Idealism, pragmatism, and the power of compromise in the negotiation of New Zealand’s Zero Carbon Act

ABSTRACT Discursive choices made by policy entrepreneurs are an important factor in the development of climate change acts (CCAs). This article examines the extent to which such choices reflect the strategic need for CCA entrepreneurs to compromise pragmatically and modulate their policy preferences in order to secure the agreement needed for CCA adoption. Drawing upon theoretical insights from discursive institutionalism (DI) and policy entrepreneurship, this article analyses discursive choices during negotiations surrounding the New Zealand Zero Carbon Act (ZCA). The analysis shows that endogenous political-ideological constraints compelled entrepreneurial actors to modify first-choice preferences for emissions reduction legislation by reframing their coordinative discursive interventions to accommodate potentially oppositional groups. Further research is required into the conditions under which such strategies become discursively operational, to provide guidance to climate policy entrepreneurs as CCAs continue to diffuse globally. Key policy insights Strategic compromise by climate advocates is crucial to the passage of enduring legislation. Climate policy entrepreneurs’ decisions about how and when to compromise to ensure legislation may have significant implications for climate policy efficacy and political durability. Compromises may only defer rather than diffuse underlying political tensions but can enable CCA adoption so as to reshape political contexts in the longer-term. Future research can inform climate policy advocacy strategies that aim to balance ambition and durability.


Introduction
In November 2019, New Zealand passed the Climate Change Response (Zero Carbon) Amendment Act (Zero Carbon Act (ZCA)), enshrining a legal commitment to achieve net-zero emissions by 2050. The ZCA represented landmark legislation not just because of its intent but also because it enjoyed cross-party support from the Labour-led government and the opposition National Party after a decade of partisan wrangling over climate policy (Ainge Roy, 2019). In June 2020, Parliament also passed the Climate Change Response (Emissions Trading Reform) Amendment Bill (ETR Bill) to strengthen New Zealand's mainstay policy instrument for reducing emissions. This time, however, the National Party voted against the legislation, arguing that New Zealand should not commit to pricing livestock and fertiliser emissions or higher emissions trading scheme (ETS) prices until the economic impact of Covid-19 was better understood (Wannan, 2020).
Our empirical questions on how cross-party agreement on the approach to New Zealand climate policy occurred and came under strain so quickly afterwards feed into our main analytical question: how do discursive choices made by climate policy entrepreneurs contribute to reconfiguring political debate on the adoption and design of Climate Change Acts (CCAs)? The role of framing and discourse in shaping conditions for climate legislation has been analysed extensively (Carter & Jacobs, 2014;Christoff, 2013;Gillard, 2016;Inderberg & Bailey, 2019;Lockwood, 2013;. Research has shown how recasting climate change from an environmental to an economic frame can enable CCA advocates to emphasise the socio-economic benefits of decarbonization (Carter & Jacobs, 2014). Other studies, however, demonstrate how alternative discourses on the costs and distributional effects of long-term emissions targets can strain cross-party agreement over CCAs (Gillard, 2016;Lockwood, 2013). In both cases, the discursive choices made by policy entrepreneurs seeking to shape political debate and legislative responses to climate change, and associated structural constraints, remain significant factors in CCA development but further work is needed to deepen understanding of the decisions CCA advocates face about compromising with other actors, particularly where major shifts in cooperation are needed.
Compromise occupies a necessary and central role in democratic politics (Bellamy et al., 2012), but decisions about whether, how and when actors choose to compromise have particular pertinence for CCAs because decisions at one juncture may create multi-decadal precedents for future institutional conditions and policy options. Inherent to the CCA concept is the idea that the self-reinforcing path-dependencies created by long-term policy commitments and mechanisms means that, within reason, a flawed but agreed CCA may be preferable to something perfect that never materializes (Levin et al., 2012;Rietig & Laing, 2017;Rosenbloom et al., 2019).
To explore how such decisions shaped the development and adoption of the New Zealand framework climate legislation, we draw on ideas from discursive institutionalism (DI) about how institutions respond to, reshape and constrain ideas to analyse the different ideas policy entrepreneurs utilize to champion policy change. These range from philosophical ideas to detailed policy ideas, and the interactions through which policy élites communicate with general audiences and coordinate the preferences and expectations of other policy actors (Schmidt, 2002(Schmidt, , 2008. By developing a framework that combines DI with the concept of policy entrepreneurship, we consider both the content and context of policy discourse from a strategic perspective.
Section 2 explains the DI framework in more detail. Section 3 then describes data sources and processtracing methods, before Section 4 outlines the New Zealand climate policy context ZCA advocates needed to navigate. Section 5 discusses the findings, paying particular attention to the role of compromise in policy actors' discursive strategies. Section 6 concludes with lessons gained and suggests that focusing on the role of compromise provides new insights into the ability of climate change advocates to contribute to enduring and progressive climate legislation.

Framework
The theoretical framework developed from the DI and climate-energy policy literature categorizes the strategic choices made by climate policy entrepreneurs by focussing on policy discourse as a primary strategic tool. In so doing, we emphasise the often pivotal choices about when and how to compromise with others in the exercise of political agency.

Consensus or compromise
The policy process is one of the main means by which policy disagreements are resolved in pluralist liberal democracies. Actors seeking to influence the form of agreements during policy processes face choices about the type of resolution they seek which, we argue, can take the form of either consensus or compromise outcomes. Compromise outcomes here refer to situations where participants are willing to agree to an outcome they regard as non-optimal but better than no agreement. Compromise outcomes contrast with consensus outcomes where relevant parties agree that a course of action is optimal (May, 2011;Rossi, 2013;Weinstock, 2013Weinstock, , 2017. Using these definitions, actors may seek consensus in two types of situation. The first is where they do not perceive a need to compromise to secure an agreement, either because the terms of agreement are non-controversial, or because they judge they can persuade or influence others to accept the superiority of their preferred outcome. The second is where the cost of compromise is greater than the cost of failing to find agreement, based on either a pragmatic calculation or an unwillingness to violate moral or ethical principles (May, 2005). An actor who attempts -but failsto build consensus for their preferred result risks an outcome where no agreement is reached, or their isolation from the policy discourse leading to an agreement over which they have had no influence.
In contrast, decisions to pursue compromise reflect assessments by actors that they are unable to realize an agreement that fully reflects their first preferences and that the cost of the compromise to secure non-optimal agreement is less than the cost of no agreement or the actor's marginalization from the decision process. Compromise may represent either a straightforward material concession or the re-framing of their position to better align with the terms of the policy debate. An example of the latter form of compromise might be the reframing of climate change from a moral to an economic issue.

Institutions, ideas, and discursive interaction
Discursive institutionalism has its roots in constructivist understandings of how institutions respond to ideas embedded in discourse (Peters, 2019) and has been used widely to examine how entrepreneurial political actors involved in climate and energy governance manipulate ideas (Fitch-Roy et al., 2020;. It proposes that actors exercise agency through both their 'background ideational abilities', which allow action within the constraints of existing contexts, and their 'foreground discursive abilities' that enable them to think outside their institutional context and communicate to change or maintain those institutions (Schmidt, 2008(Schmidt, , p. 314, 2017. DI examines ideas at several levels. Policy ideas are the most specific and relate mainly to policy provisions. Programmatic ideas underpin policy ideas and set out problems to be solved by policies. Finally, philosophical ideas can be seen as 'deep core' ideas that are rarely contested by their adherents (Schmidt, 2008, p. 306). Schmidt also distinguishes between 'cognitive' ideas (indicating 'what is or what to do') and 'normative' ideas ('what one ought to do') (Schmidt, 2008, p. 306). Cognitive ideas thus provide the methods or procedures for political action and, at the programmatic level, may define the problems to be solved and indicate solutions. By contrast, normative ideas provide values to underpin political action and validate policies.
Finally, 'discourse' centres on the processes through which actors interact and ideas are generated, deliberated, legitimated, and agreed upon (Schmidt, 2008(Schmidt, , 2010. 'Coordinative' discourse focuses on dialogue between direct participants in the policy process, usually political élites: elected officials, civil servants, and interest groups (Schmidt, 2002;2008). 'Communicative' discourse, meanwhile, focuses on conveying ideas to wider public audiences, and encompasses a wider range of political actors (such as political leaders, policy forums, and informed publics) who bring forward ideas developed through coordinative discourse for public deliberation and legitimation.

Entrepreneurship: agency and power
The DI literature ascribes an important role to policy entrepreneurs ' … as catalysts for change [who] draw on and articulate the ideas of discursive communities and coalitions' (Schmidt, 2008, p. 310). These actors invest time and effort in identifying, analysing, and advancing particular problems on the policy agenda and offering solutions (Béland & Katapally, 2018;Capano & Galanti, 2020). They often achieve greater results than their material resources suggest (Boasson & Huitema, 2017), by using their discursive abilities to facilitate cooperation within and between groups, build and maintain institutions, and reshape policy debates through creating new forms of meaning that attract other political actors (Fligstein, 2001, p. 106). This may necessitate developing frames and narratives that resonate with others' 'identity, belief, and interests, while … using those same stories to frame action against various opponents' (Fligstein & McAdam, 2012, p. 50).
Power in DI is often, therefore, ideational. Power is exerted 'when actors seek to influence the beliefs of other [actors] by promoting their own ideas at the expense of other [ideas]' (Carstensen & Schmidt, 2016, p. 322).
Thus, agency to shape policy is partially contingent on strategically manipulating ideas. Carstensen and Schmidt (2016) identify three 'types' of ideational power: (1) power through ideas: persuading others to accept an argument by invoking ideas; (2) power over ideas: imposing particular meanings on others, either directly or by resisting alternative ideas; (3) power in ideas: institutionalizing ideational structures in ways that narrow the scope for other ideas to gain purchase.
Strategic decisions by policy entrepreneurs consequently hinge on which ideational power resources they possess. They may develop what they perceive as the 'killer argument' for their preferred outcome, or be an especially persuasive political operative, and seek to exert 'power through ideas'. Alternatively, they may have the capacity to impose ideas through control of the communicative discourse in media channels or to 'shame' others by developing a powerful normative idea that others cannot rebut; as seen in the moral authority of some social movements. The recourse here is to 'power over ideas', emphasising compulsion rather than persuasion. Finally, 'power in ideas' suggests that actors may enjoy deeper structural or institutional support for their preferred idea. Underlying social, political, and intellectual landscapes inevitably place some ideas within or outside the realm of topics that may be fruitfully discussed. That some ideas may be implicitly or explicitly considered indefensible may result from historical path dependency or the predominance of particular frames or modes of thought (Bailey & Wilson, 2009).
Not all actors have access to equivalent ideational power resources: this depends largely on hierarchical or institutional status. Since climate policy entrepreneurs are expected to operate with extraordinary proficiency relative to their role (Boasson & Huitema, 2017), policy entrepreneurs are less likely to flex power 'in' or 'over' ideas that relate more to status than skill. A reasonable expectation is for policy entrepreneurship to rely heavily on powers of persuasion 'through ideas'. However, one objective of persuasion through ideas is to alter the terms of the policy debate and increase the power in a particular idea or ideas set.
Policy entrepreneurs' strategic decision-making therefore demands assessment of the distribution of ideational power resources, which itself necessitates evaluation of the political context, each actor's position within that context, and each actor's political abilities and policy preferences. Actors seeking influence must continually reappraise these conditions and, at each iteration, choose between one of two basic categories of action: to pursue their 'ideal' outcomes; or concede ideational or material ground to secure progress and avoid 'the perfect becoming the enemy of the good'. We therefore emphasise the strategic considerations facing policy entrepreneurs when deciding when to compromise by modulating their preferences or reframing their arguments. Across DI's ideational levels, the pursuit of consensus outcomes may be more likely when philosophical ideas are at stake, while compromise outcomes may be more common when philosophical ideas are unthreatened or there are severe reputational or material consequences of not reaching an agreement. Examining the terms of policy discourse across all three levels allows us to use the framework to illuminate the 'depth' of the political agreements that drive policy outcomes. Compromise to achieve 'something rather than nothing' Ability to realize first-best preferences is constrained by context or others' power over or in ideas Concession through adopting less optimal policy preferences (power through ideas allows recognition of limits to consensus) Reframing to address opponents' argument directly (and persuade acceptance of compromise through ideas) Table 1 sets out the two broad categories of agreement an actor may choose to pursue, alongside the implications for the two main components of agency in DI: the background ideational abilities through which an assessment of the context is formed, and the foreground discursive abilities that allow the deployment of ideational power.

Methods
The research was based on a qualitative 'holistic' single case study (Yin, 2014) that mirrored other theory-testing CCA studies (Gillard, 2016;, using the New Zealand ZCA/ETR Bill process as a focal point for deeper examination of relationships between discursive strategising and policy decisions. Process tracing was used to uncover the discursive strategies employed from the inception of the policy process to its adoption, with a particular emphasis on agenda setting and policy negotiation. This technique tests theoretical propositions in political processes over time by disaggregating them into key decision points (Bennett & Checkel, 2014). In this study, we trace the evolution of New Zealand framework climate legislation through four temporally sequential stages to understand how different discursive strategies influenced policy development.
To construct a timeline of the process, primary, secondary and tertiary material were analysed and events triangulated.
The empirical material was drawn mainly from public documents but one semi-structured interview was conducted with an individual involved in the NGO Generation Zero to explore and validate interpretations of informal processes and timelines. This empirical material was supported by insights on debates surrounding New Zealand climate policy gained from 23 expert interviews conducted with politicians, government officials, business groups, NGOs and independent commentators for a previous research project, which were used indirectly to help interpret the discursive and policy positions of different actor groups for the present study (Inderberg & Bailey, 2019). The following sources were used to map the policy positions of different actors at different stages of the policy process against the main arguments used to justify these stances: (1) Publications by organizations promoting the ZCA; (2) Consultations and reports on the ZCA and ETR Bill, including: government discussion documents produced to accompany the ZCA consultation; submissions from industry, NGOs and other groups and individuals; and analyses of consultation findings; (3) Texts of the Bills and Supplementary Order Papers; (4) Hansard records of the bills' parliamentary readings; (5) Cabinet papers and regulatory impact analyses of measures to manage livestock and fertiliser emissions; and industry submissions proposing alternatives to mandatory pricing of livestock and fertiliser emissions.

The development of New Zealand's Zero Carbon Act
The concept of CCAs as multi-faceted legislative frameworks that mandate long-term targets and policy actions for mitigating greenhouse gases (GHGs) (

Phase one: Generation Zero and the ZCA genesis
Although political debates on reforming New Zealand climate policy had been ongoing since prior to the introduction of the New Zealand Emissions Trading Scheme (NZ ETS) in 2008, one of the main entrepreneurial drivers creating agendas for a new legal framework was Generation Zero, a youth-based NGO that had been active on climate issues since 2010 (Generation Zero, 2020). Generation Zero published a series of reports during this early period that drew on international precedents to lobby for more ambitious New Zealand climate policy (Generation Zero, 2014). In particular, the reports stressed the UK CCA's use of legally binding emissions targets, carbon budgets, and the creation of an independent oversight and advisory committee. Emulation of the UK was also apparent in its second report's title, The Big Ask, which echoed the citizen campaign to support the establishment of the UK CCA (Carter & Childs, 2018). Generation Zero's campaign nevertheless stressed that New Zealand climate policy should reflect the needs of its trade-exposed and primarysector dependent economy.
In 2016, following adoption of the Paris Agreement, the group began advocating explicitly for legislation that would create cross-party commitment to an ambitious, long-term decarbonization programme and end partisan wrangling over the direction of New Zealand climate policy (Mclaren, 2019;Interview). Inspired by the unusually high degree of cross-party agreement exhibited during the passage of the UK's CCA, Generation Zero's view was that ensuring broad political support was indispensable to securing meaningful and durable legislation within New Zealand's plural party mixed-member proportional (MMP) and short (three-year) electoral system.
The group's tactics of direct action and conventional political activities (Dodson & Papoutsaki, 2017;Noronha, 2013) became a large-scale communication exercise to mobilize public support, simultaneously articulating philosophical ideas (climate challenges facing New Zealand), programmatic ideas (framework climate legislation) and policy ideas (the proposed Act's features). The campaign incorporated both online and traditional campaigning, including events like Zero Carbon Act: The most important legislation for our generation (June 2016); and International case studies and lessons for New Zealand, which again drew on other countries' experiences to argue for a New Zealand CCA (Generation Zero, 2017a). Effort was also invested in coordinative networking, using personal and other NGO connections to engage business and primary-sector actors, as well as influencing local politics through an 'adopt an MP' campaign encouraging communities to lobby their parliamentary representatives. Further coordinative activity focused on strengthening the political acceptability of Generation Zero's policy ideas by seeking advice from a non-partizan group of academics, politicians, and other professionals on how to design economically and politically viable legislation for New Zealand.
In April 2017, Generation Zero's policy team launched its 'blueprint' for a 'Zero Carbon Act' that adapted elements of the UK CCA to reflect Treaty of Waitangi 1 requirements, New Zealand's Pacific responsibilities, and the country's emissions profile (Generation Zero, 2017b). It also represented the group's first major decision on whether to compromise on how it lobbied for a ZCA. To promote non-partizan politics and diffuse tensions between the environmental and agricultural communities, it endorsed a 'two-baskets approach' that recommended different legal treatment of long-lived (mainly carbon dioxide and nitrous oxide), and short-lived GHGs (mainly methane associated with agriculture) (Generation Zero, 2017b). By incorporating scientifically recognized differences in methane's physical properties into its blueprint (Cain et al., 2019), Generation Zero hoped to make its proposals palatable to agricultural interests, though it never intended this discursive compromise to signal diminished ambition to reduce short-lived GHGs (Interview).
The blueprint was supported by influential organizations including Forest & Bird, WWF-New Zealand, Oxfam NZ, and some businesses (Generation Zero, 2018a). Another objective for Generation Zero was to secure support from New Zealand's youth political parties (Nationals, Labour, Greens, New Zealand First (NZ First) and Māori Party) to reinforce the idea that the ZCA transcended party politics. Support from the Young Nationals was especially important as a way of encouraging the National parliamentary party to support cross-parliamentary working on the ZCA (Interview). However, Generation Zero purposefully avoided overt political endorsements for its blueprint that might compromise its party-neutral status and when the Green Party cited the ZCA blueprint in its 2017 election manifesto, Generation Zero requested its removal.
The other major source of momentum during the ZCA's agenda-setting phase came from closer to the political establishment, when in 2015 the prominent Greens politician, Kennedy Graham, founded GLOBE-NZ to advance cross-party collaboration on climate change. 2 Graham enlisted representatives from New Zealand's six main parties to participate in briefings and discussions on climate change. The briefings from scientists, politicians and business leaders played an important coordinative role in 'bind[ing] a disparate group of New Zealand parliamentarians' towards the goal of net-zero emissions (Graham, 2018, p. 38), which GLOBE-NZ sought to extend by commissioning Vivid Economics to conduct a study of pathways for achieving net-zero emissions. Vivid examined options ranging from reliance on international offsets to strategies combining innovation, shifts from pastoral agriculture, and extensive afforestation, while its conclusions contained several measures aligned with the CCA concept, including the need for cross-party agreement to enhance policy coherence and predictability, and the creation of independent statutory institutions to assist with the development of national climate policy (Vivid Economics, 2017).
Strong public interest in the report led to a special debate in Parliament in April 2017, wherealthough a CCA was never mentioned directlythe Greens co-leader, James Shaw, committed the party to net zero by 2050, Labour called for stronger action, and NZ First and the Māori Party representatives indicated the importance of achieving net zero by 2050 (New Zealand Parliament, 2017). This was followed in 2017 by two GLOBE-NZ sponsored consultancy papers examining pathways towards net zero by 2050 and ways to strengthen emissions pricing that helped to shape New Zealand political discourse further towards how, rather than whether, to pursue net-zero emissions (Leining, 2017;Young, 2017). Reflecting on the GLOBE-NZ initiative, Graham expressed the belief that it helped to bridge entrenched divides in New Zealand politics by creating a forum where members could analyse issues and express views with 'more political fluidity than would normally be found in the debating chamber' (Graham, 2018, p. 43).
By this juncture, the Greens and Labour had become critical ZCA entrepreneurs, with Labour arguing in its 2017 election manifesto that 'It is not good enough to say we [New Zealand] are too small to mattermost countries individually could claim the same' (Labour, 2017). NZ First similarly pledged to establish a UK/Norwegian style CCA, and although National's manifesto did not mention a CCA, it did not oppose the idea. When Labour's electoral performance enabled it to form a coalition government with NZ First supported by a 'confidence-and-supply' agreement with the Greens, Prime Minister Jacinda Ardern spoke about her determination for New Zealand to be on the right side of history on climate change and appointed the Greens' James Shaw as Minister for Climate Change. Endorsement of the ZCA proposal by the Office of the Parliamentary Commissioner for the Environment in March 2018 (WWF New Zealand, 2018a) gave the government further encouragement to launch consultations on proposals for a CCA in June 2018.

Phase two: the Zero carbon consultation
The consultation, Our Climate Your Say, was pivotal in the government's efforts to build support for the ZCA (Ministry for the Environment, 2018a). In courting views from business, agriculture, forestry, civil society groups, Māori communities, local government and the public, the governmenthaving taken the reins of policy entrepreneurship on the ZCAsought to secure a mandate to underpin negotiations with potentially resistant political and business actors. The consultation discussion document emphasised the risks of climate change to businesses and society, New Zealand's moral duty to protect future generations and the Pacific region, and the country's significance to global mitigation efforts: Although New Zealand's share of global emissions is very small (0.17%) … . The Zero Carbon Bill is an opportunity for New Zealand to decide how it delivers its part in the global effort while encouraging action by others. (Ministry for the Environment, 2018a, p. 9) On the programmatic level, the document explored the capacity for a ZCA to promote innovation, employment, the Māori economy, cleaner air, resilient ecosystems, efficient and comfortable homes, and reduced traffic (Ministry for the Environment, 2018a). Having framed the government's stance, the document then sought opinions on policy options, including long-term emissions targets, carbon budgets, protecting tradeexposed sectors, an independent Climate Change Commission, and adaptation policy (Ministry for the Environment, 2018a).
The document also discussed transition challenges, including the risks of reduced economic growth, industrial decline, unemployment, and carbon leakage to countries with less stringent climate policies (Ministry for the Environment, 2018a, p. 19). Although such messages arguably provided ammunition to potential ZCA opponents, they also demonstrated the government's alertness to negative outcomes and its commitment to a fair and inclusive transition (ibid. p. 11). Equally, if the consultation gained broad-based support, it would signal to political opponents that New Zealand accepted these challenges.
The consultation received 15,009 submissions, including around 9,500 pro-forma letters submitted using templates developed by Greenpeace, Generation Zero and other activist organizations (Ministry for the Environment, 2018b). Two recurring themes are identifiable within the discourses and arguments used by affected groups and other respondents: (i) recognition of the need for action and the ZCA; and (ii) defence of respondents' interests and involvement in the policy process. Industry organizations like BusinessNZ that had previously resisted strengthening of the ETS indicated principled support for the ZCA and presented themselves as active decarbonization partners (BusinessNZ, 2018). Similarly, Fonterra, New Zealand's dominant dairying cooperative, signalled its commitment to the ZCA's principles (Fonterra, 2018). However, business acceptance of the ZCA was not unconditional. BusinessNZ criticized the government's modelling and methodology for determining emissions targets, arguing that it could: … foreshadow the implementation of policies aimed at achieving them regardless of the economic and social consequences. (BusinessNZ, 2018, p. 2) Fonterra petitioned for involvement in policy design to recognize the dairying sector's importance to the economy and global food security, the technical and economic challenges of reducing methane, and the need to manage the ZCA's social and economic impacts on agricultural communities: A significant and rapid approach to reducing methane is likely to have negative economic and social impacts on the agriculture sector that could be avoided. (Fonterra, 2018, p. 4) Forestry related organizations similarly sought to align the ZCA with their commercial interests, arguing for 30year rather than five-year carbon budgets to reflect forestry investment cycles (Wood Processors and Manufacturers Association, 2018). The New Zealand Farm Forestry Association focused on strengthening forestry representation in the Climate Change Commission (New Zealand Farm Forestry Association, 2018), while regional governments emphasised the need to give councils and non-government actors meaningful roles in New Zealand's zero-carbon strategies to ensure just and equitable transitions (Auckland Council, 2018;Greater Wellington Regional Council, 2018). Generation Zero also championed society-wide involvement in the ZCA to safeguard just treatment for Māori communities (Generation Zero, 2018b), while WWF-New Zealand cited the experience of its UK office with the UK CCA to highlight its credentials to assist in developing the ZCA (WWF New Zealand, 2018b).
Another feature of many submissions was claims by organizations to represent large constituencies. The Sustainable Business Council reported that its submission was based on consultation with 95 member companies that accounted for 29% of New Zealand's private sector revenue and 9% of its workforce (Sustainable Business Council, 2018). Similarly, the New Zealand Farm Forestry Association submission was presented on behalf 14,000 forest owners (New Zealand Farm Forestry Association, 2018). So, although the government's consultation strategy secured cross-societal backing for a ZCA, it also highlighted that significant coordination work was still needed to ensure the legislation was acceptable to groups that remained wary of its economic and social implications.

Phase three: legislating the Zero carbon Act
Despite cross-party support for the idea of a ZCA, coalition politics played an important role in shaping the legislation introduced to Parliament. In particular, NZ First opposed allowing the Climate Change Commission powers to set carbon budgets independently and expressed reservations about imposing 'worldleading' obligations on agriculture (Cooke, 2019). In seeking to resolve these issues, James Shaw was reported as saying he was 'more interested in a lasting law than a perfect one' and that he placed a premium on building bipartisan support over pressing his party's desire for stronger measures for agriculture (Cooke, 2019).
Once a compromise had been agreed, the ZCA enjoyed general support from the majority of parties, including the main opposition National Party, during its passage through Parliament. Representatives from across the political divide praised James Shaw for his stewardship of the ZCA and for seeking and securing National's agreement on the legislation, and the debates were characterized by regular affirmation of each party's commitment to shared normative and cognitive ideas, in particular: (1) the need for accelerated action; (2) the need for framework legislation to provide predictable investment and innovation signals; (3) flexibility in how policy responded to changes in scientific evidence and economic conditions; (4) the importance of justice and inclusivity within New Zealand's climate policy response (New Zealand Parliament, 2019a;2019b;2019c).
The mood of the debates nevertheless became combative at times as National invoked New Zealand's status as a small trading nation to portray the ZCA's agriculture provisions as unsupported by scientific or economic evidence and risking harm to rural communities and other vulnerable sections of society (Cooper & Rosin, 2014;Driver et al., 2018;Richter & Chambers, 2014). 3 National representatives argued that tougher regulations for agriculture than existed in other countries reflected a misguided attempt at international leadership and pressed for less demanding targets, measures to prevent productive agricultural land being converted to forestry, greater use of international credits, and clearer plans for innovation and technology development (New Zealand Parliament, 2019a).
Despite cross-party agreement on the idea of fair and just transitions, disagreements again persisted over its interpretation. For example, National argued during the ZCA's second reading that: … the so-called just transition needs to be measured and only move through at a pace in which everybody is comfortable. (New Zealand Parliament, 2019b) Government interpretations, meanwhile, stressed New Zealand's moral obligations to future generations and its Pacific neighbours, the legislation's capacity to strengthen agriculture's competitiveness, and farmers' willingness to embrace the low-emissions challenge: Whether farmers talked about how they recognised that farming practice had to change … to their credit [they] came to talk about how that should be managed. Not to oppose it, but to talk about management. (Duncan Webb, Labour, New Zealand Parliament, 2019b).
The repeated calls for amendments suggest that the political consensus over the ZCA only extended to its principles and framework, while detailed programmatic and policy measures remained susceptible to challenge. Although National's support gave the government few reasons to moderate the ZCA, the introduction of several amendments in its second reading following a further consultation by the cross-party Parliamentary Environment Committee underlined the need for continued coordinative activity and compromise to maintain industry and opposition cooperation. Scott Simpson, National Spokesperson for Climate Change, flagged the potential for a return to adversarial climate politics while proposing a Supplementary Order Paper (SOP) amendment during the ZCA's committee stage: … [we] hope that the Government will take notice of our clear intent in producing and preparing these SOPs. It's likely that if the SOPs that we put forward are not accepted … the potential for change at a future date remains high, and I'm sure that's not the purpose of the Minister's intent in trying to get cross-party support. (New Zealand Parliament, 2019d).
To navigate the legislative process, the government faced several discursive and policy choices that potentially affected the legislation's capacity to deliver economy-wide emissions reductions. These included which arguments to press to defend the ZCA's integrity, how far to accommodate alternative arguments, and whether (and how) to amend provisions on agricultural emissions. Its discursive strategy can be characterized as seeking to 'cover all bases' by simultaneously stressing philosophical ideas about climate change threats, normative ideas about the moral case for action, cognitive ideas about improving economic competitiveness and quality of life, and coordinative commitments to fairness and mitigating adverse economic and social impacts. Its policy strategy in turn could be described as 'keeping the door latched but ajar' by retaining the ZCA's ambitious targets, independent policy scrutiny, and policy priorities while deferring detailed negotiations on how to achieve the ZCA's ambitions to the next stage of the policy process (Leining et al., 2020).
Phase 4: action to achieve the ZCA's goals The Climate Change Response (Emissions Trading Reform) Amendment Bill (ETR Bill) formed separate legislation to the ZCA but is pivotal to understanding the politics of the ZCA's development for two reasons. First, the ETS remains New Zealand's dominant policy instrument for reducing emissions, so its design is critical to realizing the ZCA's ambitions (Bertram & Terry, 2010;Driver et al., 2018). Second, the ETS has been heavily criticized since its introduction, among other things, for exempting livestock and fertiliser emissions because of concerns about economic impacts and carbon leakage from the relocation of agricultural production to other countries (Inderberg et al., 2017).
To address these issues, the government proposed reforms to the ETS that included pricing farm livestock emissions and processor fertiliser emissions. The Primary Sector Leaders Group remained wary of mandatory pricing without clear mechanisms to cut emissions economically and proposed an industry-government agreement (He Waka Eke Noa) to build capability for farm-level ETS participation oragriculture's preferred optionmeeting emissions targets without ETS surrender obligations. The government's strategy for upholding the ZCA's ambitions without alienating agriculture or fracturing the political consensus underpinning the ZCA consisted of: continued negotiation with agricultural leaders to strengthen the proposed agreement's alignment with the ZCA; regulatory impact analysis of the agreement versus pricing livestock and fertiliser emissions in the ETS; further stakeholder consultation; and messaging about the government's constructive working with agriculture to counter opposition aspersions that policy was driven by ideology rather than practicality and fairness: We are pleased the Government are finally getting alongside farmers … it's nice to see they've finally learned that you can work with the sector to get the results that you need. (Erica Stanford, National, New Zealand Parliament, 2019e) The regulatory impact analysis concluded that farm-level livestock and processor-level fertiliser emissions pricing would be more effective in reducing emissions and enjoyed public support (Ministry for the Environment, 2019a). However, it also cited lack of stakeholder buy-in as a threat to the policy's durability, arguing that an industry agreement might be viable if the sector accepted agricultural emissions pricing from 2025, resolved issues around the agreement's governance and funding, and the government reserved the right to introduce a processor-level scheme before 2025 by Order in Council (bypassing parliamentary approval) if progress in implementing the agreement was unsatisfactory (Ministry for the Environment/Ministry for Primary Industries, 2019).
In October 2019, James Shaw sought final cabinet agreement on measures to reduce agricultural emissions. His recommendations recognized the need for flexibility and pricing within a broader policy package but recommended processor-level livestock and fertiliser pricing from 2021 (Ministry for the Environment, 2019b). However, cabinet opted instead for a modified sector agreement while maintaining farm-level livestock obligations and processor-level fertiliser obligations from 2025 4 (Cabinet Office, 2019). How far the cabinet rejected the climate minister's recommendations, or decided upon the outcome he had anticipated all along, is unclear. Either way, it represented a key policy response to risks to the ZCA of pursuing the government's 'first best' policy option.
This phase of the policy process centred mainly on policy choices rather than discursive manoeuvring, although the consultation provided further opportunities for public debate on agriculture's climate responsibilities. 5 The government also continued to deploy communicative and coordinative messaging to sell the deal to opposition parties and industry while reassuring activist and public audiences of its commitment to tackling climate change. Despite the government's attempts to coordinate political support by stressing its partnership with agriculture, it failed to dissuade National from voting against the ETR Bill on the grounds that its economic impacts had received insufficient scrutiny following disruptions to parliamentary procedures caused by the Covid-19 pandemic. The government's parliamentary majority ensured the ETR Bill's passage in June 2020, but National's opposition raises doubts about whether it will enjoy continued support from future governments.

Discussion
The New Zealand ZCA and ETR Bills were subject to a variety of discursive interventions by policy entrepreneurs during their development. A critical question returning to our theoretical framework (Section 2) is whether, when and how these actors chose to compromise to secure a deal, or to maintain their 'first-best' preferences and attempt to expand or enforce a consensus. We discuss the findings in relation to two core themes: the idea of consensus politics as a fundamental prerequisite for climate change legislation; and the compromises required to secure it.

Consensus
A fundamental lesson that New Zealand actors drew from the UK's experiences was that ambitious long-term climate-change legislation was impossible without broad-based political 'consensus' (New Zealand Parliament, 2019b). Seeking the broadest possible support was pivotal to Generation Zero's agenda-setting strategy and, throughout the ZCA consultations, the government worked in both the policy and discursive spheres to demonstrate that the concerns of potential political opponents and key interest groups were being addressed. Business, agriculture and forestry were particular targets for these coordinative efforts but so too were other major groups, such as the Māori community. The depth of communicative engagement with citizens is equally evident in the Generation Zero campaign and the ZCA consultation response. Public awareness of climate issues and familiarity with arguments about the ZCA rose to unprecedented levels during this period.
Maintaining political and business support for the ZCA nevertheless required the government to articulate a spectrum of messages. During Parliamentary readings, normative ideas about the moral case for action were combined with messages stressing the ZCA's economic and social benefits and the government's commitment to a fair and equitable transition. This attempt to 'cover all bases' reflected important choices between 'firstbest' idealism and 'second-best' pragmatism. Continuing to focus on the threats of climate change limited the scope for all-out attacks on the legislation but, if used to excess, could (and did) lead to accusations of marginalizing alternative concerns (New Zealand Parliament, 2019b). Conversely, economic and equity framings demonstrated sensitivity to these concerns but exposed aspects of the ZCA to criticism and renegotiation. Striking a balance between normative framings that provide the motive force for action and cognitive framings about how action should proceed can thus be critical to maintaining long-term cooperation and achieving CCA goals. They also highlight the need for governments to accommodate and assimilate alternative ideas rather than attempting to win debates by 'hard-selling' preferred ideas, exerting power through and in ideas rather than over ideas.

Compromise
The critical challenge for Generation Zero during its campaign was to engage its discursive abilities to build decisive support among political parties, business, and society for integrated, long-term climate action. It exerted considerable power through ideas as it constructed and communicated cognitive and normative rationales for ambitious climate action based on climate science, international exemplars, programmatic messages about the risks of climate change, and appeals to New Zealand's moral obligations.
While this aspirational messaging allowed Generation Zero and, later, entrepreneurs within the government to broaden support for the ZCA, they also made strategic assessments of political feasibility that led them to modulate their preferences. In particular, they decided to adopt the 'two-baskets approach' early in the process, not because they wanted less ambitious targets for short-lived GHGs, but to diffuse future challenges to the ZCA from the agricultural sector and National Party by recognizing the different physical properties of these gases (Interview). This attempt to mitigate a potential obstacle to the ZCA indicates CCA entrepreneurs' sensitivity to endogenous constraints on their capacity to use discursive skill to galvanize New Zealand climate policy. By modulating their discursive approach, Generation Zero and government negotiators sought to reconcile their motivating commitment to economy-wide emissions reductions with the pragmatic need for agreement. Whether or not the two-baskets approach was regarded as a 'second-best' alternative, the repercussions of this decision persisted throughout the subsequent process and arguably created the conditions for fractures in the political consensus to reform the ETS.
From one perspective, the fact that CCA entrepreneurs from both inside and outside government both saw legislation which did not differentiate between GHGs as politically unviable reflects the 'power in ideas' held by the agricultural lobby that became 'hardwired' into New Zealand policy-making and discourse when the entry of agricultural biological emissions into the ETS was deferred in 2008. Advocates for climate action in effect made the agricultural sector's main arguments for them in advance by emphasising agriculture's economic importance and the technical and economic challenges of reducing livestock and fertiliser emissions. However, the agricultural sector would undoubtedly have made these arguments anyway (Cooper & Rosin, 2014). By anticipating these concerns while stressing normative ideas about the need for action and cognitive ideas about protecting agriculture, climate policy entrepreneurs avoided entrenchment by inviting and challenging agriculture to work cooperatively on climate issues. So, although the two-baskets approach left difficulties in negotiating legislation to reduce agricultural emissions, it equally illustrates policy entrepreneurs flexing 'power in ideas' in the hope of facilitating the institutionalization of new ideational structures and expectations in New Zealand climate policy.

Conclusions
One assessment of the ZCA might be that developing the political agreement needed to pass and sustain the legislation necessitated strategic avoidance of difficult political challenges. The argument continues that, where points of contention exist, there is a heightened risk that political agreements remain fragile, and that interestdriven politics will, sooner or later, resurface to challenge the intentions and mechanisms of CCAs. This remains an open question. The policy consensus that enabled the UK CCA, for example, was tested during debates over setting of the fourth carbon budget, which was only confirmed after a review, but has remained generally secure (Averchenkova et al., 2020;Gillard, 2016;Lockwood, 2013. One reading is that the New Zealand ZCA has been made more vulnerable by the strength of its links to more contentious policies like the NZ ETS. Either way, the contemporary outcome of historical processes in New Zealand has been a normative and cognitive environment in which agricultural concerns continue to exert a strong influence on policy ideas and discourse. Difficulties exist in openly challenging such embedded truths, even for skilled policy entrepreneurs, and in such situations, incumbent industries can, to an extent, rely on 'power in ideas' to fend off challenges to their exceptionalism. Path dependency, from this viewpoint, appears to create severe obstacles to policy change. The critical ingredient of CCAs, however, is a refashioned version of path dependency (Levin et al., 2012;Rietig & Laing, 2017;Rosenbloom et al., 2019). Like other CCAs, the ZCA establishes principles and obligations for future climate policy-making and mechanisms for achieving and monitoring them (Nash & Steurer, 2019). As a result, CCAs create pathways for changing economic interests, or how political actors interpret and express those interests, that it is hoped will generate greater long-term support for strong climate action. In terms of DI, this involves developing strategies for deepening consensus over time, moving progressively from the normative, philosophical idea of climate action to the programmatic concept of framework legislation, and ultimately the cognitive ideas that underpin future agreements to achieve the ambitions of CCAs.
Compromise is an indispensable part of the policy entrepreneur's toolkit for achieving such shifts, especially where the objective is to re-define what is politically 'possible' over the long term. The power of compromise resides in the idea that something imperfect but 'done' is preferable to a perfect idea that never materializes. The degree to which contentious issues can be recast (and ideational power redistributed) in future policymaking as a result of the ZCA remains unclear, as does the potential for political hostilities to resume (Gillard & Lock, 2017). CCAs seek to turn climate politics away from outcomes where the compromises needed to reach agreement fatally undermine effectiveness. This contribution nevertheless indicates that navigating between ambition and durability remain central to theoretical and practical understandings of the politics of CCAs. A fruitful direction for applied policy research is therefore to develop insights that can inform climate policy entrepreneurs' decisions about when, and when not, to compromise. Notes 1. The Treaty of Waitangi was signed by representatives of the British Crown and Māori Chiefs in 1840 and has a significant role in framing political relations between the Maori population and the New Zealand Government. 2. GLOBE-NZ is a chapter of GLOBE International, a non-profit entity founded in 1989 to advance cooperation on climate change and sustainable development. 3. Only the libertarian party ACT argued that New Zealand was incapable of contributing to meaningful change through its actions (Inderberg & Bailey, 2019). 4. In another demonstration of coalition politics, this decision also included 95% free allocation of ETS allowances to honour the coalition agreement between Labour and NZ First, and processor-level obligations on livestock emissions from 2025 if farmlevel obligations were not implemented. In return, NZ First moved away from its 2017 election commitment to replace the ETS with a CCA, indicating another compromise by its leadership. 5. The ETR Bill also introduced forestry reforms, reflecting forestry's importance to New Zealand's climate strategy.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This work was supported by Economic and Social Research Council [Grant Number ES/T000600/1].