Unravelling the United Kingdom ’ s climate policy consensus: The power of ideas, discourse and institutions

As climate change policies and governance initiatives struggle to produce the transformational social changes required, the search for stand out case studies continues. Many have pointed to the period between 2005 and 2008 in the United Kingdom as a promising example of national level innovation. With strong cross-party consensus and a ﬁ rst-of-its-kind legislation the UK established itself as a climate policy leader. However, early warning signs suggest that this institutionalised position is far from secure. Through a novel application of discursive institutionalism this article presents a detailed analysis of the role of ideas in unravelling this ambition under the Conservative-Liberal coalition administration (2010 – 2015). Discursive interactions among policymakers and other political actors were dominated by ideas about governmental responsibility and economic austerity, establishing an atmosphere of climate policy scepticism and restraint. By situating this conspicuous and in ﬂ uential process of bricolage within its institutional context the importance of how policymakers think and communicate about climate change is made apparent. The power of ideas to in ﬂ uence policy is further demonstrated through their cognitive and normative persuasiveness, by imposing over and excluding alternatives and in their institutional positioning. It can be concluded that despite innovative legislation, institution building and strategic coordination of different types of governance actors the ideational foundations of ambitious climate change politics in the UK have been undermined.


Introduction
Global environmental issues such as anthropogenic climate change continue to be a significant political agenda at multiple levels, not least for nation states. As inter-, multi-and transnational environmental governance initiatives proliferate, the domestic politics of climate change leaders (and laggards) has become of wide interest (Andresen and Agrawala, 2002;Fankhauser et al., 2015;Liefferink et al., 2009;Schreurs and Tiberghien, 2007). In 2008, the United Kingdom (UK) government passed the Climate Change Act (CCA), a first-of-its-kind legislation legally binding the UK to an ambitious greenhouse gas emission reduction target of 80% of 1990 levels by 2050. The implications were significant, institutionalising climate change as a political issue within the UK but also diffusing its ambition and policy framework to other contexts (Gummer, 2014;Hill, 2009;cf. Pielke, Jr., 2009).
The political and institutional circumstances surrounding UK climate politics and the CCA have since received a great deal of attention (Bowen and Rydge, 2011;Carter, 2014;Carter and Jacobs, 2014;Lockwood, 2013;Lorenzoni and Benson, 2014). This article presents a critical and detailed analysis of changes in the way climate change is thought about and discussed since that heyday. Between 2010 and 2015 a Conservative-Liberal coalition government was responsible for continuing this ambitious climate policy agenda within the context of a global and national economic recession. Given the tendency of policymakers to backtrack, or stall, on previous commitments during difficult political and economic periods Howlett, 2014) this is a timely moment to ask: what happens to the underlying ideas and does it matter?
The Stern Review's (Stern, 2007) presentation of early climate change action as economically rational was pivotal in the UK case (Carter and Jacobs, 2014). Also, the idea of five-yearly carbon budgets to keep successive governments on track, and accountable, to the 2050 target was important (Bows et al., 2006). Discourses of low-carbon business opportunities, correcting previous policy failures and a moral sense of urgency helped to secure support from private actors, policymakers and civil society respectively (Lorenzoni and Benson, 2014). Some of these ideas and discourses were formally institutionalised in a government Department for Energy and Climate Change (DECC) and a semi-independent Committee on Climate Change (CCC). As a result, energy and climate goals became entangled through an increasingly complex mix of instruments, discourses and strategies designed to simultaneously achieve low-carbon, secure and affordable energy-known as the 'trilemma' (Kern et al., 2014;Rogers-Hayden et al., 2011).
Despite the formal nature of this institutionalisation, doubts about its longevity have been raised (Lockwood, 2013), especially as the 'competitive consensus' among political actors supporting it quickly fell away after the 2010 national election (Carter, 2014). In response to these warnings, and also to calls from climate policy innovation scholars for more research into post-adoption complexities (Jordan and Huitema, 2014), this article tracks the nature and impact of subsequent changes in the ideas and discourses of UK climate politics.
In Section 2 the merits of a constructivist approach to studying political ideas and institutions are outlined. Section 3 summarises the methodology and case study materials. Section 4 shows how economic rationality and the normative positioning of government remained important ideas but that their initial emphasis on early action and leadership shifted under the strain of austerity. In light of these findings Section 5 raises concerns about a consensus approach to climate change politics and explores the political institutional context in more detail. Section 6 concludes with reflections on the analytical framework's contribution to the study of ideas in policy and the UK case study's relevance for climate change politics and governance in other contexts.

Ideas are more than just another variable
Political science, policy analysis and governance studies have all increasingly sought to account for the importance of ideas and discourse in shaping political processes (Fischer, 2003;Gofas and Hay, 2010;Hajer and Versteeg, 2005;Hajer and Wagenaar, 2003;Kütting and Lipschutz, 2012;Yanow, 2000). This critical turn can be seen in prominent theories of the policy process where only a handful of staunchly empiricist approaches continue to ignore or black box issues of subjectivity (Cairney, 2011;Sabatier, 2007). Reactionary attempts have been made to incorporate ideas as one more controllable variable within a positivist philosophy of science (regarding policy see: Pawson, 2006) so as to 'not have to swallow the contaminated epistemological water of postmodernism in order to enjoy the heady ontological wine of constructivism' (Keohane, 2000: 129). However, far from treating ideas as freefloating epiphenomena, many constructivists have explicitly linked them to traditional political entities such as institutions, interests and policy change (Béland and Cox, 2011;Hajer, 1996).
The implication of only partially considering ideas is apparent in some of the research on environmental policy innovations and institutions mentioned above. For example, Patashnik (2008) and Pierson (2004) both subsume the role of ideas under an explanation of politics as path-dependent, thereby failing to fully grasp their diversity and potential for driving change. However, these political realist accounts tell us little about the ideational and discursive processes through which these actors interact and through which climate policies are enacted or undermined. To illustrate, Lorenzoni and Benson (2014) compare such an approach with the more constructivist discursive institutionalism (DI) framework, demonstrating the latter's ability to explain the influence of climate economics ideas and the discursive interactions among civil society, politicians and business leaders that produced near unanimous support for the CCA.
This article extends the application of a constructivist approach, and DI in particular, to present a comprehensive account of how ideas and discourse have continued to shape UK climate politics over time. Accordingly ideas and discourse are treated as particular forms of power and political processes differentiated from, but interrelated with, other forms such as laws, institutions and structures. Within the language of DI, Carstensen and Schmidt (2015: 4) define ideational power as 'the capacity of actors (whether individual or collective) to influence actors' normative and cognitive beliefs'. This is done through three observable processes: the persuasion to accept and adopt certain views (power through ideas), the imposition of ideas and exclusion of alternatives (power over ideas), and the production of subject positions as well as the constraining of what can be legitimately considered (power in ideas) (ibid.). Given this analytical depth the intention is not to simply claim that 'idea A caused policy B' but to offer a more qualitative account of how ideational elements affect the way actors interpret, influence and enact climate policy.

Rethinking institutionalised ideas
Adopting a broadly Habermasian understanding of discourse as communicative action (Habermas et al., 1990) DI focuses on the interactions between actors and the ideas they carry, thereby reducing the emphasis on entrenched formal structures found in other schools of institutionalist thought. This sensitivity to interpersonal dynamics makes it a suitable framework for analysing the post-adoption politics of the CCA, where policymakers and other actors begin to negotiate their preferred pathways towards implementation. It is in these personal and micro-political exchanges that climate policy ideas are re-formed, supporting or disrupting the achievement of long-term targets.
Following Schmidt (2008Schmidt ( , 2010 the analytical components of DI can be clearly defined to produce a framework incorporating ideas, discourses and institutions (see Table 1). Ideas fall across three levels ranging from implicit values (philosophy) to general assumptions or principles (program) and specific solutions (policy). 1 For example, hidden social norms as well as more explicit assumptions about the scientific, economic or cultural nature of climate change are all as important as the practical actions of risk assessments and carbon budgeting. Further, two types of overarching ideas are particularly adept at tying together these three levels: normative ideas that provide prescriptions by linking values to appropriate courses of action and cognitive ideas that guide analysis by appealing to prevalent logics and interests.
Put simply, discourse refers to the 'exchange of ideas' among actors (Schmidt, 2011: 56). These interactions take a variety of forms (e.g. myths, stories, and scenarios) but their common goal is to represent ideas. There are two types of discursive interaction: coordination among actors responsible for developing policy and communication between these actors and other, less centrally placed, political stakeholders. The relative importance and influence of these types of discourse is partially determined by the institutional context. In a simple, or unified, polity communicative discourse will be most prominent as policy actors make decisions centrally and then seek to justify them whereas in a compound, or dispersed, governance context coordinative discourse is more pronounced as multiple actors are involved throughout the policy process. Lastly, the particulars of the institutional context (e.g. expected logics, patterns and audiences) need to be adequately addressed by ideas and discourses if they are to be influential.
Having been applied in numerous policy areas at national and international levels DI has become established alongside, but also differentiated itself from: rational-actor, historical and sociological variants of institutionalism (see : Bell, 2012;Peters, 2011;Schmidt, 2010). Perhaps the most obvious and popularising difference is its emphasis on change rather than stability. By stressing the dynamic, as opposed to deterministic, nature of institutions DI has leant itself almost exclusively to studies of moments of change (although: Hope and Raudla, 2012). Ironically, unless more longitudinal studies are undertaken, this may serve to reinforce the view of the policy process as sequential and equilibrium seeking i.e. there is stability, then a radical change that is studied in detail, then a return to stability.
In order to increase the DI framework's sensitivity to incremental changes, the notion of ideational bricolage as theorised by Carstensen (2011) can be instructive. Herein ideas are understood as being comprised of a web or related elements of meaning whose presence, linkages, and relative importance is prone to change over time. These mutations are incremental enough to be overlooked, or even deliberately downplayed, but their cumulative effect and influence on policy can be significant (ibid). By combining this with Schmidt's equally dynamic conceptualisation of ideas and institutions it becomes possible to see how the different elements of meaning contained within normative and cognitive ideas are situated and evolving across three levels.
In Fig. 1 overarching cognitive/normative ideas are represented as being comprised of smaller ideas spanning the three levels. Reinterpreting these as a web of elements of meaning enables a more detailed description of their linkages and stability/changes. For example, over time from t1 to t2, a given element may stay the same (e1), its relative prominence may change (e2 shrinks), or it may be completely replaced by an alternative (e3 becomes e4). Mapping this web onto the three levels of DI offers a useful typology for defining the elements of meaning and showing their distribution.

Methods
Discourse analysis was carried out on transcripts from a set of original interviews carried out between May and September 2015. In order to address the research questions most efficiently interviewees were selected for their close proximity to UK climate policymaking and politics circa 2008-2015. In line with the DI framework interviewees were purposively selected for their status as either policy actors (involved in designing and elaborating  Links between (political) ideology and climate change.
Program-assumptions and organising principles that define the problem and its solutions.
Debates within government about how to define and respond to climate change. Policy-specific solutions to specific problems.
The CCA, carbon budgets and related policies. Type Normative-connect the three levels by referring to values and appropriateness.
How UK climate politics challenges or conforms to social expectations. Cognitive-connect the three levels by appealing to prevalent logic and interests.
The rationalities of science, economics, politics, culture etc.
How climate change is presented in the UK and how this shapes the message. Interaction Coordination-among policy actors to: create, elaborate and justify. The mediating role of entrepreneurs and the allegiances of different actors. Communication-among political actors to: present, deliberate and legitimize.
Strategies, media coverage, public engagement and the transmission of messages about climate change.

Institutions Form
Simple-unified governance. Climate policymaking at Parliament and its style of engagement with other actors throughout the process. Compound-dispersed governance. Specifics Convincing messages follow expected patterns, are logical and are pitched at the right audiences in the right ideational setting.
How climate ideas and discourses differ within government and how they interact with their institutional setting.
policies) or more general political actors (involved in deliberating and legitimising policies). Following Yanow (2000), individuals from different types of organisations regularly involved in climate policy and politics were recruited to ensure a representative range of perspectives (see Table 2). Interviews with such policy élites are a useful and valid tool for analysing otherwise hard to reach aspects of the political process such as decision makers' beliefs and values (Beamer, 2002). Given the exclusive nature of policy networks a critical snowball recruitment method (Noy, 2008) was used, relying on the notion of 'reputation' to recruit key individuals (Farquharson, 2005). To avoid reverting to a narrow definition of reputation as merely authoritative, interviewees were also asked to consider 'thought leaders', prominent non-state actors, and individuals with insightful positions. Transcripts were analysed for evidence of types of ideas, accounts of discursive interactions and descriptions of institutional contexts. This approach was applied to the interviewees' explicit accounts as well as to their more implicit and indirect utterances (e.g. underlying philosophical principles and assumptions). In other words, texts were analysed for both their content and function as meaning-making discourses. The same analytical approach was applied to relevant publically available documentary data sources (e.g. policy documents, government reports, public statements, and media articles) thereby enabling a triangulation of the findings. Although the emphasis of DI and this article is on interactions between actors and thus on verbal accounts, documents were particularly illustrative with regards to communicative discourses. Here texts were primarily treated as discursive practices intended to present and legitimise certain ideas about climate change through policy.

Shift in the institutionalisation of climate change ideas
Prior to the general election in 2010 climate change had rapidly risen in political salience internationally and in the UK. Described as a 'punctuated equilibrium', this attention to environmental politics and the risk of dangerous impacts from climate change highlighted the inadequacy of existing policies to reduce greenhouse gas emissions (Carter and Jacobs, 2014;Lorenzoni and Benson, 2014). Spurred on by private sector endorsement of the business case for a low-carbon economy and a vociferous civil society campaign UK politicians engaged in a 'competitive consensus' to adopt a world-leading climate policy framework in 2008 (Carter, 2014). However, as the recession dragged on and political attention turned away from climate changeespecially after the 2010 electionthe cost of implementing climate policies became a contentious issue (Carter, 2014;Ares, 2011). As Lockwood (2013) has suggested, the political sustainability of the CCA began to look less assured. Changes to the types of ideas and discourses associated with climate policy between 2010 and 2015 are a key indicator of this uncertainty and its effects on future efforts to reduce emissions.

Philosophy
Given the dominance of neoliberalism and deregulation in UK politics it is perhaps unsurprising that the philosophical level of ideas was populated with concerns about the failure of a centralised form of government to address the physical and social reality of climate change. Whilst not shared by all, these concerns were part of a wider modernist trust in the potential of technological innovation and economic liberalism to provide solutions. Similarly, the failure of any mainstream political ideology to convincingly relate to, and articulate, the social complexity of climate change was also mentioned. For example, one interviewee noted reluctance among the political élite to confront 'the confusing and incoherent fragments of postmodern politics' (CS1), suggesting that most intuitively attempt to depoliticise climate change whenever possible.
Overall, the majority of ideas identified were programmatic, with some being linked to specific policy solutions such as setting carbon budgets and conducting risk assessments. Broadly speaking they followed a logic of either 1) climate change is X and therefore governing it requires Y or 2) climate policies are bad/good for the economy. Already this suggests a predominance of political and economic, rather than social or cultural, rationalities for addressing climate change. At a general level this is an example of power in ideas, or the institutionalisation of certain ways of thinking that requires policymakers to base decisions on technical forms of evidence.

Program
Based on a belief in neoclassical rational-actor principles the first set of ideas often defined climate change as the archetypal 'tragedy of the commons' writ large. Assumptions about the need for top-down targets and principles of good governance were quick to follow. Whilst common, this view was often accompanied by a resigned acceptance that there has been a rapid loss of political appetite for state intervention. For instance, climate policymakers felt their range of options to be significantly curtailed by the newly introduced Better Regulations Framework (BIS, 2013), another example of the exclusionary power of ideas. The new guidance applied to all government officials and institutionalised a 'one-in two-out' rule based on an 'incredibly narrow minded view of regulation as a purely negative burden' (PM5). In practice this means flagship climate regulations and taxes were significantly weakened, as in the cases of the Emissions Performance Standard for fossil fuel power generators which 'was completely neutered' (PS1) during its development, and also the Carbon Price Support policy which was frozen within a year of being introduced (HMRC, 2014).
Second, the economics of climate change remained a provider of best practice principles for climate policy. Central to this set of ideas is the cost-benefit ratio argument made by the Stern Review, stressing the cost effectiveness of early mitigation action (Stern, 2007). Among policymakers a shorthand heuristic for these ideas was the principle of 'maximum return on investment', which was favourably seen as helping government to function more like a business and as wholly sensible given the adage that government never has enough money to do everything it wants to. By late 2014 the long-term relationship between the economy and climate change had not been invalidated but had been replaced by the more politically salient short-term relationship between the economy and re-election.

Policy
Relying mostly on economic assumptions many climate policymakers, including ministers at DECC and the Department for Business, Innovation and Skills (BIS) continued to try and bolster this early action imperativeciting positive feedbacks, investor confidence and infrastructure development as part of a long-term planbut ultimately the short-term demands of saving public money eclipsed them. The overbearing presence of austerity even led policymakers to actively avoid some solutions; being told directly that 'if it costs money it isn't going to happen' (PM1). More than just the coercive power over ideas embodied in the Treasury's ability to veto policy suggestions, the austerity agenda created a restrictive atmosphere within which policymakers pre-emptively excluded certain ideas based on an internalised assumption that only efficiency gains and 'doing more with less' types of policies would be deemed viable. Whilst the range of specific solutions for reducing emissions was dwindling policymakers were able to point towards carbon budgets and risk assessments as evidence of their continuedalbeit strategic or even symbolicactions on tackling climate change.

Losing the normative mandate to lead
Around the time of the CCA (2005-2008) a strong normative position for the UK as a climate leader emerged. This was based on a moral imperative to act urgently in the face of dangerous climate impacts and to reap the benefits of correcting hitherto failed policies (Lorenzoni and Benson, 2014). Since then, these elements of meaning have significantly diminished and been joined by others to produce a shift in the overarching normative idea about the government's role in tackling climate change (see Fig. 2).
As predicted by Carter (2014), the sense of urgency and momentum associated with the Big Ask civil society campaign and the competitive consensus among politicians dissipated after the national election. The prominence of morally-based ideas also shrank due to previously abstract notions of dangerous impacts being redefined as specific manageable risks (see Fig. 2) via the UK Climate Change Risk Assessment programme and policies (DEFRA, 2012). However, thanks to deep-seated democratic values of governmental duty and accountability this norm didn't dissipate entirely. The idea of target setting also remained, enabling politicians to point to the flagship CCA policy and carbon budgets when challenged with critiques or when positioning the UK as a climate leader on the international stage.
Two new additions to the milieu of meanings fundamentally altered the way the idea of leadership could be expressed. As climate change's political salience waned the meaning of leadership became more about acting responsibly than seeking out firstmover opportunities. As a result, the direct benefits of mitigation and adaptation were no longer enough to justify action, there needed to be co-benefits too. One policy actor recalled: ' possible. This was exemplified by the Coalition's Big Society agenda (Cabinet Office, 2010), which extolled the innovative problemsolving potential of shared responsibility and non-state actor empowerment. Previously civil society and industry had vociferously pressurised government to lead but they were now being asked what they were going to do.

Whose numbers count most?
Economics continued to be the most prevalent cognitive idea through which to grapple with the complexity of the climatesociety relationship. For instance, in successive reports and recommendations to Parliament the Committee on Climate Change (e.g. 2008Change (e.g. , 2010Change (e.g. , 2013 deployed economic analyses to directly appeal to: the perceived long-term vision of politicians seeking to leave a legacy, publics concerned with infrastructure investment and low-carbon enterprises looking for a stable business trajectory. However, through the adjustment of metrics and assumptions economic logic can be coherently applied in support of very different policy recommendations. This is an epistemological situation that policymakers were acutely aware of and which was built into their political strategies: 'You've got three departments [DECC, BIS, Treasury] all doing analysis on the same questions just because they don't trust each other. For example, the Treasury focused more on short-term values and so off-shore wind was hated whereas DECC saw it as a gamble for the future.' (PM4) As austerity loomed large, economic discourses began to paint a different picture to the one presented by the Stern Review. Concerns were raised about reducing public spending, protecting energy intensive industries and maintaining competitiveness with the rest of Europe. In 2013/14 arguments around reviewing the fourth Carbon Budget presented a visible example of these competing rationales. A chorus of support to be as ambitious as economically possible swept through the first three carbon budgets (in 2008) but, in contrast, adopting the fourth (in 2011) required a direct intervention from the Prime Minister and the addition of several caveats including a chance to review and revise it. Although it was ultimately upheld this challenge to the validity of the fourth carbon budget, and the CCC's advice more generally, has set a precedent for future caveats and reviews of ambition.
From the policymakers' perspective (self-labelled as pragmatic), pursuing long-term climate goals and low-carbon investment was overtaken by the imminent and interdependent need to repair the economy and get re-elected. In political terms this led to prioritising the interests of consumers/voters over those of the low-carbon sector e.g. reducing energy prices by cutting subsidy levies. In simple economic terms the discount rate proposed by Stern for calculating cost-benefit ratios into the future was rapidly adjusted to favour the present. Or, to put it another way, the political economy of climate change ideas went from prioritising long-term diffused benefits to being primarily about short-term concentrated costs (see Fig. 3).

Normative and cognitive cross-over and other ideas
Overall the various elements of meaning spanned all three levels of ideas in much the way that Schmidt (2010) anticipated. That is, with very little explicit reference to philosophical level values (e.g. political philosophy or ideology), and an abundance of programmatic level organising principles (e.g. leadership, target setting, markets and subsidies). It is also important to note that developments in the cognitive and normative ideas described above did not happen in isolation from alternatives or from each other. For instance, arguably accelerating the decline of a government leadership norm was the resurgence of the powerful cognitive idea of scepticism. Although climate science scepticism had become a marginalised position the more moderate and politically acceptable position of 'climate luke warmism' -that is, accepting the science but doubting the likelihood of severe impacts emerged. When coupled with the belief that government intervention is inefficient in areas of unpredictability or long-term planning this amounted to a form of 'climate policy sceptism'. Further, the notion of co-benefits appears in both types of ideas suggesting it to be a pertinent and malleable element; for instance several policy and political actors saw it as vital for keeping climate goals on the government's agenda and as the only possible way to secure political and economic resources during the recession.

Turning down the volume on climate change discourse
Despite speaking with a range of different actors, ideas about climate change appeared in a limited number of discursive forms. In the policy sphere, meteorological and economic scenario modelling was treated as the foundation of climate change knowledge. Repetitive policy scripts about tackling the energy trilemma and abating risk were then used to explain what was being done. Even among wider political actors and stakeholders these forms were the most prominent. Though in addition, there were some polarised protagonist/antagonist driven narratives recurring in the media that functioned more as awareness raising devices than as a means of elaborating or deliberating responses to climate change. Given the evidence base and its focus on public policy (where scientific and economic evidence-based justifications and good versus bad plotlines are known to resonate loudest) this narrow range of discursive representations is perhaps unsurprising.

A Consensus built on shallow foundations
The day-to-day political interactions within, and along the fringes of, government are such that a summary of these discursive practices can only ever be a snapshot. As has been shown, discourses of climate governance and economics were continuously present, although not necessarily stable, over time. They often produced boundary work 2 (e.g. annual reports, scenario models and framework policies) to help coordinate different types of actors around a particular set of ideas. A mixture of ideational coherence, strategic enactment and institutional context mediated their ability to coordinate actors around specific ideas. Table 3 lists some examples of the discourses used by interviewees to elaborate and justify their preferred policy options as well as the institutional context in which they were deployed and debated with various actors. Whilst they are not exhaustive or entirely representative these examples do provide insights into common themes and the importance of institutional venues and the ideas that permeate their boundaries.
Notwithstanding more detailed disagreements, the CCC and DECC consistently aligned behind an overarching discourse of positive climate economics and necessary government involvement. Generally, support for this position from politicians was unreliable. Cabinet ministers and party leaders were keen to engage in high-level rhetoric e.g. around approving the fourth carbon budget and signing a 'joint pledge' during the 2015 election to remain committed to tackling climate change. However, they were not so forthcoming on specific climate policy solutions that could drive a rapid low-carbon transition such as a power sector decarbonisation target or domestic energy efficiency. In the case of the latter, underperforming policies (e.g. the Green Deal) resulted in a big setback to the UK's attainment of long-term emissions reductions (CCC, 2014).
Both of the CCC Chairmen (Lord Turner, previously Director General of the Chamber of British Industry and then Lord Deben, previously Chairman of the Conservative Party) were able to keep potentially hostile business leaders and Conservative politicians on side. Similarly, the two Secretaries of State at DECC (Chris Huhne and then Ed Davey) fought effectively to make the economic case for climate change a more visible priority and enlisted the help of other departments (e.g. working with BIS to establish the Green Investment Bank) to mainstream climate goals across government. Despite these coordinative efforts the persuasive power of climate change economics had limitations. For instance, a significant countervailing pressure came from the Treasury who took an increasingly hostile stance toward climate policies, occasionally aligning with anti-regulation politicians and high carbon industry actors. This coordination was built around a salient counterdiscourse of climate policy as expensive but also politically dispensable À succinctly expressed in the Prime Minister's often quoted dictum to 'get rid of the green crap' (Schofield, 2013).
Here the importance of institutional constraint/enabling is particularly clear. First, DECC struggled to reconcile internal inconsistencies to do with both policy (trilemma goals) and politics (as a Liberal Democrat led department in a Conservative majority coalition). Second, the CCC was formally mandated to provide advice based on scientific and economic rationalities only. Whilst their reports were well received by most type of actors they were restricted in their ability to engage in overtly political, and therefore potentially more effective, coordinative discourses. And thirdly the Treasury, due to its institutional position as fiscal overseer and the pervasive discourse of reducing public spending, was able to occupy a contrary position vis-à-vis ambitious climate policy without fear of being isolated or losing influence. As one interviewee quipped: 'If you work for the Treasury it is your job to be disliked so we should be reasonably understanding, but they have been increasingly overreaching.' (TT5)

Telling a compelling non-story
It is important to note that the communicative interactions analysed here took place among actors with special interest in climate change but who felt that it commanded little attention or value in wider society. In fact, many actorsincluding central government and mediadefined climate change as a 'non-story' Table 3 Discursive interactions and their institutional context.

Discourses Contexts
High-level commitments to climate change such as target setting were generally supported but detailed and potentially divisive decisions were seen as the responsibility of more decentralised processes and actors.

Central Government
Positive and long-term economics backed up policy decisions and linked them to high-level targets but this was hampered due to inconsistency caused by competing policy goals and Coalition politics.

DECC policy
The economics of emissions reductions and the science of climate risks were the foundations for long-term planning but were incongruent with dominant ideas about neoliberal governance and austerity.

CCC consultations
Austerity, small government, and climate science luke warmism produced an atmosphere of climate policy scepticism. Treasury negotiations Leadership through cross-party consensus building and appealing to positive investment storylines.
NGO and think tank engagement Uncertainty was a central theme for both low carbon and industrial actors as they appealed to policymakers for consistent economic signals and policies.
Private sector statements 2 Based on Gieryn's (1983) original description of the way knowledge can be used to demarcate areas of expertise for ideological reasons. and made a strategic decision not to make it a central theme. This is a clear example of one idea's power over the alternative view that climate change could be an important political issue. Not only did this limit the overall amount of communicative discourse but it also steered what little there was towards a particular view of climate change as just another issue among many rather than as something cross-cutting and fundamental. Six communicative discourses and the way they deliberated and legitimised certain climate programs or policies were identified (see Table 4). Again, governance and economics ideas were prominent throughout. Emerging out of this assortment were two competing meta-discourses: 1) the UK is leading on climate change through a long-term and positive economic plan, contra 2) disagreement over energy and budget priorities is undermining the UK's response to climate change.
The UK's leadership role in climate change governance lost its public appeal quickly after 2008, retaining relevance only for a fairly narrow set of actors engaged in international climate negotiations at the United Nations Framework Convention on Climate Change (UNFCCC). A telling sign of its diminished influence was the relative absence of climate change from the electoral campaigns in 2015, especially given their coincidence with a milestone UNFCCC conference of the parties in Paris (COP21). In fact, by signing a joint pledge, party leaders were able to gloss over the fact that recent policy changes had knocked the UK off course to achieving its longer-term targets (CCC, 2014).
Long-term planning and positive climate economics were central pillars of DECC's communication strategy and were spelled out narratively through the language of carbon budgets, transition pathways and renewable energy roadmaps (e.g. DECC, 2009DECC, , 2011aDECC, , 2011b. However, policymakers reported getting little traction beyond the usual low-carbon sector firms and environmental NGOs. Controlling the policy message is a vital form of power over ideas for central government but one that was largely unsuccessful in this case. Instead, aided by messages from the Treasury, some media organisations and climate policy sceptics reverted to a simplified economy-environment trade-off characterisation in order to cast doubt on the government's capacity to stay the course as a climate leader during constrained economic times. Open calls from policymakers for a complete rethink of the UK's leadership position and the CCA were few and far between (See speeches by: Osborne, 2011;Patterson, 2014). Indeed, postlegislative scrutiny conducted in 2013 (DECC, 2013) broadly reaffirmed the government's commitment to its level of ambition. However, a more measured but no less damaging discursive challenge to the 2008 consensus did gain ground. Climate policy sceptics within government were quick to amplify disagreements with the help of the media's tendency to condense complex issues into combative shorthand storylines e.g. by leaking letters showing disagreement between ministers over the fourth carbon budget (Stratton, 2011). Similarly, a complex set of policy instruments had been developed to reflect the energy trilemma's goals but much communicative discourse ignored these details, focusing instead on net public expenditure and green levies, thereby pitting DECC against the Treasury, consumers against environmentalists and politicians against their constituents' least favourite energy sources.

Institutional complexity and logical inconsistency
Interacting with these shifts in ideas and discourses were the political institutions responsible for climate policy and governance more generally. Elsewhere the UK parliamentary system has been described as a 'simple' polity, meaning that the unified administration requires a relatively small amount of coordinative interactions to reach a decision before putting most effort into communicating their justifications to the wider public and stakeholders (Schmidt, 2010). Yet, several findings from this study suggest this description to be inaccurate. First is the drive towards devolution. The devolved administrations of Wales, Scotland and Northern Irelandas well as sub-national city regionshave varying degrees of autonomy over their climate change programs and policies, often outperforming and contradicting those espoused in England. Secondly, inter-departmental factions were exacerbated both by the need to integrate climate goals and by Conservative-Liberal coalition dynamics. Thus the coordinative activities and evidence gathering, or environmental policy boundary work (Guston, 2001), of Junior Ministers, Select Committees, All Party Parliamentary Groups and Special Advisers increased significantly. Lastly were the complementary trends toward multi-actor governance, open policymaking and the mediatisation of politics (Hajer, 2009). Together they demanded more pluralistic coordination around climate change issues and undermined the authority of didactic state driven messages. Taken together such discord calls into question past claims about the institutionalisation of climate change in the UK: 'We have some climate legislation and a lot of climate related policy zooming around that Westminster bubble. But I don't think the UK has ever had a position on climate change. I don't think we have a polity in the sense of a public conversation that is a true reflection of our society. There is the Westminster bubble, which has rather hijacked the public conversation, but I wouldn't call that a polity.' (CS1) Further, there were concerns about the inability of various political and governance institutions' ability to adequately express, let alone respond to, the social complexities of climate change. The ideas of 2005-2008 about cross-sector and cross-party consensus in pursuit of ambitious climate goals proved, over time, to be incompatible with several entrenched institutional logics. In policy terms, the logic of neoliberalism continued to snub regulatory options while at the same time the logic of 'the Treasury View' -an example of power in ideas institutionalised since the 1930s recession, asserting that government spending offers no net Table 4 Communicative discourses and the climate related ideas they sought to express.
Discourse Summary UK as leader Government has a responsibility to pursue first mover advantages and influence other actors to address public good problems such as climate change. Long-term plan Target setting and the flexible use of policy levers can cope with the uncertainties of climate change and are necessary for dealing with the long-term impacts that reach beyond political cycles.

Positive economics
Climate change is an investment opportunity that brings co-benefits and political capital.

Dissensus
The complexity of climate change demands more rigorous political debate and a decentralised, market-driven approach to governance. Energy transition Subsidies and energy market signals can address climate goals but only if consumers are protected from significant costs and supplies remain secure.

Budgeting
Reducing public expenditure is a top priority therefore climate policies should be discussed in terms of efficiency and return on investment.
benefit to economic activity in times of austerityrestricted any fiscal options regardless of how generously climate economists discounted the future. In political terms, the cognitive power of heuristicas opposed to holisticthinking in government quickly returned climate change to its position as a background and nonvoter issue. In this context consensus without competition is tantamount to depoliticisation, as one climate policymaker recalled: 'If you are not seen to be having a fight about something then it is not seen as being politically significant. We got some attention around the Spending Review but interest in the run of the mill [climate policy] pieces was already waning when we came in [in 2010]. ' (PM5) 5. Discussion: the power of ideas and their impact on climate politics More than just identifying and describing various ideas about climate change, the findings have illustrated their linkages, relative prominence and alterations over time. The generalised norm of government responsibility for acting on climate change remained but was diminished, losing the positive leadership emphasis and failing to connect to either political philosophies or to specific policies (Fig. 2). As the 2015 general election grew nearer climate economics ideas shifted from investment and benefits to austerity and costs, replacing policy prescriptions for low-carbon sector prioritisation with wider economic co-benefits (Fig. 3). The power of ideas, as theorised by Carstensen and Schmidt (2015) was also clearly evident throughout. Coordinative interactions among policy actors relied on the persuasive power through ideas of economics, with institutionalised government divisions and austerity ultimately overriding climate policy commitments. Two types of power over ideas were apparent in the communicative interactions of actors: the exclusion of climate change as a central political story and the inability of central government to control, and thus impose, coherent climate policy messages. Lastly, the institutionalised power in ideas of economic rationality was further strengthened by the recession, leaving the Treasury and neoliberal governance advocates in a strong position to veto many climate policy ideas.
As economic recession and the election cycle wore on politicians appeared to revert to the meta-preference of 'reelection at all costs' that Bauer and Knill (2012) warn about in their work on policy dismantling. Combining these findings with the observed trends in public and media engagement with climate change in the UK and around the world (Boykoff et al., 2015) have confirmed Lockwood's suspicion that despite occasional peaks (e.g. the Big Ask campaign in 2008) there is insufficient pressure from voters to ensure politicians will shoulder the responsibility for addressing climate change as a stand-alone public good. It is perhaps not surprising then that environmental NGOs expended great effort in coordinating a joint pledge from the three main party leaders (Green Alliance, 2015) that simply restated previous commitments.
Whilst this strategy of cross-party consensus building may be based on theoretical (Giddens, 2009;Voß et al., 2009) and experiential (Carter, 2014) foundations, there is reason to suggest that it may not always be the most effective. Several interviewees expressed dismay at the lack of ambition and its inadvertent reinforcement of the idea of climate change as a 'non-story'-even during an election campaign period. However, this is not to suggest that the other extreme of polarised climate politics would be any more effectivewitness the slow progress on climate change of the United States administrationsbut that a certain degree of contestation is necessary to limit the scope for dismantling or inaction. A more moderate and pragmatic strategy of exercising power over policymakers by highlighting slow progress at a time of high accountability and visibility may be effective. Despite the previous success of this strategy in 2008, few voices referred to evidence that policy changes under the Coalition Government were undermining longer-term emissions reductions (CCC, 2014).
The cognitive idea of climate change economics moved in the opposite direction with regards to politicisation. The positioning of its component meanings changed very little (see Fig. 3) but the switching from investment to austerity as an underlying principle and the associated shift of emphasis from benefits to costs threatened the validity of certain policy ideas e.g. green energy subsidies and the levying powers of DECC. The impact of these tensions on climate policymaking was significant; reductions in the generosity of renewable energy subsidies were made in 2011 and Parliament contradicted CCC advice about setting a decarbonisation target for the power sector in 2013; both of which have hampered progress towards emissions reductions in a key sector. The review of the fourth carbon budget evidence base in 2014 was an example of policymakers outwardly questioning the wisdom of the CCC both in terms of its programmatic assumptions and policy solutions. A precedent has thus been set for Parliament to revisit its carbon budget commitments on economic competition grounds, potentially jeopardising the attainment of its targets. Overall, it appears the battle to persuade through the power of economic ideas was won by those presenting a view of climate policy as too costly and thus incompatible with deficit reduction and, by extension, re-election.
The case study findings are a reminder of the ideational influence of economics as a rationale for governing (Flyvbjerg, 2001). Particularly important was its cognitive role throughout the policy process (e.g. for visioning a future low-carbon economy and guiding decision making based on the principle of maximum return on investment) and the institutional positioning of economists (e.g. Treasury control over DECC's levying activity and the appointment of economically-oriented leaders at DECC and the CCC). This 'political influence of economics' (Hirschman and Berman, 2014) is not inherently counterproductive to climate change actionas the Stern Review (2007) provedbut it does limit the range of policy solutions that can be considered. This latter point is linked to concerns about the increasingly technocratic nature of climate and energy policy in the UK, especially regarding its negative effect on the political capacity of policymakers (Kuzemko, 2015).

Conclusion
Starting from the assumption that a prolonged economic recession would challenge the foundations of the UK's ambitious plans to tackle climate change, this article set out to explore the ideational, discursive and institutional dynamics through which this might be observed. Combining DI with a relational definition of ideas illustrates how, through discursive interactions, ideas underpinning institutionalized policy arrangements can be reconfigured. The impact of these changes on the policy process can be described as power through, over and in ideas. To summarise the UK case in these terms: austerity economics proved most persuasive among policymakers, the non-story status of climate change kept it away from the (potentially creative) frictions of politics, and institutional constraints ensured any attempt to increase climate ambitions were delegitimised. Thus, it has been shown that ideas continue to matter after the adoption of innovative policies and that the discursive interactions and ideational bricolage through which they evolve can be analysed to reveal exactly when and how they matter.
Overall, the UK government's role in tackling climate change transformed from innovative leadership to decentralised responsibility, questioning (and also rejecting) new targets for reducing emissions in favour of 'Big Society' driven alternatives. The economics of climate change continued to be a prominent cognitive idea tying together norms, principles and specific solutions. However, its previous rationale for early action was challenged by a shift in emphasis from investment to austerity i.e. from benefit to cost. Together these reformulated ideas were represented through increasingly popular discourses of macroeconomic prudence and climate policy scepticism. In addition to amplifying divisions among political actors these changes reinforced simplified narratives and stalled the progress of climate policymaking through restrictions on both thoughts and resources.
Contrary to the image of UK climate politics as safely institutionalised by cross-party agreement and flagship legislation, the 2010-2015 period was marked by an increasing dissensus. Some direct impacts on policy were apparent (e.g. in cuts to department budgets, to renewable energy subsidies and to energy efficiency schemes) although these were obscured under the general banner of austerity rather than as an effect of the UK abandoning its position as a climate leader. This is perhaps the most worrying implication of the findings, especially for other countries seeking to emulate the UK's climate policy framework and consensual political strategy. Strong disagreement among policymakers, as well as a growing shortfall in the required policies to meet long-term targets, were partially hidden from critique. Thus, target setting and rhetorical commitments may provide political cover for inadequate policy action on the ground. At the UNFCCC COP21 in Paris in 2015 the UK, and many other countries, presented themselves as leading the fight against climate change, however the gap between targets and implementation at the national level has been quietly growing.