Mighty River Power, Meridian Energy, Genesis Power, Solid Energy, and Air New Zealand.  Since then the issue of Māori ‘rights and interests’ in water have become a prominent issue that continues to complicate questions of how freshwater might be allocated in a quantitative sense. This stands in stark contrast to the example of fisheries. In the mid-1990s when New Zealand developed its then-revolutionary quota management system to ensure fishing yields were sustainable, the Crown settled with Māori at the national level to allocate 20% of the catch quota to Māori through a new national fisheries organization. With fisheries, Māori were given a quantitative percentage of the commercially harvestable yield. With freshwater, however, Ruru observes that successive governments have continued to deny Māori their economic recompense in relation to freshwater. She argues that ‘real reconciliation in a decolonized context will remain elusive until fair, complete, and holistic restitution for water grievances is offered across all redress spectrums, including cultural, commercial, and proprietary’ (Ruru, 2013: 311). 2.3 Changing freshwater environments: knowing the national freshwater ‘problem’ Deteriorating freshwater ecosystems across New Zealand provide the biophysical basis for claims of freshwater ‘crisis’ and demands for strong state regulation of human activities (Knight, 2016). But changing material conditions are not enough to generate this diagnosis; scientific knowledge of ecological change had to be constructed, scaled, and made authoritative. Moreover, explanations for changing ecosystems had to be developed to identify which human activities ought to be regulated, and how, to protect freshwater ecosystems. It has taken several decades to build a firm scientific foundation for the 47  identification and attribution of the ecological ‘problem’ posed by ongoing forms of water extraction and land use. While New Zealand has relatively abundant freshwater in volumetric terms, its distribution across the environment, coupled with pollution from land use patterns, mean that water quantity and quality require careful management (OECD, 2017b).  In the early 1990s there were indications that New Zealand’s freshwater estate was already under strain from agricultural development (Sinner, 1992; Smith et al., 1993). However, since water extraction and pollution were not yet pushing ecosystems past visible tipping points, and since pollution from large sources (such as industrial or wastewater plants) was being reduced under the Resource Management Act (see Section 2.5), there was little political appetite to further examine these environmental effects or consider regulatory options (Consultant 2a, 2016). The OECD’s 1996 environmental performance review of New Zealand concluded that ‘Thanks to a very low intensity of water use and low overall levels of pollutant discharges fr0om point sources, New Zealand’s rivers, lakes and groundwater generally present very high water quality’ (OECD, 1996: n.p.).  2.3.1 Characterising ecological decline: State of Environment reporting New Zealand’s freshwater ‘problem’ has been made visible in significant part through legally-mandated State of the Environment (SoE) reporting conducted by the Ministry for the Environment, drawing on regional council data. While the first SoE report published in 1997 noted a high overall level of environmental quality, it recognized significant urban and agricultural pressures on aquatic ecosystems, as well as the impacts of invasive species and habitat transformation. In urban environments, sewerage overflows (and bacteria such as E. coli), heavy metals, hydrocarbons (from cars), and sediment were key freshwater contaminants (MfE, 1997). In rural areas (i.e. most of the country), water quality was generally high in mountain tributaries and decreased in lowland streams and pasture-dominated landscapes, with some rivers unsuitable for swimming due to high levels of fecal contamination (ibid). Within some areas of intensive dairy farming, water quality was even 48  too toxic for livestock to drink (MfE, 1997: 7-7). In 2007, the second SoE report concluded that while New Zealand’s water was ’still generally good by international standards’ (MfE, 2007: 304), demand for freshwater had increased more that 50% since 1997 and ‘some aspects of water quality are getting worse’ (MfE, 2007: 261). In particular, ‘increasing pollution from non-point sources, such as diffuse run-off from pasture and from paved surfaces in urban areas, poses the greatest challenge for water management in New Zealand’ (MfE, 2007: 304). Increased stocking rates and use of nitrogen-based fertilizers were identified as key contributors of nutrients (nitrogen and phosphorous), sediments, and animal effluent (including bacteria such as E. coli) into waterways (ibid: 266).  By the 2015 SoE report, multi-decadal analyses of river water quality were now possible. Over 1989-2013, 60% of river monitoring sites showed a significant increase in total nitrogen concentrations, while only 9% showed a decrease, and the rest were indeterminate (see Table 1). During this period, total nitrogen increased 12% on average across all sites (MfE and Stats NZ, 2015: 10). The 2015 and 2017 SoE reports conclude that nitrogen (mainly from fertilizer and cow urine) and E. coli (from human and animal feces) are higher in urban and agricultural land use areas, and ecological health (as indicated by the macroinvertebrate index) is lowest by far in urban sites. In 2017, the freshwater-specific SoE report found that 72% of reported native freshwater fish species were threatened or at risk of extinction in 2013, along with 34% of freshwater invertebrates and 31% of reported plant types (MfE and Stats NZ, 2017: 14-15). A summary of information on the state of New Zealand’s freshwater ecosystems from 2017 is shown in Figure 1.   49  Table 1. Trends for water quality variables in NIWA’s National River Water Quality Network 1989-2013 (77 sites). Source: MfE and Stats NZ (2015).  Variable Trend Sites showing a statistically significant increase (%) Sites showing a statistically significant decrease (%) Sites showing an indeterminate trend (%) Clarity ⇗ 64 9 27 Total nitrogen ⇗ 60 14 26 Nitrate-nitrogen ~ 52 27 21 Ammonia-nitrogen ⇘ 4 78 18 Total phosphorous ~ 38 30 32 Dissolved phosphorous ⇗ 51 14 35 Macroinvertebrate community index (MCI) ~ 5 13 83    50  Figure 1. Infographic of statistics from Our Fresh Water 2017. Reproduced with permission. Source: MfE (2017j).  51  2.3.2 Explaining ecological decline: pastoral intensification and urban development in the spotlight The picture emerging from successive SoE reports is one of widespread (albeit uneven) degradation of freshwater ecosystems, driven by the physical transformation and biochemical contamination of rivers and other waterbodies. Physical transformation of rivers has been driven by processes of urbanisation, infrastructure (e.g. roads, dams), and wider land use changes in the catchment (e.g. water use and sediment production). The chief chemical contaminants causing ecological decline – nitrogen and phosphorous – have been clearly linked to the use of industrial fertilizers and increasing stock numbers. For example, Julian et al. (2017) found that nitrogen and phosphorous levels across space and time could be best predicted by cattle density, followed by the level of plantation forest coverage. Excess nitrogen and phosphorous poisons aquatic animals and feeds algae blooms, which in turn choke the water of oxygen and smother existing habitat. In addition to nitrogen and phosphorous contamination, bacteria such as E. coli pose a direct health concern for humans and animals. E. coli and other bacteria from human wastewater and animal effluent in waterways can make animals sick, as occurred when 5,500 human residents of Havelock North became sick in 2016 from drinking water contaminated with Campylobacter species of bacteria sourced from sheep feces (Government Inquiry into Havelock North Drinking Water, 2017). Previous studies have shown that urban and intensive pastoral land uses are associated with higher E. coli and nitrogen levels in waterways (Larned et al., 2016). Pastoral land use is also associated with higher water temperatures due to the removal of shading trees, which inhibits fish growth and promotes algae blooms (Ballantine and Davies-Colley, 2014).  The scientific basis of freshwater degradation has its foundations from state and non-state scientific analyses, and has converged around a small number of variables. While SoE reports can present positive and negative spins on data, they are at least partially constrained by mainstream scientific findings. While the SoE reports may place more emphasis on 52  characterising ecological decline rather than identifying or quantifying its causes, ample scientific research and policy analysis underpins some general claims about water quality (see PCE, 2013, 2015). This resulting scientific consensus has emphasised a set of common pressures: • Nitrogen, sourced from fertilizer (via animal effluent) • Phosphorous, sourced from fertilizer • E. coli and other bacteria, sourced from animal and human feces • Sediment, sourced from erosion-prone land uses 2.4 Freshwater as a public problem: economic discourses and political demands Public meanings of freshwater have changed over time, and have been shaped by media representations, economic changes, and changes in how people interact with the environment. Here I briefly survey how freshwater has been framed in the public sphere as an economic resource vs a public commons, as this affects how political demands regarding freshwater have been made upon governments.  Māori have had deep spiritual and ancestral relationships with the land and water long before European colonization (see Robb et al. 2015, Knight 2016, Salmond 2017). Since colonization, farming for sustenance and then the market has become a significant and enduring part of cultural identity for many pākehā (European) New Zealanders, as well as Māori (King, 2003). Early frontier narratives described settlers inhabiting dangerous new territories in need of taming and cultivation, who saw rivers as transportation routes, hazards, picturesque settings, and drains for industrial and residential production (Knight 2016). Acclimatisation societies planted willows on riverbanks, introduced mammalian predators (to ecosystems where mammalian predators had not existed), and introduced trout to waterways, all to make New Zealand feel more like England (ibid). In more recent times, recreational pastimes have developed, such as angling, kayaking and hiking, and dams were built for both water 53  supply and, later, hydropower. Flood control and engineering structures radically transformed many rivers in urban and rural settings. These settler colonial narratives of freshwater landscapes continue to influence New Zealand’s freshwater politics. For example, contemporary aspirations for ‘swimmability’ in New Zealand are anchored in personal memories of a more idyllic and rural New Zealand, where families took kids to swim in rivers over summer or picnicked down at the lake (interview, C. Delahunty 2016; public meeting notes, N. Smith 2017). Compared to other developed nations, New Zealand has a relatively high standard of living, wellbeing and environmental quality, albeit with lower than average income, wealth, and productivity (OECD, 2017a). Agriculture continues to play a major part in the otherwise service-focussed New Zealand economy, with dairy accounting for 23% of exports, the largest share from a single industry (Statistics New Zealand, 2015). From 1991 to 2018, the New Zealand population grew from 3.5M to 4.9M (Statistics New Zealand, 2018), with 14.9% of residents registering as Māori in the 2013 census (Statistics New Zealand, 2013). After major neoliberal reforms to the public sector of the 1980s, the fourth National government (1990-1999) sought to open the New Zealand economy through free trade agreements in 1990s. The Uruguay round of the General Agreement on Tariffs and Trade resulted in lower barriers for NZ exports, creating a windfall gain in market share for low-cost agricultural producers from New Zealand (PCE, 2004). This led to a ‘dairy boom’ that transformed land use across New Zealand in the 1990s (Consultant 2a 2016, interview, D. Parker 2016). Massive land use conversion from sheep and beef farming to dairy commenced, particularly in Canterbury, Southland, and Manawatū-Wanganui (also known as Horizons) regions, coupled with an intensification of existing dairy operations (see Figure 2). The scale and pace of environmental change was highlighted by environmental groups in the late 1990s and early 2000s, through both formal and informal strategies.    54  Figure 2. Dairy cattle by region 1994, 2002, 2015. Reproduced under a Creative Commons Attribution 4.0 International Licence. Source: Statistics New Zealand (2017a).  In the early 2000s, the New Zealand Fish and Game Council (henceforth Fish and Game) ran a major public campaign drawing attention to the environmental impacts of ‘dirty dairying’ (Holland, 2014). This coincided with the influential 2004 Growing For Good report published by the independent Parliamentary Commissioner for the Environment. That 236-page study of nitrogen and water in farming systems identified concerning trends toward pastoral intensification (especially around dairy farming) and projected how these trends would likely impact water quality (PCE, 2004). The ‘dirty dairying’ campaign and Growing For Good report fuelled much public debate and a challenge to the fifth Labour government’s (1999-2008) attempt to brand New Zealand as ‘clean and green’ to its trading partners. A key outcome of the ‘dirty dairying’ campaign was the creation of an industry-driven voluntary 55  agreement called the Clean Streams Accord, which promised to fence stock out of waterways. While some observers are critical of the Accord and other voluntary measures to improve environmental performance (see Holland, 2014), Kevin Hackwell from the environmental organization Forest and Bird emphasises that the Clean Streams Accord led to industry and environmental groups working together in a positive way outside of the courts (interview, K. Hackwell, 2017).  The year 2009 marked the beginning of the contemporary phase of intense public interest in freshwater issues. In that year, the trialling of a novel ecological assessment method on a large North Island river produced headlines such as Manawatū River ‘among the worst in the West’ (Morgan and Burns, 2009), which led to intensifying national debate about farming and water quality. In the Manawatū region, this resulted in the creation of the Manawatū River Leaders Forum, a stakeholder collective promising a range of actions to clean up the river (e.g. see Naovalat, 2015). In early 2011, centre-right Prime Minister John Key – who was also Minister for Tourism – appeared on the BBC’s interview show Hard Talk and was challenged about New Zealand’s marketing claim to be ‘100% Pure’. When the interviewer cited ecologist Mike Joy’s research and claims of widespread ecological degradation, Key dismissed Joy’s diagnosis as the views of ‘one academic.’ This sparked much commentary in New Zealand about the objectivity of science, the state of freshwater, and the relationship between the environment and the national economy (e.g., Manhire, 2011). By this time, freshwater was firmly on the political agenda and a matter of public discourse. In the 2011 general election the Green Party campaigned on ‘Rivers, jobs, and kids’, and their representation rose from 9 to 14 seats in a 120-seat parliament, a level which it then held in the 2014 election, in part by campaigning for environmental care (Levine and Roberts, 2015). As will be discussed in Section 2.6.2, the term of the fifth National government (2008-2017) involved the creation of the Land and Water Forum as a civil society group to guide major reform to national freshwater policy; the creation of a National Policy Statement on Freshwater Management in 2011, with revisions in 2014 and 2017; and in 2012 the 56  reassertion that ‘no one owns water’, as the government sold off 49% of major public assets (see Section 2.2.3). Since 2014, much news and commentary has been produced about freshwater management, driven in part by the significance and pace of national level freshwater policy developments. Table 2 presents an indicative sample of in-depth news articles, commentaries, and investigative studies (including documentaries) from 2016-2018 alone. Over the last two years, at least five types of societal concern about freshwater can be distinguished; illustrative headlines are summarised in Table 3. One is the increasing visibility of ecological degradation and collapse of freshwater ecosystems due to pollution and over-extraction. The drying up of the large Selwyn River in Canterbury, and the ecological crises of the Waituna Lagoon, Te Waihora/Lake Ellesmere, and Lake Horowhenua, are major examples (see Table 3). Second, direct concerns for human and animal health have been brought into focus through recent events in the Hawkes Bay and Auckland regions, among others. In Hawkes Bay, there have been reports of pet animals dying from drinking environmental water and playing in shallow lakes, and an outbreak of human bacterial infection in Havelock North caused national headlines when fecal bacteria from farm animals leaked into the town’s water supply in 2016. In Auckland, several popular beaches have had to close due to bacterial contamination. Third, outcry has arisen regarding the regulatory status of bottled water. News coverage of water bottling plants in Canterbury and Hawkes Bay has highlighted how these firms extract water free of charge and export it to foreign markets. This has roiled environmental groups and political parties on the left; the Green Party for example, campaigned in 2017 on placing a charge on the use of water for bottling and sale. Fourth, the aspiration for ‘swimmable’ rivers and lakes has emerged as a political demand due in significant part to the public and political backlash to the fifth National government’s freshwater policy in 2014 (discussed below in Section 2.6.2). This aspiration intensified significantly in 2017 when the government proposed ‘90% of rivers and lakes swimmable by 2040’ (MfE, 2017a), and public discourse became filled with scientific critiques of the government’s definition of ‘swimmability’ and 57  cries of statistical chicanery (see Table 3). Finally, there has been a new interest in freshwater policy, agricultural industry, and the roles of regional and national government in securing the environmental public good. This takes the form of coverage and commentary about the roles of regional councils as environmental regulators.  Table 2. A sample of commentaries, documentaries, and in-depth investigative articles on New Zealand’s freshwater crisis (from 2016-2018). EDITORIALS 100 percent pure or 60 percent polluted? (2016) Radio NZ Can you believe we're fighting for clean water in New Zealand? (2016) Manawatū Standard  Water shaping up to be hot topic at next election (2016) TVNZ  Jack Tame: Talking 100% Pure and water till the cows come home (2017) NZ Herald 'Unprecedented' freshwater rescue plan unveiled by advocacy groups (2017)  Economist blames cows for state of NZ's rivers and lakes (2017) Radio NZ Farm impact on water quality blown out of proportion - Federated Farmers (2017) Newshub NZ’s economic growth model pushing environmental limits – report (2017) Radio NZ Sustainability of New Zealand's dairy industry under question (2018) Newshub  INVESTIGATIVE ARTICLES Take me to the river (2014) The Listener River stance: Mike Joy's controversial crusade to save New Zealand's waterways (2015) The Listener Water fools? (2017) Special report series by Radio New Zealand, includes: Worry in Waihora Pumped dry – central Otago farmers’ fight for water ‘Recharging’ Canterbury’s aquifers Southern lakes: a changing landscape Pollution in a Piha paradise The river is me River damage from forestry in Northland Sacred pipi beds polluted Insight: A line in the sand for freshwater quality? (2017) Radio New Zealand  Brighter Future? Murky prognosis for freshwater health (2017) Radio New Zealand Troubled waters (2017) New Zealand Geographic Special report: how polluted are New Zealand's rivers? (2017) Newshub 58  Insight: Fighting for NZ's Rivers and Lakes (2017) Radio NZ  Farming for our future (2018) New Zealand Geographic  DOCUMENTARIES Saving the Selwyn (2017) Television documentary, TVNZ (16m) New Zealand: Polluted Paradise (2017) Two-part television documentary, Aljazeera. Part 1 (25m), Part 2 (25m), dir. N Zalk Seven rivers walking (2017) Documentary film, 82m, dir. K Gallagher, G Barnes By 2018, freshwater issues have become a prominent feature the national public discourse and politics, with a recent poll showing that over 75% of New Zealanders (n=1000) indicate that they are ‘very’ or ‘extremely’ concerned about the pollution of waterways (Radio New Zealand, 2018b). A longitudinal study of New Zealander’s attitudes towards freshwater recorded a recent and sharp uptick in the number of people who consider the quality of rivers and lakes as ‘bad’ or ‘very bad’ since 2010 (see Hughey et al., 2016). In the arena of international reputation, the OECD (2017b) report that the production and maintenance of New Zealand’s ‘100% Pure’ brand has become a major matter of concern for many industry and trade sector organizations.   59  Table 3. A sample of media articles about freshwater (2016-2018), organized into five themes. Theme Title  (Year) and source Freshwater ecosystems in crisis Fresh water results worst ecology professor has seen  (2016) Radio NZ Lake Ellesmere algal bloom warning lifted after three years  (2017) The Press Selwyn's water woes becoming 'worse and worse' (2017) The Press Shocking sight of Waikato stream running green with effluent from 'deliberate' dairy farm discharge  (2017) TVNZ Public warned to stay out of Waituna Lagoon  (2017) Deformed snails found in toxic runoff-plagued Wellington stream  (2017) Dramatic maps released showing state of crisis in nation's wetlands  (2018),nz Lake Horowhenua 'stuffed' Forest and Bird say (2018) The Manawatū Standard Cow pee: 200 tonnes of nitrogen leaching a day  (2018) The Country Lake Ōmāpere turns green and smelly with algae bloom  (2018) Newshub Hundreds of eels found dead in Christchurch stream  (2018) Human and animal health Gastro bug hit 5000 in Havelock North  (2016) Radio NZ Canterbury's poisonous Lake Forsyth kills sheep, full of green slime  (2016) Toxic algae warnings in Hawke's Bay after dog's death in Tukituki River  (2016) The Dominion Post Christchurch's pure drinking water could be contaminated due to farming  (2017) Swimming off limits at 50 Auckland beaches (2018) Newshub Water bottling  ECan accused of 'bending the law' over consents for water bottling plants  (2018) Hawke's Bay water bottling plant lies dormant for four months  (2017) Ashburton residents demand council ditch water bottling plan  (2016) TVNZ Elation as Ashburton council backs out of controversial water bottling deal  (2016) Call for halt on water bottling consents  (2017) Radio NZ Swimmability New Government target to see 90 per cent of rivers and lakes 'swimmable' by 2040 (2017) 60  New swimmable standard 'less stringent' – NIWA  (2017) Radio NZ Changes to clean-water standards a 'shambles', say conservationists  (2017) NZ Herald Swimmable waterways standard 'confusing' – scientist  (2017) Radio NZ Labour wants waterways 'genuinely swimmable' in five years  (2017) New Zealand Geographic Councils are already struggling with swimmability targets. Here's why that's a problem (2018) The cost of making our rivers swimmable  (2018) NZ Herald Regional councils, central govt, and industry By the numbers: ECan's 'relaxed' monitoring regime leaves thousands unchecked  (2016) The Press Federated Farmers boss admits spilling effluent into stream  (2016) Radio NZ Illegal water-take data a 'wake-up call' for all water users  (2016) Timaru Herald Irrigators are Canterbury's biggest rule-breakers (2016) No plans to clean up Auckland aquifer  (2016) Radio NZ RMA failed to protect New Zealand - new report  (2016) NZ Herald 61  2.5 Freshwater as a problem of environmental planning: the Resource Management Act 1991 The human activities affecting freshwater and its ecosystems are subject to regulation by many New Zealand laws, foremost of which is the Resource Management Act 1991. The Resource Management Act 1991 (henceforth RMA) is the primary organizing environmental law in the New Zealand, and was intended to provide a framework to ensure the sustainable use of environmental resources. Considering the contemporary crisis of deteriorating freshwater ecosystems, many policy actors have diagnosed the RMA and its planning system as having failed. These diagnoses have heavily influenced current understandings of freshwater as a regulatory ‘problem’, and it is important to contextualise these diagnoses as well as consider which other readings of the problem have not gained traction in the present. This section describes the origin and philosophy of the RMA (Section 2.5.1), how it works in practice (Section 2.5.2), and explores arguments for how the RMA ‘failed’ to prevent the contemporary freshwater crisis.  2.5.1 Context and philosophy of the RMA The RMA was initially developed toward the end of New Zealand’s fourth Labour government (1984-1990), which was responsible for undertaking some of the deepest neoliberal reforms to the public sector across the industrialized West (Kelsey, 1995). Inspired by the 1987 Brundtland Commission’s report advocating for nations to pursue sustainable development as a normative goal, deputy Prime Minister Geoffrey Palmer advocated for a major restructuring of environmental management law around these principles (Randerson, 2007; Palmer, 2013; Knight, 2018). Although Palmer’s legislation could not get passed in parliament within Labour’s term, the incoming environment minister of the fourth National government (1990-1999) asked Tony Randerson, Guy Salmon, and other environmental experts to lead a review of Palmer’s proposal and develop what became the Resource Management Act 1991 (Randerson, 2007).  62  The RMA was recognized internationally as the first piece of legislation to enshrine the concept of sustainable development into law (Consultant 2a; Peart, 2007a; EDS, 2016). Philosophically the RMA is built around three main pillars (Randerson, 2007). The first is ‘to promote the sustainable management of natural and physical resources’ (See Table 4). In principle, all planning decisions and regulations made within the powers granted by the RMA must fulfil the sustainable management criterion. The second pillar is effects-based management. The RMA, in contrast to the Town and Country Planning Act 1977 that preceded it, does not and cannot direct land use based on societal or economic benefits; it enables only the acceptance or rejection of development applications based on environmental effects. The intent was to allow land users to improvise to reduce environmental effects, instead of prescribing activities (see Memon and Gleeson, 1995; Palmer, 2013). Third, the RMA embraced the principle of subsidiarity by delegating responsibility for implementing the Act at a regional scale to 16 elected regional councils (see Figure 3)3. As Palmer (2013) himself describes it, ‘the central philosophy behind the Act [was] to move away from a system of centralised activity-based planning to one driven by the effects of activities, with the assessment of those effects to be made at the local level’ (p22). The RMA was thus intended to provide strong measures to ensure environmental protection, without reaching too far into decision making about the nature of development (see Memon and Gleeson, 1995; Randerson, 2007; Salmon, 2007b).                                                     3 At first 17 areas were designated, but within a few years the Act was revised to the 16 regional boundaries that have endured to the present. 63  Table 4. Part II, Section 5 of the Resource Management Act 1991, as of 18 October 2017. Available at Purpose (1) The purpose of this Act is to promote the sustainable management of natural and physical resources. (2) In this Act, sustainable management means managing the use, development, and protection of natural and physical resources in a way, or at a rate, which enables people and communities to provide for their social, economic, and cultural well-being and for their health and safety while— (a) sustaining the potential of natural and physical resources (excluding minerals) to meet the reasonably foreseeable needs of future generations; and (b) safeguarding the life-supporting capacity of air, water, soil, and ecosystems; and (c) avoiding, remedying, or mitigating any adverse effects of activities on the environment   64  Figure 3. Map of New Zealand’s 16 regional council boundaries. Reproduced with permission. Source:  65  2.5.2 How the RMA works The RMA prescribes a nested hierarchy of decision making layers (OECD, 2017b). At the top of this hierarchy lies the Purpose and Principles (i.e. RMA Section 5) of the RMA itself, to which Regional Policy Statements and plans (and logically, resource consents or permits) are accountable. To implement the Act, regional councils are tasked with producing Regional Policy Statements, which may be supported with regional plans. Regional Policy Statements (RPSs) are high-level visions which contain i) objectives for achieving the principles of the Act within a region, and ii) policies which describe how they will realize those objectives. Table 5 compares examples of objectives and policies within an RPS with the rules and other methods found in a plan. In contrast to the high-level policies of RPSs, regional plans (henceforth plans) are detailed documents containing specific methods for achieving the objectives of an RPS (Consultant 2a, 2016). Table 5. Comparing the high-level objectives of a Regional Policy Statement with the detailed rules of a regional plan. Source: Environment Canterbury (2017, 2018). Regional Policy Statement Regional Plan Canterbury Regional Policy Statement, operative 2013 OBJECTIVE 7.2.1 Sustainable management of fresh water. The region’s fresh water resources are sustainably managed to enable people and communities to provide for their economic and social well-being through abstracting and/or using water for irrigation, hydro-electricity generation and other economic activities, and for recreational and amenity values, and any economic and social activities associated with those values, providing:  1. the life-supporting capacity ecosystem processes, and indigenous species and their associated freshwater ecosystems and mauri of the fresh water is safe-guarded;  2. the natural character values of wetlands, lakes and rivers and their margins are Change 1 (Selwyn/Waihora), operative February 2016, Canterbury Land and Water Plan  RULES 11.5.8 From 1 January 2017, the use of land for a farming activity in the Selwyn Te Waihora subregion is a permitted activity, provided the following conditions are met: 1. The nitrogen loss calculation for the property does not exceed 15 kg per hectare per annum; and 2. No part of the property is located within the Phosphorus Sediment Risk Area as shown on the Planning Maps; and 3. No part of the property is located within the Lake Area in the Cultural Landscape/Values Management Area; and 66  preserved and these areas are protected from inappropriate subdivision, use and development and where appropriate restored or enhanced; and  3. any actual or reasonably foreseeable requirements for community and stockwater supplies and customary uses, are provided for. POLICY 7.3.1 Adverse effects of activities on the natural character of fresh water  To identify the natural character values of fresh water bodies and their margins in the region and to:  1. preserve natural character values where there is a high state of natural character;  2. maintain natural character values where they are modified but highly valued; and  3. improve natural character values where they have been degraded to unacceptable levels… 4. The practices in Schedule 24 are being implemented and the information required is recorded in accordance with Schedule 24, and supplied to Canterbury Regional Council on request.  The primary mechanism for regulating environmental effects in New Zealand lies in the issuance of resource consents (OECD, 2017b). Within a plan, councils can classify named activities as permitted, controlled, discretionary, restricted, or non-complying. If an activity is permitted, then no resource consent is required. If an activity requires a consent, then applicants must follow the designated procedures to acquire a consent. When applying for a consent, applicants must provide an assessment of environmental effects stating what efforts will be taken to avoid, mitigate or remedy adverse environmental impacts (Palmer, 2013). Councils can place conditions upon resource consents or even upon permitted activities, directing applicants to undertake specific practices in relation to reducing effects on the resource. For freshwater, resource users must consult plans and potentially acquire consents if their activities involve i) the extraction of water (RMA Section 14), ii) any activities on the beds of rivers or lakes (RMA Section 12), or iii) discharge pollutants into water or soil (RMA Section 15). 67  In addition to praise for its focus on environmental effects, the RMA has been positively evaluated internationally for creating multiple avenues for public participation in the planning process (Peart 2007). Councils are required to publicly notify all RPSs, plans, and RPS/plan changes, as well as some types of resource consents. When these are notified, any person may submit feedback on the proposal (RMA Section 96). Further, any decision made by councils about an RPS, plan, RPS/plan change, or resource consent can be appealed to the Environment Court, where greater scrutiny is placed on the evidence base for environmental effects, and more financial resources are needed (Palmer, 2013). In Environment Court, submitters and/or resource consent applicants can challenge the chain of logic spanning from the RMA principles through to an RPS, plan, rule, or consent: ‘So with the RPS you can say “this policy won’t achieve the objective”, “this policy is inconsistent with the objective”, “this policy is inconsistent with Section 5 of the RMA”, “this policy doesn’t get the balance right”, “this policy doesn’t give effect to the principles is sections 6, 7 and 8 [of the Act]”’ (Consultant 2a, 2016). Finally, and most important for this study, from its inception the RMA has assigned power to the Minister for the Environment to produce National Policy Statements (NPSs) and National Environmental Standards (NESs). These national regulations would sit above RPSs and plans, where RPSs must be amended to ‘give effect to’ any NPS or NES (RMA section 55(2)(b)). The Minister can notify an NPS or NES as a national regulation without having to pass a law through parliament. For the first two decades of the RMA (prior to May 2011), only a single NPS was notified – the Coastal Policy Statement 1994 (updated 2010), and three NESs (air quality 2004, drinking water 2008, electricity transmission 2010). 2.5.3 Explaining degradation of freshwater ecosystems under the RMA: toward a sociology of environmental planning How is it that New Zealand’s freshwater ecosystems have been able to degrade, given the strong principles of sustainable management embedded in the RMA 1991? Has the RMA 68  failed, and if so, how? There has been no shortage of observations of and explanations for the ‘failure’ of the RMA to protect the environment, and freshwater deterioration has often figured as a prominent example of such failure (see e.g., EDS, 2007; Turnbull Group, 2009; Office of the Auditor-General, 2011; Palmer, 2013; MartinJenkins, 2015; EDS, 2016). A 2007 survey of experienced resource management practitioners (n=70) by the non-partisan Environmental Defence Society reported that 40% of respondents felt that the RMA was not performing satisfactorily (Peart, 2007b), and an intensive follow up investigation in 2016 (n=48) reported that 38% felt that the RMA has not achieved its environmental goals, with another 60% feeling that the RMA had only partly achieved them (EDS, 2016: 42). A litany of complaints and analyses have been put forward to explain this situation. Here I focus on two. Unforeseen costs of implementation Practitioners have noted that unforeseen costs and delays in the RMA planning process have caused the actual implementation of the RMA to take many years. Consider that both RPSs and plans tend to go through processes of initial public consultation, followed by notification of a draft, followed by a submission period, and then a notified decision, which then can receive multiple appeals through Environment Court before becoming operative. Given that RPSs contain high-level policies whereas plans contain detailed rules, practitioners describe RPSs as ‘relatively ineffective’ documents (Consultant 2a, 2016), adding a costly layer of public submissions, hearings and court processes that provide very little certainty about environmental or economic outcomes. Plans, rather than RPSs, are the mechanisms that directly influence environmental effects, and yet plans are not even required by the RMA (MfE official 2a, 2016; Heitzmann, 2007).  The RMA’s strong provisions for public participation has seen Māori, industry, and environmental groups battling in Environment Court up and down the country, litigating RPSs, plans, and RPS/plan changes in their attempts to secure environmental and economic outcomes. These participation provisions have also meant that the cost to participate in legal 69  processes – as opposed to simply submitting texts – is high in terms of time, financial resources, and expertise. This has resulted in a sense of ‘consultation fatigue’ (Heitzmann, 2007: 159), as ‘elements of discursive democracy introduced by the legislation have subsequently been impaired by technocorporatist legal formalism’ (Jackson and Dixon, 2007: 107; see also Oram, 2007).  This costly structure of legal process means that there is less resourcing available for monitoring and enforcing plan rules and consent conditions. In 2014/15, regional councils designated 47,535 consents as requiring monitoring, of which only 49% were monitored. In Canterbury, where water extraction and pollution are perhaps the most severe in the country, only 47% of the 12,230 consents that require monitoring are actually monitored (MfE, 2018a). Regional councils lack resourcing to conduct monitoring, and have not received any resources from central government for this task (Crawford, 2007; Brown, 2017). Additionally, Environment Court has historically lacked adequate resourcing, though this is improving. From 2500 cases pending in 2003, in 2016 there were only 384 (Heitzmann, 2007; Registrar of the Environment Court, 2016). For Māori, capacity constraints have been a major barrier to substantive participation in resource management (Crawford, 2007).  5.3.2 Ignoring cumulative environmental effects? Despite the RMA providing a foundation for explicitly identifying and preventing deleterious environmental effects, New Zealand’s freshwater ecosystems have been drawn down (sometimes disastrously) and polluted past visible ecological thresholds. There are three broad explanations for this situation. The first is that resource consents, and their associated assessment of environmental effects, are processed on a case-by-case basis, and as such it has proven difficult to reject any single consent based on ‘significant’ environmental effects. As one veteran practitioner put it, ‘the way in which the Act is constructed, you cannot turn something down because of the incremental change that arises from the additional amount of 70  resource use from that. Unless you set a limit, you can’t turn it down’ (MfE official 2a, 2016). A consultant recalled that: Federated Farmers very successfully ran a line that... ‘we don't believe that its us, we don't believe that you've got the right solution to this problem’ or ‘you can’t demonstrate the linkage here, that this solution will fix this problem’. They used the effects-based narrative and the need for proof as a justification for not regulating farms. (Consultant 2a, 2016) Since financially-powerful consent applicants could afford to hire top scientists to argue that their proposal was not going to significantly degrade the ecosystem, this led to what Peart (2007a) calls a ‘mitigation mentality’, where council consent staff would tend to approve consents so long as efforts were made to mitigate, remedy, or avoid negative environmental effects. This mitigation mentality came to dominate and crowd out an alternative interpretation of the RMA, which could have seen the Court instead place burden of evidence on resource applicants to prove that their activity would not harm the ecosystem in significant way (Consultant 2b, 2016). A second part of the cumulative effects problem is that in order to reject a consent due to the crossing of a significant environmental effect threshold, a maximum level of resource use had to be constructed scientifically. This concept of an ecological threshold or ‘limit’ is now understood to be a core part of governing for cumulative effects, and yet there was very little scientific or policy guidance on how such a concept might be operationalized (Consultant 2b, 2016; MfE official 2a, 2016). Practitioners from regional councils bemoaned that such guidance ‘has been missing for a long time’ (Hawkes Bay planner 2a, 2016). Thus, a lack of scientific and planning tools to regulate with limits has also contributed to the mitigation mentality. 71  A third aspect explaining the ineffective protection of freshwater from cumulative effects under the RMA lies in what the Environmental Defence Society and others have called ‘agency capture’ (EDS, 2016). This refers to the dominance of agricultural interests within regional councils and the subsequent steering of council culture and priorities. This arises through two mechanisms. The first mechanism involves the steering of formal policies such as RPSs and plans by the electoral priorities of elected regional councils, who tend to be overrepresented by agricultural and industrial interests (e.g. see McNeill, 2016). In the processes of making RPSs and plans, internalization of agricultural interests by council staff has led to a strategic setting of objectives that specifically exclude effects of agricultural land use such as conversion to dairy or intensification of dairy farms. For example, in the late 90s and early 2000s, environmental NGO Ecologic took the agriculture-dominated Waikato Regional Council to Environment Court for failing to identify diffuse pollution from agriculture as a ‘discharge’ requiring a consent under the RMA (Consultant 2a, 2016). By not specifying diffuse pollution as requiring a consent, Waikato’s plan permitted massive blocks of land in the region to be converted from exotic forestry to dairy farming, resulting in major nitrogen, phosphorous and E. coli increases into the regions waterways (Waikato Regional Council staff 2a, 2017). Another mechanism of agency capture lies in the steering the operational functions of councils by politicians as well as by bureaucrats internalizing agricultural interests (e.g. see Stewart, 2016). This can be seen in the selective interpretation, monitoring, and enforcement of a council’s own rules and policies. Studies by the Office of the Auditor-General (2011), MfE (2016a), and a major review by EDS (2016) found overreaching by elected councillors into the operation of the RMA functions of council, including the issuance of consents and the enforcement of consent conditions. Council bureaucracies have also been criticized for selective implementation of their plans in favour of agricultural interests. In an infamous recent example, Horizons Regional Council’s 2014 plan included rules controlling fertilizer use based on estimated effects from the nitrogen modeling tool Overseer® (see Chapter 5). 72  By the time Horizons’ plan became operative, the science involved in Overseer® had become more conservative, which meant that stronger restrictions on fertilizer use would be required to realize the plan’s environmental objectives. Horizons responded by not requiring farmers to abide by the stronger restrictions, so Fish and Game and the Environmental Defence Society took Horizons to Environment Court in 2016, and won, thus forcing Horizons to implement their own plan4. 2.5.4 Evaluating the RMA: diverse failures, but a common solution? Regardless of whether the RMA is flawed or its implementation is to blame, most agree that the resource management system is clearly not achieving ‘the sustainable management of natural and physical resources’ (RMA Section 5). Practitioners have identified multiple reasons for this: the significant costs of implementing the RMA, the uneven ability to marshal scientific evidence for RMA decision making processes, the lack planning tools to set and enforce biophysical limits, and the capture (or at least steering) of regional council functions by agricultural industry and development interests. For this diversity of problems, practitioners have offered an array of solutions, ranging from streamlining the RMA to reducing costs through to abolishing regional councils and replacing them with an independent Water Commission responsible for issuing consents and ensuring enforcement (Turnbull Group, 2009; LAWF, 2010). However, a common refrain from across the political spectrum has been that the failings of the RMA have been due in significant part to a lack of direction provided to regional councils by central government through the mechanism of a National Policy Statement (interviews, D. Parker, 2016; E. Sage, 2016; Former senior parliamentary official 2a, 2016; Hawkes Bay planner 2a, 2016; MfE official 2a, 2016; K. Hackwell, 2017; and see also EDS, 2007; Smith, 2007b; Rouse et al., 2016; Knight, 2018). RMA architects Tony Randerson (2007) and Geoffrey Palmer (2013) argue that the                                                  4 Wellington Fish and Game Council v Manawatu-Wanganui Regional Council [2017] NZEnvC 38 (21 March 2017). 73  implementation of the RMA was always intended to be guided by national direction from the Minister for the Environment, and they emphasise that such direction never came. Indeed, for the first 20 years of the RMA, only a single National Policy Statement was enacted – the Coastal Policy Statement 1994 (updated 2010). 2.6 Freshwater as a national policy project The collective critique of the RMA had converged around a specific policy mechanism – i.e. a National Policy Statement – as the most obvious and appropriate way to address the cumulative effects problem. Since 1999, successive governments have mobilized the mechanism of a National Policy Statement for Freshwater Management to selectively assemble, frame, and respond to public concerns about freshwater. This section outlines the origins of the National Policy Statement within Helen Clark’s centre-left fifth Labour government (1999-2008), before detailing the actions of John Key’s centre-right fifth National government (2008-2016, led by Bill English 2016-7). Tracing the contest over and evolution of this policy illuminates how the concepts contained within the NPSFM are interpreted in the context of wider state and societal developments. 2.6.1 Freshwater policy low on the agenda: the fifth Labour government, 1999-2008  The Clark Labour government has been praised for its social and economic policy, and for naming ‘neoliberalism’ as an object and enemy, all of which has been debated among leftists (Larner et al., 2007; Roper, 2015). A government staffer at the time recalls that ‘They had a big focus on social policy, and a bit of a focus on economic policy, but very little on the environment’, evidenced in part by the fact that the environment portfolio was given to the 20th ranked member in cabinet, Marion Hobbs (Former senior parliamentary official 2a 2016).  While the fifth Labour government did not enact a freshwater NPS during its three terms, it produced a draft NPS on freshwater management. During David Parker’s 6-week stint as 74  acting Environment Minister in 2007, he recalls feeling that ‘MfE was just completely at sea with what to do with freshwater quality.’ Parker contracted a solicitor to draft an NPS on freshwater quality, instructing that ‘at heart… it needed to prescribe a minimum standard for freshwater bodies that was a swimmable standard’ (interview, D. Parker, 2016). Months after giving back the environment portfolio, Parker recalls that ‘all of the forces of darkness on water quality from other government departments came out and relitigated what was in that document’. Parker says this required him to expend political capital to draw cabinet’s attention to the importance of the original proposed provisions. Eventually cabinet agreed on a compromise text and submitted the draft NPS to a Board of Inquiry. The RMA allows government to submit National Policy Statements to an appointed panel called a Board of Inquiry for independent advice. This Board of Inquiry was chaired by retired Environment Court judge David Sheppard. The Board reported back in 2009 and published its recommendations in 2010, well into John Key’s fifth National government (Board of Inquiry into the Proposed National Policy Statement for Freshwater Management 2010). Parker recalls: [T]he Judge Sheppard National Policy Statement… effectively said ‘clean water bodies ought not to be allowed to get dirtier, dirty water bodies ought to be cleaned up over a generation. In order to effect that change, until you’ve got adequate rules in plans to control things getting worse, you need to have - through the National Policy Statement - a rule inserted into all plans that says that increases in land use intensity should no longer be a permitted activity.’ And he was right, it was a very elegant solution. (interview, D. Parker, 2016). A consultant at the time recalls the politics of this manoeuvre: Part of the point of it was to put in train this process with a very good Board of Inquiry that was going to then turn around and say ‘actually we need to strengthen all 75  these different parts of the NPS’. Labour then doesn’t collect all the flak for having proposed this stuff in the first place prior to the election, but they set in train a process through which a Board of Inquiry is going to recommend some much more rigorous stuff. Whether that’s what they intended or not, that’s what happened. The Board of Inquiry recommended a really good NPS. (Consultant 2a, 2016) The draft NPS addressed the problem of cumulative effects in part by bringing land use intensification more explicitly within the responsibilities of councils. It would have required all land owners who wish to intensify their land use (and associated pollution) to apply for a resource consent (i.e. permit). In regions such as Waikato, some polluting land uses were previously classified as permitted activities which meant land use could be converted to these activities without explicit permission. The draft NPS closed this loophole by not allowing regional councils to classify polluting activities in this way. Perhaps unsurprisingly, the Board of Inquiry-recommended NPS found favour with environmental groups such as Forest and Bird (interview, K. Hackwell, 2017) and the Green Party (interview, E. Sage 2016), but caused ‘a great deal of anguish’ for primary sector organizations (Land and Water Forum secretariat 2a). 2.6.2 Whole-of-government freshwater reform: the fifth National government, 2008-2017 In contrast to Labour, John Key’s fifth National government invested significant resources into freshwater policy development. A government staffer at the time remembers the prioritization of water as a constellation of electoral and economic Realpolitik: Through that period you had the growing dirty dairying concern, and the Canterbury irrigation situation was developing on both sides of the ledger. You had very high dairy prices in 2007 and 2008… that kept New Zealand from a sovereign debt crisis. So the incoming National government really clearly understood that water, 76  irrigation, Canterbury, [were] driving the economy: very, very valuable. And obviously with the growing public concern, they could see that they were going to have to manage the politics of it pretty carefully. (Former senior parliamentary official 2a, 2016) The freshwater policy arena has been aflutter with activity since 2008. The Land and Water Forum was created to guide the whole-of-government reform of freshwater regulation, multiple National Policy Statements have been issued, and other regulatory changes and investments in and around environmental policy have been made. The following sections trace key political developments and conflicts arising during the time periods leading up to the 2011, 2014, and 2017 versions of the National Policy Statement for Freshwater Management. By looking at each iteration of the policy and the milieu in which it was enacted, we can see which public concerns have been incorporated by the state into policy, and how. NPSFM 2011: forcing regions to set limits  When John Key’s fifth National government was elected in 2008, incoming environment Minister Nick Smith (who has a PhD in geomorphology) had to respond to the Sheppard Board of Inquiry, and issue a decision on the draft freshwater NPS. As it happened, Minister Smith already had a plan to roll out a different mechanism for forming and legitimizing freshwater policy. It would be called the Land and Water Forum.  In the mid-2000s, prominent environmentalist, RMA-architect, and researcher Guy Salmon, with other researchers from Ecologic, undertook a comparative study of Swedish, Danish, and Norwegian environmental governance institutions and processes. Salmon returned home with a transformed understanding of where the RMA had gone wrong and what might be able to fix it (Salmon, 2007a). These Nordic countries provided models of ‘collaborative governance’, where multi-stakeholder groups of environmental actors could generate 77  consensus on policy architectures and priorities, and this consensus could provide continuity and certainty across successive national governments (Salmon, 2007a). In the context of regional councils’ failure to deliver environmental outcomes under the RMA, Salmon proposed that a national stakeholder collective could resolve some difficult distributive decisions and generate enduring policy solutions, taking the heat off central government and potentially resolving the legitimacy deficit of regional councils (Salmon, 2008). Salmon’s work influenced Nick Smith’s thinking (e.g. Smith, 2006; Smith, 2007b) and Salmon’s presentation at the 2008 Environmental Defence Society conference in Auckland has become a common point of reference in the environmental community as a key source of New Zealand’s collaborative turn in freshwater management. Consider, for example, that seven interviewees – including sitting members of parliament – explicitly mentioned Guy Salmon by name when discussing the milieu of New Zealand freshwater policy. At the 2008 Environmental Defence Society conference, an informal body was created to eventually advise government, and in 2009 this group was named the Land and Water Forum (LAWF). The LAWF was given terms of reference by Minister Smith to make recommendations (by consensus as much as possible) on the nature of freshwater management needed for New Zealand (see LAWF, 2010). The LAWF initially consisted of 58 organizations including iwi representatives, industry groups (e.g. Fonterra, Federated Farmers), environmental organizations (Forest and Bird, Fish and Game, the latter choosing to be an ‘active observer’ rather than a participant in consensus statements), local governments, infrastructure providers, university scientists, recreation groups, and others. Alastair Bisley, a former trade negotiator, was appointed to chair the LAWF, which met frequently over its first phase of work to deliver its 68-page Fresh Start for Fresh Water report to government in 2010 (LAWF, 2010). This report included 53 recommendations, covering everything from setting biophysical limits for waterbodies through to promoting collaborative governance arrangements and creating a national Land and Water Commission which would include an equal number of Crown-Māori Commissioners and would report to 78  the government. The 2010 LAWF report stands as a major turning point in New Zealand’s freshwater management regime; unlike the submission process under the RMA where organizations submit separate feedback to the government about a specific policy, the LAWF report comes with a broad seal of consensus across civil society groups. Minister Smith deferred his response to the Board of Inquiry until after the LAWF report had been published and analysed (LAWF, 2010: 61).  Parallel to the LAWF developments, a major legislative change in New Zealand’s environmental democracy requires discussion. Based on the argument that the regional council in Canterbury (Environment Canterbury or ECan) was not discharging its functions to effectively manage water, the fifth National government passed the ECan (Temporary Commissioners and Improved Water Management) Act 2010, which dismissed the elected regional council and replaced them with central government-appointed Commissioners. ECan did not have an elected regional council between 2010-2016, and since 2016 it has been in a transition period with half of the Commissioner slots given back to elected regional councillors, and a plan to have a fully elected council in the 2019 regional elections (Environment Canterbury [Transitional Governance Arrangements] Act 2016). The implications of this Act for freshwater management and environmental democracy in New Zealand have been critiqued by others (Thomas and Bond, 2016; Kirk et al., 2017), and are discussed at relevant points in this dissertation. In May 2011 the government responded to the LAWF recommendations by gazetting the National Policy Statement for Freshwater Management 2011 (New Zealand Government, 2011). Most significantly, the 2011 NPSFM requires regional councils to set enforceable water quality and quantity limits for all waterbodies within their jurisdictions, and create plans to achieve them. Crucially, it devolves responsibility or limit-setting to regional councils, rather than setting national standards for environmental protection. The LAWF’s very first recommendation to government in 2010 was that ‘Central government should define national 79  objectives for the state of our waterbodies and set an overall timeframe within which they will be achieved’ (LAWF, 2010: 1), yet such national objectives were notably missing from the NPSFM 2011. Instead, regional councils were directed to determine limits based on ‘local and national values’, where national values consisted of 19 generic topics such as irrigation, drinking water, historical heritage, recreation, and traditional relationships of Māori with fresh water (see Table 6). Additionally, while the LAWF recommended collaborative governance as a framework for water management, the 2011 NPSFM did not require a specific process for regional councils to engage with communities to set limits. The 2011 NPSFM thus presents a selective response to the LAWF recommendations and to the problem of limits and regional governance. The government also rejected the provisions recommended by the Labour government regarding a halting of intensification, and instead forced councils to set limits without the use of that mechanism. Councils were required to implement the policy by 2030, and the Minister for the Environment was required to seek an independent review of the policy by May 2016 (five years after gazetting the NPS). Six months after the NPSFM was gazetted, MfE published a 50-page implementation guide to aid in interpretation of the policy (MfE, 2011). NPSFM 2014: instituting national bottom lines In response to public outcry over the lack of ecological bottom lines to protect waterways, Nick Smith asked the LAWF to advise on how a limits-based regime might work with national bottom lines. LAWF’s 92-page second report and a 113-page third report elaborated the Forum’s consensus on what such a regime might include (LAWF, 2012a, 2012b). Responding to the LAWF’s first three reports, the Ministry for the Environment published a report titled Freshwater Reform 2013 and Beyond (MfE, 2013a), which outlined the government’s proposals for comprehensive reform to the freshwater regulatory regime. These proposals involved 1) promoting collaborative community planning processes (as recommended in LAWF 2010 and again in detail in LAWF 2012a), 2) instituting a National Objectives Framework to specify and enumerate national bottom lines for freshwater, and 3) outlining 80  central government support for helping councils shift to a limits-based regime for freshwater management, including the production of guidance by central government and investment into capacity-building across regional councils. The 2013 MfE report introduced a normative model for ‘planning with community’ which provided guidance for interpreting the government’s freshwater reform activities. In a cartoon replete with community members of distinct occupations and myriad land use practices, an ideal policy cycle is presented (see Figure 4). First, a community group comes together and looks up at a table of national ‘values’, picking one (e.g. ecosystem health). This value is then broken down into specific ‘attributes’ (such as periphyton or fish) which can be measured. After selecting ‘periphyton’ (i.e. algae/slime), the community then reflect on the current state of this attribute, and consider which land uses cause this state. They then ‘Decide on what limits need to be set and what management options are required’ (MfE, 2013a: 28) to achieve the desired state for that attribute. This would be guided by nationally determined ‘bands’ for different attributes (i.e. A grade, B grade, C grade, D grade). Only a band that is the same or improved for that attribute may be selected. The community then consider the costs and benefits of reaching the desired state of an attribute, and can either say yes or can rethink their chosen value and/or bands. The 2014 NPSFM gazetted in July 2014 gave force to MfE’s vision (New Zealand Government, 2014). While it contained several differences to the 2011 version, its primary contribution was the introduction of the National Objectives Framework. The National Objectives Framework prescribes maximum allowable states for two national values (ecosystem health and human health for recreation) that apply to each ‘freshwater management unit’ (see Table 6). For ecosystem health to be achieved, monitoring sites must report within specified numerical bands for attributes such as total nitrogen, periphyton, and dissolved oxygen. For human health, measurements of E. coli and cyanobacteria must be kept below certain numerical bands. Importantly, the minimum acceptable state for E. coli permitted a high (>5%) risk of 81  infection for activities involving immersion (New Zealand Government, 2014: 31). A list of eleven non-compulsory national values are contained in an appendix, which includes mahinga kai (customary food collection) and wāhi tapu (ceremonial practices)(see Table 6). In addition, the 2014 NPSFM also added a statement acknowledging ‘Te Mana o te Wai’ (New Zealand Government, 2014: 6), foregrounding the indigenous valuation of the intrinsic integrity (mana) of the water (wai) itself.   82  Figure 4. Planning with community diagram. Reproduced with permission. Source: MfE (2013a).   83  Table 6. Compulsory and non-compulsory national values in NPSFM 2014 (New Zealand Government, 2014). Compulsory national values Ecosystem health • Attributes measured for lakes: phytoplankton, total nitrogen, total phosphorous • Attributes measured for rivers: periphyton, nitrate (for toxicity), ammonia (for toxicity), dissolved oxygen Human health and recreation • Attributes measured: E. coli, cyanobacteria Non-compulsory national values (no attributes specified) Natural form and character – Where people value particular natural qualities of the freshwater management unit.  Mahinga kai (i) – Kai are safe to harvest and eat.  Mahinga kai (ii) – Kei te ora te mauri (the mauri of the place is intact).  Fishing – The freshwater management unit supports fisheries of species allowed to be caught and eaten.  Irrigation and food production – The freshwater management unit meets irrigation needs for any purpose  Animal drinking water – The freshwater management unit meets the needs of stock.  Wai tapu – Wai tapu represent the places where rituals and ceremonies are performed.  Water supply – The freshwater management unit can meet people’s potable water needs. Commercial and industrial use – The freshwater management unit provides economic opportunities to people, businesses and industries  Hydro-electric power generation – The freshwater management unit is suitable for hydro electric power generation  Transport and tauranga waka – The freshwater management unit is navigable for identified means of transport.  The meaning of the bottom lines included in the 2014 NPSFM need to be interpreted within the wider program of government activities. In particular, ongoing and massive investments in agriculture, the reorganization of public science to subsidize agricultural profitability, and the privatization of water all effect the interpretation of provisions within the NPSFM. First, while the LAWF work and freshwater reform agendas were in motion, in 2012 the Ministry for Primary Industries publicly advertised its goal to double the real value of agricultural exports from $32B to $64B by 2025, as a part of the whole-of-government Business Growth 84  Agenda. This number, along with the provision of major public funding by MfE for private irrigation schemes, has been cited by environmentalists as directly contradicting the objective of cleaning up waterways (e.g. Joy, 2015). Even the OECD has criticized this export policy as fundamentally contradicting the governments stated aspiration to improve water quality: It is unclear how the twin objectives of reducing environmental impacts and doubling primary industry exports in real terms will be achieved, and whether the government assessed use of finite freshwater resources and impacts on water quality before setting such objectives. (OECD, 2017b: 156-157) Second, alongside this agricultural strategy, then-Minister for Science and Innovation Steven Joyce reorganized the government funding of environmental research around ‘mission-oriented science goals’ (Joyce, 2012). Instead of funding environmental research on a project-by-project basis, the government would put most of its funding into a single $100M, 10-year project with an explicit aim ‘To enhance primary sector production and productivity while maintaining and improving our land and water quality for future generations’ ( Thus, the organization of science in New Zealand had been reconfigured toward subsidizing the continued profitability of agricultural industry. Third, as outlined in Section 2.3, in 2013 the government sold 49% many state-owned assets including Mighty River Power. The Waitangi Tribunal hearing leading up to this concluded that Māori had ‘proprietary’ rights and interests in water that were equivalent to those of the Crown, and that any assertion of property (e.g. as required for sale to occur) must first address and legally clarify Māori rights and interests in water. The Prime Minister and the government claimed the common law position that ‘no one owns water’ (Young, 2012) and proceeded with the sale. Reading these developments alongside in relation to the addition of Te Mana o te Wai into the NSPFM, it appears that with one hand the government added symbolic language into freshwater policy recognizing Māori relationships to water, while 85  with the other hand it refused to recognize Māori rights and interests in water as declared by the Tribunal (Ruru, 2013). The 2014 NPSFM responds to the perceived lack of environmental protection of the 2011 version by instituting national bottom lines through the National Objectives Framework. This fundamentally centralized control over regional decision making processes by forcing all regional councils to articulate quantitative limits for the nine biophysical attributes specified in the new policy, as well as requiring compliance with the maximum allowable levels for these attributes in cases where they are exceeded. The National Objectives Framework is a selective response to the problem of nationally inconsistent environmental protection through the planning system, highlighting nine variables and requiring their measurement and management. The addition of Te Mana o te Wai within the policy document provides a further government response to the problem of Māori rights and interests in freshwater and its governance. At the same time, however, it was also refusing to acknowledge the proprietary interests of Māori in water. NPSFM 2017: the politics of swimmability The introduction of the National Objectives Framework spurred considerable debate across scientific and policy communities. Several prominent freshwater scientists agreed that the Framework represented a significant step forward in securing environmental bottom lines, but there was dissatisfaction with the substance and ambition of the Framework. Concerns were raised that: • the bottom lines were far too permissive for humans and non-humans,  • important ecological variables such as water temperature and sediment were not included, • biological indicators were needed in addition to chemical measures, • wetlands and estuaries were not protected, and 86  • improving ‘overall’ water quality ‘within a region’ allowed for spatially concentrated degradation (Science Media Centre, 2014). Outside of government, environmentalists latched onto the language describing the E. coli bottom line in the 2014 NPSFM. Waterbodies complying with the E. coli bottom line of 1000 bacteria per 100mL of water were considered suitable for ‘activities with occasional immersion and some ingestion of water (such as wading and boating)’ (New Zealand Government, 2014: 31). The 2014 NPSFM thus became synonymous with setting the national standard at the level of ‘wadeability’, which was clearly not aspirational and which went against many New Zealanders’ understandings of New Zealand as a clean and green environment, economy and nation. A grassroots group called Choose Clean Water campaigned across the country for a swimmable aspiration for water. They pressed the Minister in public forums and the popular press, rallying environmental commentators and raising public awareness, culminating in a 12,000-signature petition that was delivered to parliament in 2016 alongside thousands of blue paper drops in the shape of a river, representing the signatories (Burry, 2016). Related campaigns from the Green Party, public intellectuals such as Dame Professor Anne Salmond (2016), and a growing chorus of critical commentary from across business, science, and policy communities, collectively pushed the Minister to becoming ‘open-minded’ about clarifying that the NPS would lead to increased swimmability, though he felt the existing NPSFM 2014 would already have led to more swimmable waterways through its requirement to ‘maintain or improve’ freshwater quality (Mitchell, 2016). In February 2017, the government proposed amendments to the 2014 NPSFM (MfE, 2017a), with a press release boldly promising ‘90% of rivers and lakes swimmable by 2040’. Upon scrutiny, observers noted that that the way the government classified swimmable rivers had changed, and environmentalists accused the government of chicanery (Small and Mitchell, 2017). The Ministry had developed a complicated statistical procedure in which the median E. 87  coli count of a river was related to the amount of time that it exceeded the attribute bands of the NPSFM 2014. Swimmability was no longer measured by comparing a single number (median annual E. coli) to another number (the bottom line); the bottom line had been redefined to integrate a temporal weighting (Hansford, 2017; Webster-Brown, 2017). NIWA scientists calculated how many waterways would be designated as swimmable under the existing 2014 versus proposed 2017 bottom lines, and found that under the 2014 standards only 30% of waterways would be considered swimmable, compared to 43% under the new time-integrated grading system (Gudsell and Bramwell, 2017). Thus, the government had made 13% of the country’s waterways swimmable overnight through definitional fiat.  In addition to criticizing the level of the swimmability standards, environmentalists have also criticized how the swimmability targets exclude small streams (interview, K. Hackwell, 2017; public meeting notes, N. Smith, 2017). In a public meeting in Wellington, Green MP Catherine Delahunty argued that the swimmability maps produced by government are only for two variables (E. coli and cyanobacteria), and they do not include many things that affect swimmability such as sediment. Minister Smith conceded that other variables do affect swimmability, but argued that the new swimming maps provide members of the public with more information to make decisions about how much risk they are prepared to live with (N. Smith, public meeting transcripts, 2017). Indeed, the government intended to support a shift to swimmability by publishing the most-up-to-date information on E. coli and cyanobacteria to the public through a web-based interactive map (MfE, 2017a). The proposals for swimmability, seen in this light, are about more than just targets of bacteria levels. They are also about constructing individualised swimming ‘subjects’ who are intended to become responsible for the risks of swimming that the maps describe. In March 2017, the day after Nick Smith presented his policy proposals to a public meeting in Wellington, the Choose Clean Water campaign ran a billboard in central Wellington critical of the more permissive swimmability definition. The billboard featured a picture of a child standing over water and 88  looking into it, with bold text: ‘The government thinks you won’t notice more poo in your water’ (Cann, 2017). The 2017 NPSFM gazetted in August 2017 compels regional councils to monitor and publicly report on the swimmability of waterways in their regions. It requires them to provide a plan for how they will increase the swimmability of waterways in their region to 80% by 2030 and 90% by 2040 (New Zealand Government, 2017). The new time-integrated swimmability monitoring applies to all rivers that are longer than 1500m and at least 40cm in depth. In addition, some changes have been made clarifying avenues for Māori participation in decision making, and new clauses have been added specifying that nitrate and phosphorous levels for algae blooms should be taken into consideration (but are not included numerically). Further, a biological indicator, the Macroinvertebrate Community Index, has now been made compulsory for councils to measure and report upon, and if a stated numerical threshold is breached a regional council must prepare a remedial action plan. The contributions of the 2017 NPSFM for limit-setting and Māori participation are summarised in Table 7 and compared to the 2011 and 2014 versions of the policy.   89  Table 7. A summary of three versions of the National Policy Statement for Freshwater Management.  2011 National Policy Statement for Freshwater Management - limits 2014 National Policy Statement for Freshwater Management – ecological bottom lines 2017 National Policy Statement for Freshwater Management - swimmability Primary function Requires regional councils to set ‘enforceable water quality and quantity limits’ (p3) for waterways, with councils ‘imposing conditions on discharge permits to ensure the limits… can be met’ (p6). The aim of these regulations is to ensure that ‘The overall quality of fresh water within a region is maintained or improved’ (p6).  Councils must identify overallocated waterbodies and propose methods to phase out overallocated situations. Establishes the National Objectives Framework. This includes numerical thresholds for nine chemical attributes. No attribute in New Zealand may fall below the national bottom line. E. coli threshold set for ‘wadeability’ at 1000 bacteria per 100mL.  Compels regional councils to monitor and publicly report on the swimmability of waterways in their regions, and provide a plan to government for how they will improve the swimmability of waterways in their region to 80% by 2030 and 90% by 2040 (New Zealand Government, 2017).    Secondary function Such limits ‘must reflect local and national values’ (p3). A bullet-point list of national values is provided, including 11 uses such as domestic drinking water, electricity generation, and irrigation, and a second list of ‘water’s intrinsic values’ (p3) including healthy ecosystem processes, cultural relationships of Māori with freshwater, among others. Numerical attribute bands ensure that water quality is ‘maintained or improved’, as any attribute must not fall into a lower numerical band.  Standardises local planning processes by requiring numerical limits for specified attributes. Clarifies that ecological effects of nitrogen and phosphorus in addition to human toxicity need to be considered in the limit-setting process.  Macroinvertebrate Community Index must be measured for all freshwater management units, and regional councils must ‘seek to improve on a Macroinvertebrate Community Index score if it is below 80’ (p21). 90  Māori rights and interests Regional councils must ‘take reasonable steps to…  • involve iwi and hapū in the management of freshwater…  • work with iwi and hapū to identify tangata whenua values and interests... and…  • reflect tangata whenua values and interests in the management of, and decision making regarding, fresh water and freshwater ecosystems in the region’ (p10). The NPSFM now contains a statement acknowledging ‘Te Mana o te Wai’ (p6), foregrounding the indigenous valuation of the intrinsic integrity (mana) of the water (wai) itself.   Māori values such as mahinga kai (customary food collection) and wahi tapu (ceremonial practices) are added to the list of non-compulsory national values.  No change to 2014 version. Parallel 2017 changes to the RMA allow iwi to initiate ‘participation agreements’ between iwi and councils 91  Ultimately, the 2017 NPSFM responded to the accusations of weak environmental standards in the 2014 version by forcing regional councils to measure, report upon, and improve swimmability in their regions. Instead of raising the national ecological bottom lines, the government opted for a ‘report and improve’ approach focussing on swimmability as a specific culturally valued activity, and in response to a public concern that was carefully curated and boosted by environmentalists, public intellectuals, and elements of the private sector. Nevertheless, the strategic selection of variables used to represent swimmability has been fiercely contested. The state has also sought to patch up earlier gaps in the National Objectives framework by adding another mandatory ecological indicator (the Macroinvertebrate Community Index) and by adding an explanation to the nitrogen and phosphorous attributes. Through these amendments, the government has further clarified and elaborated its version of the freshwater ‘problem’ requiring implementation by regional councils. 2.7 Selecting and framing public concerns: five keywords in New Zealand freshwater policy While an environmental policy may introduce novel concepts into the environmental governance landscape, such concepts must be recognizable to the communities tasked with implementing them. Even radical concepts must have a history, and this history will affect their interpretation. Previous sections have sketched the colonial, scientific, political-economic, and regulatory dimensions of the freshwater ‘problem’ in New Zealand. These problem-dimensions are woven through the language of the National Policy Statement, and have been codified through the strategic and selective use of specific keywords. Keywords are used to gesture out to a broadly accepted public problem, while also collapsing that problem into a specific defined entity. For example, the problem of cumulative effects and the need for national direction has been codified within the National Policy Statement through the language of limits, the enactment of the ecological bottom lines in the National Objectives Framework, and broad reference to ‘local and national values’. These keywords can be 92  understood as the state’s strategic and selective portrayal and response to wider public concerns. In this concluding section, I discuss five prominent policy keywords within and surrounding the National Policy Statement that have emerged to organize conflict around freshwater politics. The NPSFM codifies complex public problems – of colonialism, ecosystems, economy, and regulation – through the keywords of ‘rights and interests’, ‘limits’, ‘collaboration’, ‘local and national values’, and ‘swimmability’. Through these keywords, the state both accepts the existence and importance of these public problems, while at the same time defining these concerns in ideologically selective ways. These keywords have come to provide touchstones for debate in freshwater politics, and their precise meanings carry significant implications for the relationship between state policy and wider public problems. Empirically, these keywords act as a useful guide for readers to help decode the concepts and language of New Zealand freshwater politics in the chapters that follow. State and non-state actors vie for authority by mobilizing these keywords as proxies for the wider public concerns described in this chapter. Theoretically, these keywords provide a way of grasping the incipient structure of the institutional void within New Zealand freshwater policy. The NPSFM is a strategically selective state response to the various public problems of freshwater outlined in this chapter. Historical analysis allows us to unpack the interpretive context surrounding a new policy concept such as ‘limits’, situating the keyword within the context of the (contested and selectively formulated) problem that it has been intended to address.  Māori rights and interests in freshwater have been an object of legal contestation and uncertainty since the Treaty of Waitangi was signed in 1840. Though the RMA contained several provisions for iwi involvement in environmental management, often these were vague, and they ultimately hinged upon goodwill within council staff rather than being understood as a required part of a councils’ core business (Crawford, 2007; Robb et al., 2015). Increasing iwi involvement in environmental management has been enabled more effectively 93  through the Treaty settlement process, as this carried with it resourcing to hire and train iwi resource management staff (Williams, 2007; Ruru, 2018). The sale of state-owned assets in 2013 forced the government to reassert its common law position that ‘no one owns water’, while acknowledging that Māori have special rights and interests in water that require recognition by the Crown. The LAWF advocated for further specifying Māori ‘rights and interests’ in freshwater as a matter of priority for the freshwater reform program (LAWF, 2015), and the Ministry for the Environment has described Māori ‘rights and interests’ as an ongoing priority (MfE, 2013a, 2017a). Through the 2011, 2014, and 2017 NPSFM as well as 2017 RMA reforms, further provisions for Māori participation in freshwater management have been added, though these do not address the Tribunal’s ruling that Māori have proprietary interests in water that are equivalent to economic ownership. In 2016, the Māori Council took another case before the Waitangi Tribunal to demand the recognition of proprietary rights guaranteed by the Treaty (Forbes, 2016; Williams, 2016). The nature of proprietary and non-proprietary Māori rights and interests in freshwater continue to be a major feature of freshwater policy discussion.  The contemporary idea of environmental limits is situated within a longer history of debate about the failure of the RMA and regional councils to manage cumulative environmental effects. The 2011 NPSFM introduced this concept into regulation, forcing regional councils to set numerical limits to resource use to protect water and its ecosystems. As noted in the Ministers’ foreword to MfE’s Freshwater Reform 2013 and Beyond, ‘this precious natural resource offers us significant potential for economic growth. But this will only happen if we use and manage water carefully within environmental limits’ (MfE, 2013a: 5). If New Zealand’s environmental policy community can succeed in limiting the cumulative effects of diffuse pollution in freshwater, this will legitimate the state in its domestic context and secure New Zealand’s competitive positioning in the international arena, from the marketplace of agricultural produce through to rankings of ‘livability’ (e.g. see MfE, 2017b). In the 94  practitioner community, limits have enjoyed wide consensus yet loose definition and varied interpretations (MfE, 2017e).  The terminology of local and national values refers to the balance between local discretion and national prescription in setting freshwater limits. The 2011 NPSFM required councils to set limits that ‘reflect local and national values’ (New Zealand Government, 2011: 3), while identifying eleven national values in a broad qualitative way. This gave regional councils significant discretion to define objectives and organize monitoring and compliance frameworks at the local scale. However, the National Objectives Framework in the 2014 NSFPM included two compulsory national values (ecosystem health and human health) and required all regions to measure and achieve specified attributes for these objectives. The National Objectives Framework provides the foundation for realizing MfE’s (2013a) ideal ‘planning with community’ model (see Figure 4), wherein communities are expected to specify local limits for each nationally-designated attribute. After scientific criticism of the narrowness of the National Objectives Framework, the 2017 NPSFM added a biological indicator to the list, and some qualifying language about how existing nitrogen and phosphorous measurements should be interpreted. The 2017 NPSFM also forces regional councils to monitor and report on swimmability of rivers in a prescribed fashion. In these ways, what once began as a fully decentralized policy has slowly been recentralized, as more and more standardized requirements have been added to the policy. Early indications from the sixth Labour government are that they may further add and revise attributes to the National Objectives Framework, as well as providing a standardised template for ‘limits’ (Parker, 2018b).  Since the Land and Water Forum, collaboration has become a keyword in freshwater policy and environmental management in New Zealand (LAWF, 2010; MfE, 2015; Rouse et al., 2016; Sinner et al., 2017). Guy Salmon’s translation of Nordic experiences with collaborative governance into the New Zealand context have proved pivotal to mainstreaming a specific 95  vision of collaboration. The LAWF operationalized this vision for a specific purpose, assembling high-level political representatives from major sectoral interests and forcing them to produce a consensus on the direction for freshwater management in New Zealand. At the local scale, starting in Canterbury, regional councils took up the collaborative governance mantra with gusto. ECan, newly free from accountability to locally elected politicians, carved the region up into ten zones and created community collaborative groups to make recommendations on freshwater policy for each zone (see Nissen, 2014; Kirk et al., 2017). In Waikato, Hawkes Bay, Northland, and Wellington, and elsewhere, collaborative processes were set up to determine local objectives for freshwater for their regions (see Chapter 4). While the NPSFM does not require nor mention collaboration as a preferred mechanism of community engagement, the RMA was amended in 2017 to allow a tightly-defined version of collaboration to be granted special legal status. Finally, while the swimmability of waterways has always been a concern for environmental advocates, since the release of the ‘wadeable’ standard in 2014 the concept of swimmability has come to centre stage in public debate about freshwater. From this mounting public pressure, Minister Smith became ‘open-minded’ about declaring an aspiration for swimmability, and in 2017 the revised NPSFM included a target of reaching 90% of rivers being swimmable by 2040. Not only does the revised 2017 NPSFM aspire to increased swimmability; it also forces regional councils to measure and report on swimmability regularly. Nick Smith himself wanted to make rivers and lakes swimmable not only in a physical sense, but also in the sense of creating citizen-subjects who see and utilize their waterways for swimming through the maps and metrics provided by the state, while deciding how much environmental risk to accept as a personal choice. Swimmability has thus emerged in public discourse as a significant scientific object as well as a political aspiration.  While the metaphor of an institutional ‘void’ might conjure images of a terra nullius of policy meaning, even radically novel policies have histories that shape their interpretive 96  possibilities. Limits should address the cumulative effects problem, iwi rights and interests should address the Treaty of Waitangi concerns, collaboration should address the agency capture problem. Rather than treating the terrain of freshwater politics as either flat or as chaotic, policy keywords provide structure to political demands and policy debates. Such keywords encourage us to think about how state and non-state actors attempt to refer to existing public concerns while also shaping what those concerns include. This can involve promoting keywords from outside of the state (swimmability, collaboration), arguing for different interpretations of existing keywords (e.g. rights and interests), and contesting whether existing keywords are valid referents of the problems they are supposed to address. This ongoing promotion, critique, and refinement of policy keywords provides a way of understanding how state and non-state actors seek to structure and influence the meaning of new policy concepts in an institutional void. In the chapters that follow, I will explore how differently scaled contemporary policy actors are drawing on their unique capacities and roles to structure the interpretation of the NPSFM as it is translated into practice.  97  Chapter 3 – State action for water policy implementation: coherent rollout of an ideological vision? 3.1 Introduction Once the state has crystallized its objectives and methods into a policy text, what is there left for the state to do? Back in the days of ‘command and control’ regulation in Western liberal democracies, the central state bureaucracy might have then been responsible for implementing this policy as well (Benson and Jordan, 2017: 5). With the decentralization of policy implementation to local governments, however, the practices and outcomes of policy implementation are no longer strictly guaranteed to align with the state’s objectives. In response to this, many governments have undertaken a diverse array of non-prescriptive or ‘soft’ activities that selectively privilege certain methods and outcomes of policy implementation (Jarvis, 2012; Mol, 2016; Perkins, 2017b). Through the generation and circulation of guidance material, the provision of information, and the co-development of implementation programs with local governments, for example, the state bureaucracy can and does attempt to structure the institutional void of policy implementation. Understanding the nature of these reconfigured state activities – and elaborating on their material and democratic implications – is thus an urgent analytical and political task.  This chapter explores how New Zealand’s environmental state is attempting to influence the terrain of water policy implementation. What activities is the state undertaking to support water policy implementation, and what objectives are being sought through this process? In answering these questions, I will seek to make an empirical, analytical, and political contribution. Empirically, this chapter offers an account of the structure and function of New Zealand’s environmental bureaucracy, the Ministry for the Environment. Characterizing the Ministry’s activities around water policy will enable researchers and practitioners across the world to understand how New Zealand’s environmental state is set up, and how it might 98  differ from others, and what unique opportunities and constraints it might experience. Analytically, the chapter investigates whether and how the Ministry for the Environment’s activities are configured around a coherent ideological agenda to structure the institutional ‘void’ (Hajer, 2003b) of policy implementation. By examining the specific projects and programs of implementation support undertaken by the Ministry, I consider whether and how these activities relate to – and might be explained by – environmental neoliberalism, and what other logics may guide policy action. Politically, this chapter advances a pragmatic agenda for working with/in the state for environmental justice and sustainability. After examining the logics and practices of bureaucrats within the Ministry, I resist calls for environmentalists to abandon the state as a site of political struggle, and instead call for environmentalists to enrich their activist repertoires by considering the specific factions, organizing logics, and actually-existing intentions that shape bureaucrats’ activities in unique historical-geographical settings.  Section 3.2 proceeds by defining the environmental state and reviewing its study by critical social scientists. I argue that while ‘broad-brush’ studies of state action (Heyman, 2004) provide valuable foundations for understanding state logics and activities, they tend to reduce state intentionality to a singular, coherent, effective, and often malevolent logic, such as capitalism or colonialism. In contrast, fine-grained empirical studies of actual bureaucracies provide can characterize and explain state activities in ways that identify the discretionary and strategic agency of bureaucrats. Concepts and approaches drawn from anthropology, I contend, can support attempts to build an understanding of how politics within the environmental state actually works. Section 3.3 briefly recounts relevant aspects of my methodology and access to the Ministry for the Environment (MfE), and Section 3.4 outlines the history and administrative context of MfE as an organization within government. Section 3.5 describes the work of the bureaucrats tasked with supporting implementation of the NPSFM, and Section 3.6 situates this work within the broader activities of the Water Directorate, which is responsible for steering a whole-of-government approach to freshwater 99  regulation. Section 3.7 then scales up to examine the primary organizing logics and tensions of MfE, to think about how organizational practices are structured and how they change (or not). Section 3.8 deploys the dual lenses of constraint/agency to consider the relative capacity for action of differently situated state actors within the state bureaucracy. Section 3.9 concludes by revisiting the notion of the institutional void and reflecting upon the analytical and political utility of a fine-grained and broad-brush understandings of the environmental state. 3.2 Theorizing the environmental state This section defines what is meant by the environmental state and reviews pertinent research and debates about it. After defining the state, I outline broad-brush explanations of state behaviour based in political economy (see Heyman, 2004), and I consider the merits of a fine-grained empirical approach to analysing the politics of the state bureaucracy.  3.2.1 Defining the environmental state Sociologist Max Weber (1978) famously defined the state as a system of administration and law which claims legitimate interests over a population and territory, and, through wielding a monopoly of legitimate violence, can compel action. To police and enhance the welfare of the population, Weber argued that a powerful bureaucracy was needed. This bureaucracy would require a clear chain of command led by elected officials and use standardized protocols to ensure consistent administration of the law (see Heyman, 2004; Jessop, 2016). Much scholarship on the state – including the environmental state – draws a link from Weber to present (see e.g., Whitehead, 2008; Jessop, 2016). For purposes of this chapter, I define the environmental state as a system of government administration concerned with managing the biophysical environment and associated socio-environmental interactions (Duit et al., 2016). The precise form and functions of environmental states need to be situated spatially and historically. This involves positioning the features of a case against wider observations of 100  state transformation. Historically, Meadowcroft (2012) observes that across the industrialized West, state concern for the environment has both broadened in terms of the spatial remit of environmental problems under its mandate, and deepened in terms of the increasing complexity of science and regulation needed to manage these problems. The environment has become an increasingly central and integral priority for states (Jordan and Lenschow, 2010), and the environment has become a relatively permanent focus of political action and struggle (Meadowcroft, 2012; Whitehead, 2017). The spatial structures of environmental states are also being reconfigured through decentralization (or denationalization, see Mol 2016). This involves the ‘scaling down’ of state responsibility for environmental protection to local administrative levels, the ‘scaling up’ of regulation to transnational policy and industry networks, and ‘scaling out’ environmental responsibilities to civil society (Reed and Bruyneel, 2010; Cohen and McCarthy, 2015). Environmental bureaucracies are increasingly connected and reflexive international actors, drawing policy ideas from other states and international organizations (Duit, 2014; Jörgens et al., 2014). The regulatory practices and internal workings of environmental states are also changing. Mol (2016) observes that across the OECD, there is considerable variability in the size and scope of environmental bureaucracies, and that while many have been decreasing in staff numbers since the early 1990s, a few have remained stable, and some (such as the UK’s Environment Agency and New Zealand’s Ministry for the Environment) have increased in size. From previous use of top-down, command and control regulation (still evident in the US for Lejano and Funderburg, 2016), many environmental states are increasingly embracing market-based policy instruments (Lemos and Agrawal, 2006; Castree, 2011; Jordan et al., 2013), network-steering facilitation roles (Torfing et al., 2012; Klijn and Koppenjan, 2016), and non-interventionist forms of policy that rely less on compulsion and more on the production of guidance and incentives (Mol, 2016; Perkins, 2017b). As their regulatory approaches change, so too do their workforce capacities: Mol (2016) argues that 101  environmental bureaucracies have devalued specialist expertise in favour of mediation, communication, and management skills. 3.2.2 Explaining the state and its activities: broad-brush approaches Many scholars have used what Heyman (2004) calls broad-brush approaches to understand the diverse activities of states. Such approaches ‘characterize all bureaucracies… as having similar tendencies, modes of thought, and patterns of action’ (ibid: 490). For Marx and Engels (1848), the state’s activities can be understood as a mechanism to secure bourgeois (upper class) interests in capital accumulation. Political economic perspectives drawing on Marx contend that the state’s function is to ensure capital accumulation within its territory while avoiding a crisis of legitimacy with its population (McCarthy and Prudham, 2004; Castree, 2008b; Ioris, 2015). The environmental state, within this argument, functions to secure economic growth for the upper classes of society while translocating and disguising ecological degradation. To achieve this, it is argued, the environmental state makes nature legible to capital and markets (Robertson and Wainwright, 2013; Meehan and Molden, 2015; Parenti, 2015; Collard and Dempsey, 2017), violently appropriates land and resources for capitalists (Smith, 2007a; Bridge, 2014), reinforces class-power through privileging certain concepts of nature (Robbins, 2000; D’Alisa and Kallis, 2016), and distributes the ecological effects of capitalism in racist and unjust ways (Collard et al., 2016; Pulido et al., 2016).  Critical scholars have identified several distinctive changes in the modus operandi of environmental states across the world (especially the industrialised West), which can be linked to the proliferation of neoliberalism as an economic and public policy doctrine (Heynen et al., 2007; Himley, 2008; Castree, 2010; Collard et al., 2016; Perkins, 2017a). For these scholars, the neoliberalization of environmental governance is a coherent project of policy reform built around (Castree, 2010): • Private property rights 102  • Market-based institutions, e.g. commodification of nature • Roll-back of state responsibilities and capacities for environmental protection (including decentralization without resourcing) • Shift to private provision of public services • Creation and promotion of self-sufficient individuals  • Shift from prescriptive regulations to ‘soft’ policy tools such as incentive and voluntary schemes (Jordan et al., 2013; Perkins, 2017b) While theoretical arguments for characterising all of these changes together as neoliberalization have been elaborated and critiqued (McCarthy and Prudham, 2004; Castree, 2010; Collard et al., 2016), in practice the neoliberal script has become a common point of reference for scholars who observe one or several of these shifts in specific settings. The neoliberal script provides a way of connecting on-the-ground changes with powerful international networks and ideas about government, and identifying a coherent purpose behind otherwise disparate activities. In this way, the neoliberal script – in which the state acts to ensure economic growth and profit for some actors in society – provides a benchmark against which to analyse differences of experience (see Castree, 2008a). On the one hand, it is crucial to identify the patterns of state activities and how they tend to be associated with certain social and environmental effects (Castree, 2008a; McCarthy, 2012). In this, rigorous and ongoing critique of neoliberal environmental policies, economic ideologies, and their associated assumptions, remains an urgent task. On the other hand, such all-encompassing critiques have led to the complete rejection of the state as a potential vehicle for environmental and social justice objectives: Instead of seeing the state as a helpmate or partner, it [the environmental justice movement] needs to see the state as an adversary and directly challenge it (Pulido et al., 2016: 27) 103  Pulido et al. cite ‘industry capture’ (ibid: 17) of the state as a primary reason for the lack of state response to the concerns of the environmental justice movement, echoing the spirit of Marx and Engel’s (1848) conclusion that ‘The executive of the modern state is but a committee for managing the common affairs of the whole bourgeoisie.’ In this view, the state is fundamentally captured by the interests of capitalists, and state policies are a universal (if spatially heterogeneous) expression of this relation. Pulido et al.’s injunction to reject and oppose the state in its entirety gives us cause to reflect: have critical scholars succeeded in explaining the environmental state and fully delimited its political possibilities? 3.2.3 Fine-grained studies of bureaucratic practices Heyman (2004) argues that while broad-brush approaches are useful and indeed necessary, ‘A totally critical framework does not allow space for constructive struggle with and against varied bureaucratic factions’ (p491). An important complement to broad-brush analyses is what he calls particularistic or fine-grained studies of bureaucracies. Fine-grained studies of the internal workings of bureaucracies can map out how bureaucrats make decisions based on multiple personal, interpersonal, and organizational – as well as formal political – goals and values. ‘A workable ethnography of a power-wielding organization has to take into account the life of the organization itself as it translates upper-level mandates down to the level of everyday practices’ (Heyman, 2004: 489). By following the practices of bureaucrats as they interpret formal rules and official narratives, the aim is not to produce an evaluation of a ‘successful’ or ‘failed’ bureaucracy (Mosse, 2005; Hoag, 2011). Rather, the aim is to understand how bureaucratic action is shaped by formal and informal organizational metaphors, intra-organizational factions, professional identities, and personal values.  Analytically, fine-grained accounts help to understand how state activities are ‘caused’ by multiple forces, including – but not limited to – the political ideology of the Minister and government of the day. Politically, this opens up terrain of political struggle. Identifying the moments and qualities of bureaucrats’ discretionary practices can support a strategic ‘war of 104  position’ (Gramsci, 1971) as bureaucrats with sustainability and environmental justice values seek to transform both the state and civil society. As one example, through creative mobilization of organizational metaphors and polysemic keywords in the U.S. Environmental Protection Agency, ‘staff-level employees can and do craft counter-neoliberal content that will pass political review without objection’ (Robertson, 2010: 8). Understanding how bureaucracies work enables environmentalists to think about their engagements with/in the state differently. As Heyman (2004) contends, ‘Effective action anthropology demands insights into potential allies, enemies, coalitions, key metaphors, and ideologies’ (p 491). Fine-grained studies can help to reveal how the bureaucracy is ‘a life-world populated by actual buildings, specific objects and people with anxieties and dreams’ (Hoag and Hull, 2017). If broad-brush approaches have led us to prematurely rejecting the state as site of political struggle, perhaps fine-grained studies of the state can foster a nuanced approach understanding how the state and its interests are constituted in practice. This brief review identifies three sets of perspectives on the environmental state that can guide the analysis moving forward. First, the environmental state can be characterised spatially and historically by its administrative responsibilities and internal capacities. Second, through documenting a range of MfE’s activities in relation to water policy, we can ask: are these activities ideologically coherent, and (how) do these accord with global tendencies in neoliberal restructuring of environmental states? Third, in addition to broad-brush political economic explanations, are there more-than-neoliberal logics that underpin state activities, such as organizational and individual values?  3.3 Methods This chapter draws primarily from interviews, participant observation, field notes, and document analysis collected between February 2016 and March 2017 within New Zealand’s Ministry for the Environment (see Chapter 1). Over the forty working days I spent within MfE, I conducted 40 semi-structured interviews with 28 unique individuals, ranging from 105  analysts and senior analysts through to team managers, the Directors of the Water Directorate and past and present members of the executive leadership team (including the current chief executive). Interviews ranged from 16 to 80 minutes, with most around 50 minutes. In this chapter I use the term ‘officials’ to refer to any bureaucrat working at MfE. I use the term ‘manager’ to refer to team managers, directors, and executive leadership, and the term ‘analyst’ to refer to analysts and senior analysts. This distinction allows me to highlight differences between managers and analysts’ perspectives, while maintaining a level of anonymity for individual managers by classifying them together.  3.4 Ministry for the Environment: mandate, characteristics, and reputation New Zealand has a relatively independent state bureaucracy. Administrative heads of government departments are appointed by the State Services Commissioner rather than the government of the day, and public officials’ principal duty is to the New Zealand public rather than the government in office (Shaw and Eichbaum, 2018). Under the State Services Act 1988 and the Public Finance Act 1989, Ministers and chief executives agree on budgets and deliverables for their Ministries, but the administration and operation of the Ministry lies with the chief executive. The Ministry for the Environment (henceforth MfE) was created through the Environment Act 1986 to ‘advise the Minister on all aspects of environmental administration’ (Section 31a). While the Environment Act does not specify regulatory functions for MfE, under the Resource Management Act 1991 the Environment Minister has regulatory powers which MfE can be requested to discharge, such as reviewing regional council implementation of the RMA or writing national policy statements and national environmental standards. Under the Environment Reporting Act 2015 the Secretary for the Environment (chief executive of MfE) is co-responsible for publishing national State of the Environment reports. Importantly, since MfE’s budget must be negotiated with the Minister and signed off by cabinet, the size and scope of MfE’s activities are significantly affected by the priorities of cabinet and the ability of the Minister to secure funding for MfE’s work. 106  With 329 full-time equivalent employees in 2016, MfE is a medium-sized government department, with more capacity than small Ministries such as transport (123 employees) or defence (93) but considerably less than Ministries such as health (1095), education (2607) or primary industries (2413, see State Services Commission, 2016: 33). MfE has the second youngest average staff age of all government departments (39.6 in 2016) with strong female representation in the workforce (60.5%) typical of many government departments. It also has the third-lowest length of tenure (3.8 years) and the third-highest rate of annual turnover (24% in 2016, see State Services Commission, 2016). In broad terms MfE can be described as a relatively young, female, and pākehā (of European descent) Ministry.  Prior to the mid-2000s, MfE was largely considered an irrelevant entity in freshwater management, for reasons discussed in Chapter 2. Apart from publishing State of the Environment reports in 1996 and 2007, and producing technical guides and occasional policy discussion documents, MfE’s activities were not considered significant by practitioners within and outside of government. One practice that was recalled favourably was that, in the 1990s, MfE used to submit comments on regional plans as a way of contributing a national environmental interest perspective into local planning processes (interviews, Consultant 3a, 2016; K. Hackwell, 2017), but this stopped by the early 2000s. When water reform was prioritized by the cabinet of the fifth National government, reform involved a re-envisioning and resourcing of MfE to proactively provide policy support for the National Policy Statement for Freshwater management (NPSFM) 2011 and 2014 and to help accelerate implementation. In 2012 the Water Directorate was created within MfE to drive a whole-of-government approach to water policy reform. The Water Directorate involved a merger between freshwater teams of MfE and the (much larger) Ministry for Primary Industries, as well as secondments from Treasury (MfE manager 3a, 2016).  Through the increasing frequency of freshwater policy developments since 2009, MfE has become a significant actor and power broker in freshwater policy, and through this it has had 107  to confront two sets of challenges. The first is practical: MfE has been missing in action for decades on freshwater, and is having to build its relationships and reputation effectively from scratch. This task, as we will see, is partly affected by the youth, gender, expertise, and high turnover of MfE staff. The second challenge is public and political; in leading these freshwater policy processes – including providing the secretariat for the Land and Water Forum and modeling for Nick’s Smith’s swimmability targets (see Chapter 2) – MfE is seen also as responsible for the political outcomes of these processes. Although government Ministries are supposed to provide ‘free and frank advice’ to their Minister (see Shaw, 2015), MfE is increasingly perceived as a political actor in ideological allegiance with the Minister (interview, K. Hackwell, 2017). The value-laden choices, inclusions, and exclusions that MfE makes through these policy making processes are of significant consequence and deserve critical analysis. 3.5 Into the breach: the brave new world of policy implementation A first type of bureaucratic discretion is that of individual bureaucrats over specific budgeted projects. Here I describe the work of the Implementation team within MfE, which has been specifically charged with supporting regional councils to implement the NPSFM. By examining three sets of projects – relationship management, economic capability-building, and the NPSFM Review – we can see how bureaucrats have exercised discretion over the structure and substance of state projects in ways that do not simply align with a coherent rollout of the government’s neoliberal ideology. Crucially, while there are distinctively neoliberal elements within these projects, there are also other significant logics at play, and bureaucrats are creative in working within the constraints of their mandates.  In 2013 a group of officials were tasked with documenting and supporting the implementation of the 2011 NPSFM. Initially the team’s work involved surveying councils about their implementation progress and intentions, compiling information into a database that might eventually be mapped spatially (MfE official 3a, 2016). In 2014 a permanent 108  manager was hired for the team, who upon arrival initiated a significant redirection of the team’s approach and focus. A team member recalls: The work that had been outlined by the team was to pick some councils that were in the forefront (of implementation) and concentrate on developing their skills and learning from them. [the new manager] came from a council that wasn’t on that list and [the manager] suggested even at the interview for the job that that was not an appropriate way to take the councils of New Zealand with us. Because it was either picking winners - and what do you do about the rest? - or it was funding the haves, and what was the position for the have-nots? (MfE official 3b, 2016) This official emphasises an egalitarian approach to implementation that helps the ‘have-nots’ and not only those councils that conform with visions of implementation success. The approach developed instead involves a ‘relationship management model’, which was laid out in project documents and was well-understood by all in the team (Internal MfE documents, 2015, 2016; MfE manager 3b, 2016). Two team members were assigned to each regional council as ‘relationship managers’. Their relationships involve regular (at least twice-yearly) visits to each regional council, as well as regular contact via telephone calls, emails and interactions through involvement in MfE-supported workshops. In the initial phase of this work, the team undertook training in basic qualitative research (communications, marketing, how to ask open-ended questions) and asked councils ‘where are they going, what do they need, what are the stumbling blocks, what’s their plans’ (MfE official 3b, 2016).  Guided by a mandate to support councils and empowered by relative administrative discretion, the Implementation team pursued several projects under their mandate (MfE official 3b, 2016; MfE official 2c, 2016). The team organized ‘exchanges’ with regional council staff addressing the topics of collaboration, economics, and other technical issues in freshwater planning, to facilitate sharing of experiences (Internal MfE document, 2015). When I first arrived in February 2016, the team was eight people. By the time I left five weeks 109  later, one had been seconded to another department, and two new hires had joined. As of September 2018, only three of the original team members still work within MfE. This flux is not unusual, and it illustrates the wider problem of staff turnover in the Ministry (22.7% in 2017, MfE, 2017b: 35).  3.5.1 Relationship management – a new modus operandi for government?  We are trying to change the way the Ministry works with councils, and that will make a difference. (MfE official 3e, 2016) The relationship management approach provides a sense of identity for the Implementation team. MfE officials imagine their purpose as thoroughly bottom-up, wherein MfE supports the reflexive diagnosis of issues by councils themselves, and provides strategic brokering and knowledge-based resourcing. One team member notes: ‘We’re not into recommending a particular model or path, or expecting them necessarily to come together on one’ (MfE official 3b, 2016). Rather, ‘What we’re trying to do is to put in front of them some ways of doing it – economics, accounting whatever it is, collaborative planning – and get them to share what works and what doesn’t for each other so that they can then pick up… what they think will be best for their region’ (MfE official 3b, 2016). In addition to the regional councils, relationship management portfolios were created for national-scale sectors of primary industry, environmental organizations, and iwi. Because of the high stakes involved in building these relationships, the ‘primary’ relationship managers were all initially senior analysts, to provide stability (seniors stay longer at MfE) as well as increase the perceived authority of the MfE officials in their interactions with councils. Team members readily offer benefits of the approach. An initial benefit was the engagement of councils that had historically shunned MfE’s influence (MfE official 3b, 2016). According to one team member, ‘two… councils who have rather operated at arm’s length [from MfE]… are starting to see that there’s value in participation’ (MfE official 3b, 2016). Even though the 110  councils do not necessarily trust MfE, officials hope that by engaging in dialogue with other regions that this may lead to sharing and circulation of experiences, ideas, and perhaps, practices. For some, relationship integrity became a matter of personal pride, leading to strong ownership of relationships by individual officials and even protective behaviours (MfE official 3f, 2016). Further, the relative discretion afforded to the team meant that they would ‘almost 100% be able to turn around and be able to supply something that we’ve thought up that will meet [councils’] needs’ (MfE official 3d, 2016). As summarised by one official, The networking’s been brilliant… I thought we might be coming to our natural end because we’ve covered a lot of ground. And so I asked them [council staff] and they unanimously said… that they wanted to continue having them twice a year for two days and here are some ideas of things they wanted to hear about. So they are still getting value. (MfE official 3d, 2016) Relationship management has not been without its challenges, tensions and contradictions. For one, sometime projects fail. One team member was building a web-based platform for document-sharing amongst regional council staff (MfE official 3a, 2016), but when I checked in with this project later, it had dissolved because the web-based security requirements were too much of a logistical hassle for council staff to use easily. Another challenge is that even though councils may participate in exchanges to learn about MfE’s future policy directions, MfE officials have to guard such information very carefully, which places at least some constraint on their relationship-building (MfE official 3b, 2016). Perhaps most significantly, the Ministry-wide problem of staff turnover poses an acute challenge for the Implementation team, as they are the ‘outward facing team’ (MfE official 3g, 2016) of the Ministry on water issues. The team confront a legacy of recent negative MfE stakeholder perception surveys that all recommend greater continuity of contact persons at MfE (Internal project document, 2016). These challenges endure. For example, upon the transfer of emails from a previous colleague, one official recounts overlooking a single sentence at the bottom of an email, which 111  ended up having negative consequences for their relationship with that council (MfE official 3h, 2016). 3.5.2 Building economic capability  Section 32 of the Resource Management Act 1991 (RMA) was amended in 2013 to change how regional councils make and justify planning decisions. Where councils previously had to account for the benefits and costs of any plan change in generic terms, the 2013 amendments specified that costs and benefits must refer to the explicitly economic, environmental, cultural, and social effects of any proposed provisions. Further, any effects upon economic growth and employment opportunities must now be assessed and feature in decision making. Third, costs and benefits are now required to be quantified, where practicable (MfE, 2013b). When I joined the Implementation team in 2016, I was surprised to learn that a significant component of their work involved building economic capability with regional councils to support the new Section 32 regime (see Table 8).  Table 8. A snapshot of key projects in the economics capability-building workstream. Project Key features Economics exchange 2014-ongoing (as of 2016) • Twice-yearly workshops with regional council staff • Topics solicited from councils, e.g. S32 analysis, costing scenarios, ecosystem services, iwi values, infrastructure planning, limit-setting ‘Scalable’ economic assessments (2014-2017) • $2M of funding over four years for council-led projects ‘that in some way built their organization’s capability to do economic analysis… filling a local gap in knowledge and a national gap in knowledge’ (MfE official 3a, 2016).  • Included a modeling study of sediment in the Kaipara Harbour north of Auckland, and the uptake of farm management plans in the Horizons region.  • Officials manage contracts, seeking updates from councils, and facilitating the feeding back of results and learnings to other regional councils at the end of the projects (MfE official 3a, 2016). 112  Economics 101 workshop (2016) • One-day training course for council staff • Held in regions to increase participation • Course content produced by MfE officials, but ‘taught’ by an academic economist on secondment from a university • Plan to run course ~10 times by end of 2016 A national economics ‘tool’ (2016-17) • Idea successfully pitched as utilizing the ethos of ‘co-design’ promoted by the new MfE executive leadership team • Output is radically open – could be tool, framework, guidelines • Plan to i) Use open-ended ‘empathy interviews’ to learn council staff needs, ii) Iterate tool between MfE-councils, and iii) Fine-tune output through two-day workshop in Wellington The economics exchanges and the Economics 101 workshop can be read as elaborations of the team’s general approach to issue diagnosis and collective dialogue. For the Economics 101 course, officials report that ‘the thing that other people find really valuable is actually just being able to have that discussion about their skepticism with economics on the day… and have a bit of a dialogue about it’ (MfE official 3i, 2016). Officials adopted a bottom-up sense-making strategy oriented toward developing critical and practical competence rather than ideological alignment. They sought pragmatic examples in which economic reasoning might make environmental sense: ‘We’ve looked at farm-level analysis, upscaling that to a catchment... They’ve looked at mitigation cost curves so they wouldn’t be so scared of them anymore. And you know they go like this (demonstrates curve) and so we can get this much [pollution] mitigation… You know, don’t be scared of a graph because you’re not a mathematician’ (MfE official 3d, 2016). The economic assessment funding and the economics ‘tool’ projects were similarly built out of the economics exchange, and strengthened the team’s relationships with council staff. In the case of the economics tool, officials pitched the idea of trialling the concept of ‘co-design’ which the then-new (in March 2016) executive leadership of MfE were advocating across the organization. Here, officials constructed a mandate for a radically open type of output. ‘I have no idea what this tool’s going to look like’, noted one official, ‘we will do whatever [the council staff] want. It’s not even going to be called a tool, we need to decide what we are 113  going to call it. But a tool does sound like “is it going to be a framework, is it going to be a box, is it... going to be a spreadsheet?” So we will find a way of framing the question, the real question we want answered. How can we help you... what can we develop that would be of use?’ (MfE official 3d, 2016). While the ‘tool’ project could be open and inclusive of all councils, the economic assessment funding could only be disbursed to councils willing and able to put together a successful proposal. As an official clarified in 2017, ‘the regions that we are working with are the regions who have said “yes pick us, pick us, we want to work with you guys’” and most of them are quite well-resourced councils’ (MfE official 3c, 2017). Thus, despite the egalitarian intentions of the Implementation team, in practice there are organizational imperatives to fund those projects that already have the very capability that the state seeks to cultivate (see Mosse 2005). Across these projects, an unspoken concept of capability has been internalised by the Implementation team. Consider that regional councils have variable scientific and social scientific capacity in their organizations, as noted by MfE’s (2017e: 47-48) own review of NPSFM implementation. When I asked around the regions, one council official explained that what they really wanted from MfE was either i) a list of experts who could be contracted to undertake economic analysis, or better yet ii) an MfE-funded hub of consultants to do economic analysis for regional councils (interview, regional council official 3a, 2016). While the economics workstream projects may help planners and others to understand the role and utility of economic analysis, these projects do not increase the material resourcing for councils to employ economists or modelers. An Implementation team member explained that  You’re not allowed to fund what is basically council business as usual, right. Councils should have a certain level of professional development that they do for their staff… Having two or three economists in-house who then went out to councils would basically mean we'd be giving councils staff, and we wouldn’t have enough to help all the councils that need it. (MfE official 3e, 2016) 114  While the Implementation team have exercised agency and creativity in the substance and structure of their economic capability projects, these improvisations have been undertaken within a context that renders certain types of support as off-limits. This is an issue that we will encounter further below and throughout this thesis. 3.5.3 Juggling relationships and ethics with the NPS Review  The 2011 NPSFM signalled that ‘The Minister for the Environment intends to seek an independent review of the implementation and effectiveness of this national policy statement in achieving all its objectives and policies and in achieving the purpose of the Act’ (New Zealand Government 2011: 5). In early 2016 the Implementation team were assigned the task of designing and undertaking this review (henceforth the Review). Being asked to evaluate regional councils – with whom officials had now spent years trying to build trust – caused a fair bit of angst and uncertainty. The NPSFM was the centrepiece policy of the government’s freshwater reform, and to send any indication to the public that the policy had failed or was ill-conceived would not be a wise move for officials (who might be negatively evaluated by their managers) or for the Ministry (who might be negatively evaluated by its funder, the Minister). In the public sphere, MfE was trying to build name recognition as an independent environmental steward, however, so to align too closely with the perspectives of either regional councils or to the Minister could compromise their public reputation.  Based on the advice of a working group of senior regional council staff, the Review team decided to focus ‘more [on] the processes the councils are using than... if they are meeting the government’s outcomes’ (MfE official 3a, 2016). In this way, the Review could describe and compare council practices without providing a final evaluation of them against some normative criteria. ‘We are not going to pass or fail them, they just need to have something in place... there has to be a sign or some record of meaningful engagement of how you got the community's views. I’m prepared to be pragmatic, they just need to defend it’ (MfE official 3j, 2016). The decision to not issue a pass/fail verdict was also based on the limited view of 115  regions that the government could have from a distance. ‘We can’t or we shouldn’t sit here in Wellington and go “from what I’ve seen in my three visits and the odd person I’ve chatted to [that] you’re not doing a good enough job,” you know. We don’t have that level of information’ (MfE official 3j, 2016).  The team’s approach, then, involved documenting the formal planning procedures that have been undertaken in relation to the NPSFM – such as plan changes, collaborative processes, setting limits (see MfE, 2017e). Team members acknowledged that this narrower scope posed its own problems, but stuck to their decision:  If a council’s coming out with whole regime of managing water that’s not going to work for some technical reason or whatever (laughs)... Yeah you’re right, it’s not designed to find out that sort of stuff because I don’t think we can in the time that we have, or even that we are the right people. (MfE official 3k, 2016) The Review involved a standardised written questionnaire sent to every council, and a two-day site visit in which the Review team met with local politicians, council staff, iwi, and local stakeholders. The participation numbers demonstrate council capacity differences: for the West Coast, two council staff attended, while for Hawkes Bay there were ‘probably 12’ (MfE official 3k, 2016). Council staff attendance was procured through councils, while for iwi and stakeholders MfE had to rely on their own relationships and suggestions of councils, which the Review team acknowledged was a significant limitation (see MfE, 2017e). The Review teams made a special effort to connect with iwi through the Mana Taiao unit of MfE and other networks rather than through council suggestions, because they did not want councils to ‘cherry-pick’ those representatives for whom they have positive relationships to report on (MfE official 3k, 2016). The Review teams also met with national-level industry and environmental groups, again relying upon existing management networks.  116  The Review team asked participants about the character and quality of council engagement strategies, limit-setting processes, relationships with iwi, and other topics. Notes taken were circulated condensed into legible bullet-point format and circulated to all involved for fact checking and clarification. The Review team then went through an internal sense-making process to identify and select the 11 themes that feature in the published Review (MfE, 2017e).  The Review process threw up several challenges for MfE officials involved. The issue of network saturation was evident. When officials approached council staff and stakeholders, some were frustrated about the time commitment involved, and one complained that they had already contributed to a freshwater policy survey conducted by a scientist at the National Institute for Water and Atmospheric Research. Further, ‘at the same time we have the NPS amendments going on, so... they are all telling us the NPS keeps changing, they’re sick of it, and they want to be left alone to get on with it’ (MfE official 3k, 2016). Some councils were defending their plan changes in Environment Court at the time, and were concerned that acknowledging any non-ideal aspects of their planning process could be used against them in court arguments.  MfE officials also had to go through a careful process of sifting through the accounts they were presented and deciding how to represent them. As one official recalled, People don’t often sit around and tell you all of the good things that are happening. It’s the same as when someone comes back from a holiday, you know. They don’t tell you about all of the good times they had, they’ll focus on... that they got mugged or they had to pay a bribe here, or... their plane got delayed here because... that’s kind of interesting to people. I think we get a bit of that. (MfE official 3k, 2016) When the draft report was circulated, some stakeholders and council staff contested claims made by others and demanded proof. For example: 117  Someone from an NGO might say “I don’t care what the council says, this waterway is stuffed, and the cause of it is farming.” So we can say, “according to stakeholders this waterway is stuffed and it was caused by animals” and the council will write back to us and say “bullshit, where’s the evidence?” or “that doesn’t match up with our monitoring”’. (MfE official 3l, 2017) In one case, a council called the MfE Review ‘a debacle’, as the Review team had only interviewed a single environmental and iwi representative, both of whom gave unflattering accounts of the council (MfE official 3l, 2017). The published Review (MfE, 2017e) contains several caveats in response to these concerns, such as emphasising the incomplete coverage of stakeholders, and arguing that stakeholders’ views are valuable even without external validation. The Review team chose to defend against these concerns rather than retreat to producing a more conservative document, and the final report includes some strong evaluative statements (albeit not directed at councils by name).  The Review team pitched the report as ‘being critical of ourselves’ as much as – if not more than – being critical of regional councils (MfE official 3l, 2017). The quietness of the report’s publication and reception is telling. For such a significant piece of work and an important progress marker for the government’s freshwater reform program, I was surprised to see it receive no Ministerial press release or any news coverage. After asking around MfE, I learned that ‘higher ups’ within MfE were unhappy with the report and sought to prevent – or at least downplay – its release. The report got through relatively intact, I understand, through the virtue of its rigorous process and officials who were prepared to defend it.  The purpose, conduct, and content of the Review were actively shaped by bureaucrats and not simply a mechanical translation of the Minister’s political ideology. The Review experience shows how MfE officials have navigated their competing roles, trying to avoid any clean ideological or normative allegiance to either the Minister or regional councils. The Review produced multiple challenges for officials to overcome, and they did so in ways which, while 118  logical, were not the only legitimate choices available. Officials improvised within the constraints of their situation to create a distance (if not independence) from the Minister, they invited councils into setting the agenda of the Review, and made significant efforts to solicit Māori perspectives beyond their councils’ influence. They also enacted a concept of professional integrity that put them at odds with some councils and stakeholders, even though the Minister and ‘higher ups’ within MfE might have preferred a document with fewer criticisms of regional councils and central government. By examining the moments and contexts of discretion that bureaucrats encountered through the conduct of the Review, we can see how bureaucrats can engage in intentional forms of struggle to construct the structure and substance of a state project such as the Review. 3.6 The Water Directorate as the ‘war room’ of regulatory reform The projects of the Implementation team are undertaken within the larger unit of the Water Directorate, which has a distinct founding rationale, coordinating objectives, and division of labour. In this section I sketch the work of different teams across the Water Directorate to consider what this whole-of-government approach to water reform looks like, whether it amounts to a coherent ideological program, and how bureaucratic discretion both constitutes and directs this work. The Water Directorate (henceforth the Directorate) is a ~45-person unit (as of November 2017) within MfE. The Directorate was set up in October 2012 to help give effect to the recommendations of the Land and Water Forum’s second and third reports (LAWF, 2012a, 2012b). MfE reallocated personnel to translate the LAWF’s recommendations into specific policy proposals spanning various legislative, regulatory and non-regulatory actions (MfE official 3b, manager 3a, 2016). The Directorate brought together officials from MfE, the Ministry of Primary Industries (MPI), the Ministry of Business, Innovation and Employment, the Department of Conservation, and Treasury (MfE manager 3c, 2016). At the core of this collaboration was MfE and MPI, a collaboration which the government felt was crucial to 119  integrating economic production concerns within any resulting freshwater regulation (MfE managers 3a, 3b, 3c, 2016). The Directorate was allocated significant budget, as ‘Ministers… went “oh shit, we need to provide a lot more resource around this”’ (MfE manager 3c, 2016). From 2012 to 2016, the Directorate doubled its original staff of 25 and quadrupled its original budget of $4M through successful budget bids supported by the Minister (MfE manager 3b, 2016).  One manager described the early years of the Directorate as being ‘the war room on water’, where high stakes policy was being developed, especially the National Objectives Framework (MfE manager 3a, 2016). The Directorate translated many of the LAWF recommendations into discrete policy levers (see MfE, 2013a). For example, the introduction of a community planning model would be included as an option in amendments to the RMA, and stock fencing would require a new regulation separate from the NPSFM. To generate national bottom lines for freshwater quality, the Directorate convened a group of over 50 scientists to build the National Objectives Framework, and in 2014 the government gazetted the revised NPSFM which included the Framework. Since 2014 the Directorate’s focus has shifted from policy development towards implementation support (MfE manager 3c, 2016).  3.6.1 Realizing freshwater reform: organizing logics of the Directorate When I joined the Directorate in February 2016, it consisted of eight teams. Table 9 presents a stylised summary of each team with indicative projects for some teams. Officials from policy and implementation both described the work of the Directorate in relation to four broad orientations to the NPSFM (MfE officials 3j, 3m, 2016): • implementing the existing NPSFM (guidance, implementation) • testing whether the NPSFM is working (evidence-reporting) • amending the NPSFM in response to specific issues raised (NPS amendments, evidence-policy) 120  • forward-thinking policy development outside the current NPSFM (policy-strategy, rights and interests) Table 9. Water Directorate team descriptions and some Illustrative projects, as of February 2016. Team name Description Water policy and strategy (henceforth policy-strategy) Developing options for allocation and economic efficiency • Waiting for LAWF to provide recommendations • Options for addressing over-allocated catchments • Options for transfer/trade of water consents • International review of allocation systems, supported with council case studies • Synthesising submissions from public consultation Rights and interests Options for resolving iwi rights and interests in freshwater • Working on RMA amendments specifying requirements for participation of Māori in resource management • Working on NPS amendments specifying requirements to include tangata whenua values in freshwater planning NPS amendments Developing proposals for NPS revisions based on public submissions and feedback • Clarifying “overall water quality across a region” • Clarifying “limits” Evidence-policy Synthesizing science to inform amendments to the National Objectives Framework and economics to inform limit setting • Considering information about the Macroinvertebrate Community Index as a metric to incorporate into the NOF • Considering adding new nitrogen and phosphorous attributes to the NOF Evidence-reporting State of environment reporting and public communication of scientific information • Working with Land Air Water Aotearoa (LAWA) as web-based tool to communicate scientific information to the public Program management Internally-focussed team charged with coordination of work across the Directorate Guidance Develop and publish guidance for interpreting and implementing the NPSFM • Guidance documents on collaboration, Section 32 analysis 121  • Running a “limits exchange” Implementation Build capacity and maintain strong relationships with stakeholders • Relationship management • Economic capacity building • NPS implementation review  This conceptual division of labour provides a coherent sense of purpose for the officials in the Directorate. Across these eight teams, officials experienced different feelings of constraint and agency. For the policy-strategy team, the issue of water allocation (developing methods to allocate water quantity or quality to users) was politically sensitive, and officials were required to ensure that they did not send any public signals about their direction of thought, or even the potential range of thought on the topic (MfE official 3m, 2016). The iwi rights and interests team were constrained in a different-but-related way, as the government’s position that ‘no one owns water’ meant that Māori would not be allocated the right to any (proportional or absolute) quantum of water (see Ruru 2013). The team’s function was to determine what Māori rights and interests in water might be in a legal sense, without giving Māori proprietary rights to water. As an official from another team reflected, ‘it seemed a bit peculiar to have the whole allocation debate... sitting within that rights and interests team, [who] were then given very little latitude to actually give anything’ (MfE official 3g, 2016). In contrast, the implementation and guidance teams felt they had significant discretion set their own agenda and frame their own projects, drawing on input from stakeholders (MfE official 3b, manager 2b, 2016). Where a bureaucrat sits within the Directorate affects their ability to frame and undertake projects, and the discretional capacity of the implementation and guidance teams needs to be placed into context. First, there was a sense – expressed by officials – that the implementation and guidance work was non-political, as it was ‘about the stuff that’s already changed from a central government point of view’ (MfE official 3m, 2016). Second, perhaps owing to the relative novelty of the implementation program within MfE, 122  officials had considerable discretion to formulate their own agenda and projects (MfE officials 3b, 3d, 2016).  3.6.2 Styles and forums of bureaucratic agency Conceptualizing the qualities and arenas of bureaucratic agency (i.e. discretion) across the work of the Directorate can be undertaken through three lenses. Since Section 3.5 examined the agency of implementation officials in designing and undertaking projects, here I consider the agency exercised by a more constrained team (rights and interests), as well as the Directorate more broadly in relation to freshwater reform agenda and its role in reshaping policy and project development. Decolonising around set policy parameters On issues for which the government of the day has clear policies or stated preferences, explicit constraints on officials are arguably the strongest. These constraints come in the form of heightened secrecy about the nature of ideas being considered and the type of work being undertaken, and internalised censorship regarding certain types of consultative activities. Even within sensitive issues, however, creativity and improvisation are still required to forward policy work in these areas, and bureaucrats’ work is far from some machine-like translation of pre-formed ideological interests of politicians.  Consider the work of the rights and interests team. The allocation and economic use of freshwater is directly relevant to guarantees in the Treaty of Waitangi that Māori would retain sovereignty over – and possession of – their lands and taonga (or treasures, see Chapter 2). As a part of the freshwater reform undertaken by the fifth National government, the freshwater rights and interests of Māori were designated as a key work area (MfE, 2013a). Within the Directorate, a ‘rights and interests’ team was set up, and a new governance collective was initiated to deliberate with MfE. The Iwi Leaders Group consisted of eight representatives selected from the Iwi Chairs Forum, itself a nation-wide collective of 123  iwi leaders. The Iwi Leaders Group had a special mandate to interface with the Crown on freshwater policy, and were supported by an advisory team of experts and officials, including the rights and interests team. One official notes that while this engagement was not formal, it was very useful: That engagement doesn’t replace consultation and engagement nationally with other iwi and hapū, but it sort of happens in a safe space. So it’s a good faith engagement, it’s free and frank discussions, information gets shared on a confidential basis and that has helped us to test ideas, share ideas and develop proposals that hopefully better meet iwi expectations. (MfE official 3n, 2016) Key achievements from these interactions include: acknowledging Te Mana o Te Wai (the spirit of the water) within in the 2011 NPSFM, explicitly identifying Māori values such as mahinga kai (traditional food gathering) in the in the 2014 NPSFM, strengthening the Te Mana o te Wai provisions in the 2017 NPSFM, and strengthening expectations for regional councils to engage with iwi in freshwater planning, as included in the 2017 RMA reforms. These achievements were both supported and driven by officials who took notes from over 100 regional hui (meetings) and read, analysed, and summarised feedback from Māori and the wider community about the Next Steps for Freshwater consultation document (MfE, 2016b). In addition, officials worked with the Iwi Leaders Group to articulate recommendations and policy language for a revised NPSFM and a revised RMA (MfE official 3n, 2016). Dialogues between the Iwi Leaders Group and the Crown have been often referred to prominently in MfE policy documents to enhance the perceived legitimacy of government policy (e.g., MfE, 2013a: 8; 2016b: 27). However, policies resulting from these conversations do not necessarily reflect the complete aspirations or consent of the Iwi Leaders Group (MfE official 3o, 2016). Indeed, the Iwi Leaders Group has publicly criticized the government’s decisions on numerous occasions, such as the government’s addition of language around 124  economic values in the 2017 NPSFM amendments (e.g. see Neilson, 2017). MfE officials are also quick to challenge the view that government policy fully reflects the Iwi Leaders Group’s aspirations. When I asked if Iwi Leaders Group-Crown dialogue was an example of co-governance, I was immediately corrected: My understanding of co-governance is that there is joint decision making, it’s important to state that the Crown retains decision making for public policy, so this is not a situation of co-governance. What it is is a process of collaborating and sharing ideas and information to develop better policy. (MfE official 3n, 2016) Given that the Iwi Leaders Group has been prominently cited by the fifth National government, this placed pressure on the government to incorporate as many of the Iwi Leaders Group’s recommendations as possible. In the Next Steps for Freshwater document, MfE (2016b: 28) note that ‘feedback from regional councils and the Freshwater Iwi Leaders Group… is that… [Te Mana o te Wai] is unclear and provides ambiguous and inadequate direction.’ While the revised 2017 NPSFM included a new controversial Objective A4 emphasising the provision of economic well-being (New Zealand Government 2017: 12), the new description of Te Mana o te Wai in the same document clarified that ‘it is intended that the health and well-being of freshwater bodies is at the forefront of all discussions and decisions about fresh water’ (ibid: 7). Although it remains to be seen if this Te Mana o te Wai language will be influential in legal settings, the addition of both statements into the 2017 NPSFM is likely better for both Māori and the environment than the addition of the economic provisions alone.  In their efforts to support the creation, articulation, and uptake of recommendations from the Iwi Leaders Group, officials would not describe their work as opposed to the neoliberal environmental agenda of the fifth National government. As Mosse (2005) argues, bureaucratic power can be practised not only through changing the stated objectives of an organization, but also through changing how existing objectives are interpreted and 125  connected to objectives elsewhere in the system (see also Heyman 2004). MfE officials were instrumental in working with the Iwi Leaders Group to frame and prioritize their recommendations, and narrating these recommendations in ways that could be perceived as non-threatening to a neoliberal centre-right government. While certain freshwater policy outcomes for Māori have been significantly limited by the government’s stated policy positions, officials and the Iwi Leaders Group have worked to secure significant gains in the policy process that would not have been achieved in the absence of their energy and commitment to decolonizing the state. Strategic lesson-drawing from LAWF and Europe In contrast to tightly constrained policy issues, the more open question of how to design a system of national environmental baselines created two new moments in the policy process: translating recommendations from the Land and Water Forum, and distilling lessons from overseas. Bureaucrats exercised agency by interpreting these ‘sources’ in different ways. In 2010 the Land and Water Forum (LAWF) advocated for the development of a national regulatory regime built around bottom lines for freshwater quality. Since the LAWF recommendations are branded with the mark of consensus of over 50 major civil society organizations, the Directorate made significant efforts to work with LAWF, staffing the LAWF secretariat and sequencing MfE policy work around LAWF timelines and outputs. This was a consequential decision. For example, one MfE manager wanted to develop a limit-setting toolkit that councils and stakeholders had requested, but their proposal for funding was declined on the basis that it would ‘interfere with what the LAWF are thinking about’ (MfE manager 3d, 2016). A potential interpretation of this decision is that Directorate managers wanted to work with the political power of LAWF to advocate strengthening environmental limits, and needed to ensure LAWF members felt central to the policy process. 126  Upon reading LAWF’s recommendations, however, MfE managers felt many of these recommendations were not ‘policy-ready’ (MfE manager 3a, 2016). This created space for officials to exercise discretion in adding further agenda items as well as criticizing specific consensus-produced LAWF recommendations. One official noted, for example, that while the LAWF recommendations covered many things, ‘they couldn’t really land on… the crunchiest, hardest stuff’ such as allocation methods or what do about new water users (MfE official 3m, 2016). Another official argued that the LAWF’s recommendation to force councils to report on the Macroinvertebrate Community Index (MCI) of rivers as a measure of ecological health was ‘completely flawed’ because it was ‘never designed for that use’ (MfE official 3p, 2016). In this way, officials have had to tread a careful line between recognizing LAWF’s legitimacy versus augmenting LAWFs proposals based on technical merit. These discretionary moments of deciding what is ‘policy-ready’ and what is not can risk undermining the legitimacy of the LAWF process. For example, one member of LAWF recounted feelings of frustration with MfE’s apparent resistance to LAWF’s recommendation to include MCI: I think the officials fuck it up - good technical term. It’s mind boggling… I mean [named official] was busy arguing all the way that we couldn’t do MCI, and everybody is just rolling their eyes and going “We’ve been doing this for 30 years, every regional council manages MCI. MCI is a major part of resource consents and stuff, we really know MCI.” Despite every single scientist saying “no, it’s fine”, every regional council saying “no, it’s fine”, but [named official]… the Directorate sent out another view. (interview, K. Hackwell, 2017) If the selective adoption and resistance of LAWF recommendations can be described a type of discretionary cherry-picking, a second type of cherry-picking can be identified in efforts to learn from Europe. In the early days of the Water Directorate, several Directorate officials undertook information-gathering trips abroad to learn about the European Water Framework Directive. Adopted in 2000, the Water Framework Directive requires all member nations to 127  achieve ‘good ecological status’ for all waterbodies by 2027, and provides an analogue for New Zealand’s experience with the NPSFM ( The OECD has recently compared New Zealand’s environmental bottom lines with those of the Water Framework Directive (WFD), finding that ‘New Zealand bottom lines are set below the equivalent WFD bottom line’ and that the NPSFM includes looser criteria for pleading exception to these bottom lines (OECD, 2017b: 170). The OECD suggested the NZ government might consider ‘setting the default water quality level high’ and having strict criteria for any downward negotiation from a high default (OECD, 2017b: 171). Against this, however, an MfE manager during the creation of the 2014 NPSFM recalls that European practitioners they spoke to ‘were pretty adamant about what not to do’ (MfE manager 3a, 2016). European practitioners warned against using European-style bottom-lines in policy because ‘if you have raw numbers that people have to meet, and they fail, and they’re always going to fail’ (MfE manager 3a, 2016). Having strong ecological bottom-lines in Europe rendered most implementing districts non-compliant, ‘which meant that actually the government was reluctant to do anything because it felt so hopeless’ (MfE manager 3a, 2016). When policymakers looked at a map of their countries with most of the colors were red (designating failure to meet bottom-lines), this became paralyzing politically. European officials also warned against allowing a regime of exceptions, because ‘in Europe everyone has got an exception’, and ‘an exception was a free pass, and we didn’t want a system that was based on free passes, we wanted a system that was based on planning to get better’ (MfE manager 3a, 2016). Anothe