An unexpected climate activist: central banks and the politics of the climate-neutral economy

ABSTRACT Recently, an unexpected public actor has joined climate activists’ forces: central banks increasingly care about and act on climate change. Their (potential) contribution to the climate-neutral economy is immense, as they possess vast institutional resources to influence investment in the economy. Central banks’ advocacy is puzzling, as they put their legitimacy at risk by stepping outside their mandates and area of expertise. To explore this puzzle, I study why and how climate change came on the agenda of the Dutch central bank, one of the first movers among central banks. By opening the black box of organizational politics within central banks and building on unique interview data, I show how the policy shift of the Dutch central bank can be traced back to a small group of policy-seeking staff. This process-tracing case study contributes to our understanding of transformative change in climate policy in particular and public policy more generally.


Introduction
'As a Baron von Münchhausen we initiated this platform, without anyone asking us, without a higher committee to report to, we just did it' Dutch central banker 1 Whilst countries around the world have been struggling for decades to transition to a climate-neutral economy and prevent irreversible damages to the planet we live on, many politicians, journalists, citizens, and interest groups increasingly care about climate protection. More recently, anotherunexpectedpolitical actor has joined their forces: central banks and financial supervisors (hereafter: central banks) have lately turned into climate advocates and set the first steps to change their supervision and monetary policy accordingly (NGFS, 2021). Central banks' (potential) contribution to the transition is considerable. A crucial component of the transition is (private) financial investment: there is need for a 'tectonic shift' of financial streams towards the climate-neutral economy (IPCC, 2018). Nowadays involved in a multitude of policy fields at almost every step of the policy cycle (Fernández-Albertos, 2015), central banks and experts claim that they possess vast institutional resources to influence the financial sector and thereby the allocation of credit toward the transitioning of the economy (NGFS, 2021;Schoenmaker, 2021;van 't Klooster & Fontan, 2020). The advocacy of central banks is puzzling, in the first place because usually climate goals are not (explicitly) part of their tasks and mandate (Dikau & Volz, 2021). Central banks consequently put their legitimacy at risk by stepping outside their mandate and areas of expertise (Johnson et al., 2019;Schmidt, 2013). Given this paradox, it is surprising that there are almost no studies on this topic. The few related existing studies explore what central banks do on the issue of sustainable finance and whether this is covered by their mandates (Dikau & Volz, 2021;Steffen, 2021), or how the institutional context (including limited central bank mandates) can explain why Europe is staying behind in ambitious financial climate policy compared to other jurisdictions (Baer et al., 2021), but not why and how climate change arrived on the agenda of central banks.
To explore this puzzle, this paper focuses on the first years of their advocacy, i.e., the phase in which the first central banks started to become involved (2014)(2015)(2016)(2017)(2018). In particular, the paper studies why and how the central bank of the Netherlands, De Nederlandsche Bank (DNB), changed into a climate advocate. As I will show in detail below, DNB is a crucial case, as it set the scene for the unprecedented advocacy of central banks together with other first movers (Bank of England and Banque de France). Moreover, the Dutch central bank is an interesting least-likely case because, compared to other first movers, DNB has a narrower mandate, possesses fewer resources, and is relatively conservative.
I argue and show empirically, based on interviews with central bankers that were directly involved within DNB and extensive document analysis, how a small group of policy-seeking staff started to push for climate action and was able to switch the bank's policy position. Above all, protagonists created new external 'venues' to build up support and expertise internally and 'framed' climate change as a financial risk to legitimize the involvement of a central bank (Princen, 2011). In contrast to the expectations from the traditional central bank literature, the government and the financial sector played a rather minor role (Fernández-Albertos, 2015). This paper makes four contributions: first, I highlight and show that central banks haveunexpectedlybecome important actors in the politics of the transition to a climate-neutral economy. Second, by opening the black box of central banking and providing unique first-hand insights into the organizational politics within DNB, I explain why and how this surprising shift came about in the crucial Dutch case. Third and more generally, the processtracing case study expands the literature on independent agencies, drawing attention to the importance of studying the use of stakeholder bodies by agencies and the policy impact of individuals within agencies. Fourth, by showing how a U-turn on climate policy came about, this paper contributes to our understanding of transformative change in climate policy in particular, as well as in public organizations in general, thereby adding to our understanding of the politics of public policy-making.
The next section of this paper gives a short introduction to the role of central banks and sustainable finance in the transition to a climate-neutral economy. The theoretical framework reviews the existing literature on central bank agendas and develops the main argument. Thereafter, the sections describe the research design and results, and the last portion of the paper concludes and discusses.
Sustainable finance and the sudden advocacy of central banks Central banks and the battle to stop climate change are connected through the role of the central bank in the economy and in particular the financial sector. In order to prevent global warming, the whole economy needs to transition to a carbon-neutral economy. This affects every sector, from agriculture to energy and from clothing production to housing. This transition, as any change in the economic system, needs financial investment, for example, to develop and consequently implement new technologies, to redesign and reorganize value chains, and to train employees in new skills. Scientists estimate, for instance, that to transform energy systems alone, 2.4 trillion USD in investments are needed annually between 2016 and 2035 (IPCC, 2018, p. 22). One of the most important investment sources in modern capitalist societies is private investment from the financial sector. So, in order to change the economy, the streams from the financial sector into the economy need to be (re)directed to sustainable purposes (Colgan et al., 2021;Steffen, 2021).
How are the actors in the financial sector, e.g., banks, investment funds, and insurers, to be activated? Here is where central banks come in: central banks have a wide array of powers to influence the financial sector. At the moment, more than 80 central banks have committed to this role by joining the Network of Central Banks and Supervisors for Greening the Financial System (NGFS), a global platform for central banks that support the transition (NGFS, 2021). In terms of actions, central banks have been the most active in their role as financial supervisors. As of 2020, almost 40 central banks had adopted guidelines for financial institutions on sustainable lending and/or management of climate change risks (Dikau & Volz, 2021). The European Central Bank (ECB) announced it is investigating whether banks' risk frameworks are in line with ECB's climate risk guidelines and, if they are not, will consider raising bank-specific capital requirements (Elderson, 2021). Furthermore, a majority of central banks take sustainable investment criteria into account in their own investment portfolios, for example, by excluding certain sectors (NGFS, 2020a).
When it comes to monetary policy, most central banks are considering making changes, but substantial differences in positions on what to do exactly remain in this area (NGFS, 2020b). The basic idea is that by being more selective in which assets they buy and which collateral they take, central banks can lower the funding costs for climate-neutral companies (Schoenmaker, 2021). At least two central banks are already taking actions: first, the ECB accepts sustainability-linked bonds as collateral and adopted an action plan to incorporate climate change into its monetary policy strategy (ECB, 2021); second, the UK government announced in March 2021 that it is changing the monetary policy mandate of the Bank of England to include climate goals (Hodgson et al., 2021). Some scholars are critical of central banks' climate policies, arguing, for example, that they include measures that promote greenwashing or do not actually fine the ownership of non-sustainable assets (Dafermos et al., 2021). However, this short overview shows that central banks are actually (in the process of) adopting policy measures that scholars have advocated for, e.g., a more adequate risk framework for supervision (as proposed by Campiglio et al., 2018), or more sustainable criteria for buying assets and taking collateral (as advocated for by Schoenmaker, 2021;van't Klooster & Fontan, 2020). 2 The advocacy of central banks for a climate-neutral economy is quite new. Until recently, nobody connected central banks with climate change, but after 2018 the salience of the issue increased rapidly. To illustrate, Figure 1 plots the frequency with which 'central banks' and 'climate change' were Note: The Twitter indicator is an 18-month, backward-moving average. mentioned in the Financial Times and on Twitter. However, and as I show in detail below, the first central banks in Europe had already started working on the issue in 2014. Not only was there little media attention at that time, but this was also before the Paris Agreement, the (subsequent) surge in public attention to climate change, and the involvement of other European Union (EU) institutions (Baer et al., 2021;Dikau & Volz, 2021). In the remainder of the paper, I focus on explaining these puzzling first years of central banks' advocacy.
Opening the black box of central banking

Understanding central banks' agenda
Hitherto, no study has analyzed how and why the topic of climate change has entered into the agenda of central banks in the first place. There is, however, a rich literature on central banks that can inform us on the sources of and mechanisms behind the uptake of new topics. I review this literature in three clusters: with a focus on the government, on the financial sector, and on central banks themselves.
The first strand of literature follows a principal control logic and stresses that central banks are very much dependent on what the government mandated them to do. Central banks are seen as independent authorities, public bodies that are formally independent from elected politicians and specialized to perform the narrow task that has been delegated to them (Gilardi, 2007;Majone, 1997). However, most scholars in this tradition also stress that this independence is gradual: some central banks coordinate more intensively with their government than others (Fernández-Albertos, 2015). The second strand of literature stresses the influence of the financial sector on central banks. Scholars in this tradition agree that the financial sector is a key influence on central banks. They disagree, though, on the nature and extremity of the relationship, ranging from lobbying (Epstein, 2020) to arguments about (infra)structural power (Braun & Gabor, 2020) or regulatory capture (Adolph, 2013).
Both strands thus point to central banks' external dependencies. Although these literatures make an essential contribution to understanding central banks in general, they cannot explain the sudden and unexpected engagement of central banks in climate policy. As for the government, the timing points in another direction: central banks' mandates only changed after central banks started themselves first or did not change at all (Dikau & Volz, 2021). Second, the financial sector most likely had little interest in promoting central bank climate policies, because these would limit their investment opportunitiesthe whole point is that the financial sector should only invest in the transitionand lower the value of their existing non-climateneutral assets (Colgan et al., 2021;Dafermos et al., 2021).
The third branch of literature, focusing on central banks themselves, is more promising. Central banks, according to this perspective, are powerful agents, able to push for their own priorities because of their expertise and international networks (Marcussen, 2006). These priorities range from securing new tasks, e.g., macro-prudential policy (McPhilemy, 2016) and austerity measures (Ban & Patenaude, 2019). A key insight from this literature is that central banks are motivated by reputational interests and are, for example, keen on using strategic communication in response to public concerns (Moschella & Pinto, 2019). Furthermore, reputational scholars point to conflicting demands. After the financial crisis, for instance, it was on the one hand paramount for central banks to adjust their failed policies, but on the other hand, central bankers 'expressed discomfort with branching out into new policy areas and legitimizing new policy tools' (Johnson et al., 2019, p. 558). Although central banks are presented in most of this literature as forming a tight epistemic community, scholars also point to contrasting policy positions, for example, in their reaction to the global financial crisis (Schulz, 2017) or in implementing macro-prudential policy (Thiemann, 2019).
I take this literature as a starting point and argue that we do indeed need to look first and foremost at central banks themselves. I extend the literature by presenting a framework to analyze the dynamics within central banks, which in the current literature largely remains a black box. This shortcoming is partly explained by the methodological challenge that 'central bankers are very reluctant to disclose information' (Fontan, 2018, p. 164). The next sections provide a theoretical framework to open this black box; the research design section explains how I was able to do so empirically.

Bureaucratic politics
To analyze the role that internal dynamics play in the process of how (new) policy issues enter central banks' agendas, I borrow theoretical concepts from the bureaucratic politics and EU agenda-setting literature. The bureaucratic politics literature focuses on 'the process by which people inside government bargain with one another on complex policy questions' ('t Hart & Wille, 2012, p. 329). The perspective has two important implications for the present case: first, central banks are not homogenous groups of civil servants that work in a conscientiously coordinated way. Central bankers are actors with their own interests and ways of thinking and they battle over what the central bank should do. Second, these differences matter for policyoutput: actors use strategies to implement their interests and positions into policy ('t Hart & Wille, 2012).
Following these implications, I formulate expectations on the main actors within the battle and their motivation (1) and how they battled, i.e., their strategies (2). I hypothesize that there werein the vocabulary of Korpi (2006) -'protagonists' within the central bank in favor of taking up climate change as a central bank issue. The first part of my core argument is that 'policy-seeking' protagonists were the main source: central bankers motivated to have an impact on the economy, with the goal to actually implement climate-neutral policiesnot just talking about them (Adolph, 2013, p. 10). The first protagonists were thus not so much motivated by reputation gains (Thiemann, 2019) nor career interests (Adolph, 2013), precisely because it was risky to engage with climate change, as no-one in their professional field was even considering the topic at that time (see above). 3 As I explain in more detail below, the first central bankers risked being seen as incompetent and illegitimate, putting the reputation of their organization at stake. Expectation 1: Policy-seeking central bankers started to advocate for climate goals within the central bank.

The agenda-setting challenge
My second contribution is to theorize (and test) how the protagonists managed to switch the bank's policy position: the protagonists used deliberate agenda-setting strategies to turn the central bank into an engaged institution. To analyze these strategies, I rely on Princen (2011), who distinguishes between four strategies. First, protagonists need to mobilize other protagonists. To do so, they need to pick the institutional arenas that are most likely to be responsive to the issue, i.e., 'venue shopping', or, if no suitable options are at hand, creating new venues (Princen, 2011, p. 931). I expect that protagonists, above all, made use of the 'boomerang' strategy: find supporters on a global level and let the issue trickle down again (Keck & Sikkink, 1999), because central banks are highly interconnected on an international level and susceptible to rules, norms, and knowledge produced by the international community (Marcussen, 2006). Second, protagonists need to 'arouse interest' and find the right frame to motivate others to participate. They can use 'big words'; to connect the issue to existing norms, values and priorities or 'focusing events'; or take 'small steps': by doing a study or organizing a conference with emphasis on the technical, non-disputed elements of the issue (Princen, 2011, p. 933).
With the last two strategies, protagonists have to persuade others that climate change is not only an important issue that is worth making policy on, but also that a central bank is the legitimate venue in which to do so. Note that this is not self-evident at all if a central bank has no mandate and no organizational capacity for dealing with climate change. At the same time, this step is crucial because reputation is critical for central banks (see above). So, thirdly, protagonists need to build this capacity and develop expertise on the subject, for example, by measuring and publishing the impact of the financial sector on climate change. Fourth, they need to 'claim authority', again by framing, arguing why central banks are a legitimate actor, for example, by linking the topic to central banks' core tasks and competences (Princen, 2011, pp. 934, 936). Altogether, I expect that with these strategies, protagonists secured enough awareness and legitimacy to make engagement with climate goals a self-evident task of the central bank.
Expectation 2: Through mobilizing supporters, arousing interest, building capacity and claiming authority for climate goals, protagonist set climate change on the agenda of the central bank.

Research design
This paper aims to explore the unexpected activism of central banks by focusing on why and how DNB emerged as a climate advocate. DNB is the central bank and supervisory authority of the Netherlands. DNB forms an integral part of the EU governance for monetary and supervisory policy (the Eurosystem and the Single Supervisory Mechanism) and is thus responsible for the implementation of monetary policy and supervision of banks. Besides performing these tasks, DNB also supervises, amongst others, pension funds and insurance companies. 4 The Dutch case is an interesting one in two regards. First, and as I will show in detail below, it is a crucial case and therefore important for the full picture: DNB was one of the first central banks to engage with the topic and was one of the co-founders of the NGFS. Furthermore, understanding how the topic emerged on the agenda of one of the national central banks provides the first puzzle piece in explaining how the topic arose on the agendas of the EU and ECB in particular. Indeed, I am deliberately not studying the ECB. Although the ECB is an important venue when it comes to monetary and supervisory policy-making, the ECB was not a first mover (as I show below) and thus cannot contribute to understanding the very beginning of central banks' advocacy.
Second, among first movers, DNB is a least-likely case for three reasons: to begin with, some central banks, such as the Banque de France and Bank of England, have the explicit mandate to support the general economic policy of their government, which gave them an option to defend themselves with reference to the economic policy of their national principal. 5 DNB does not have this broad mandate. Moreover, DNB is a small central bank compared to the other two and thus has scarcer resources, which means that allocation of resources toward any particular issue is not a given. Last, using monetary and supervisory instruments for climate goals presumably required the rethinking and resetting of policy views. DNB is a rather conservative central bank when it comes to monetary policy (Schulz, 2017), and the Dutch financial supervision tradition does not support a strong steering role of supervisory authorities (Quaglia, 2010).
To explain the puzzling engagement of the Dutch central bank on the topic of climate change, an in-depth process-tracing case study is the appropriate method, as it allows me to unravel the 'situations, actions and events, [and] traces of motivations' that led DNB to take on the role of climate advocate (Blatter & Blume, 2008, p. 319). The timeframe is approximately 2014-2018. I go back as far as necessary to explain the beginning of DNB's first publications on climate change until the year that the NGFS took off and the issue appeared on the global agenda (see Figure 1). My most important sources are semi-structured interviews with central bankers and supervisors who were directly involved with the issue of sustainable finance during this period. I interviewed 11 central bankers and supervisors, of whom two are managers and one of whom is a board member (see Section 1 of the online appendix for the topic list). 6 Having access to this kind of first-hand evidence is crucial and quite unique because, as mentioned above, central banks are hesitant to open up. I also conducted one interview with a civil society organization as an extra validation check. I analyzed the transcribed interviews in NVivo. I complemented the interviews with document analysis (see Section 2 of the online appendix for an overview).

Results
In this section I first provide a chronological overview of the first years of central bank advocacy, focusing on the Dutch central bank and other first movers. Then I zoom in on the positions, policy views, and interests of the main actors in the Dutch case. Lastly, I analyze the agenda-setting strategies of the internal protagonists.

A chronological overview
The first time DNB signaled it was taking climate goals into consideration was in 2011, when the new board adopted a new 'mission'. Thereafter, DNB's mission has been to 'contribute to sustainable prosperity in the Netherlands' (D2, emphasis added). 7 Three years later, DNB considered the issue in connection to one of its core tasks for the first time. In May 2014, the Minister of Finance requested an investigation from DNB into the carbon bubble and DNB concluded in August 2014 that there was no excessive credit risk to the Dutch financial sector from the oil, gas or coal sectors (D4). At the end of the same year, DNB's pension division decided to pursue a theme study into sustainable investment in the pension sector (D3). The report was published in October 2016 (D5).
In September 2015, the supervisory branch of the Bank of England started to speak out on the subject, publishing a report on climate change and the insurance sector. The report introduced the classification of climate risk into physical, liability and transition risks (D13). Around the same time, Bank of England governor Mark Carney held his now-famous speech, 'Breaking the tragedy of the horizon', introducing the classification to a large international audience (D1). At the end of 2015, 196 countries adopted the Paris Climate Agreement and around the same time two relevant international initiatives were launched: the Financial Stability Board (FSB), chaired by Carney, established a task force on climate-related financial disclosures (TCFD) and the G20 Finance and Central Bank Deputies announced the launch of a Green Finance Study Group (D8, D17).
Meanwhile, in 2015, the issue again appeared internally within DNB, in two different places. The economic policy division received a request from the board to do a study on the implications of the energy transition for the Dutch economy (R2, R4). While the preparations for this study were underway, the Green party asked the Minister of Finance to examine financial risks caused by climate change for the state as a whole, referencing Carney's speech. The minister promised in return to ask DNB to update the analysis from 2014 (D9). DNB incorporated the request from the Minister by broadening the existing energy project, leading to the publication of 'Time for transition' in March 2016, the first report of a central bank to measure the impact of the transition to a carbon-neutral economy on the financial sector (D10, D16). The report stated that DNB 'sees energy transition as one of the greatest challenges that the economy faces in the long term' (D16, p. 9), a large contrast to the short and comforting response of DNB to almost the same request in 2014. One and a half years later, in October 2017, the study on the financial sector was repeated and broadened, and published in the report 'Waterproof' (D14).
In December 2016, the European Commission established a High-level expert group on sustainable finance (HLEG) to give advice on the direction of further EU legislation on sustainable finance, with representatives from the sector, academia and civil society. The only central banker in the HLEG was member of the DNB executive board Frank Elderson (D7). 8 Around the same time, Elderson also became the chair of the Sustainable Finance Platform, a Dutch network that aims to stimulate sustainable finance, with representatives of the financial sector, the government, academia and other regulators (D6).
After Trump became President of the United States in 2017, US representatives started to avoid further discussion of climate change in international institutions, for example, in the G20 Green Finance Study Group, the FSB and the IAIS (R1, R7-R9). Trump withdrew from the Paris Agreement in June of that year and shortly afterwards French President Emmanuel Macron announced that France would host a summit in December 2017 to boost sustainable investment. The Élysée urged all governmental organizations to contribute and the Banque de France took the initiative to bring together eight central banks and supervisors (D11, R6-10). Elderson was invited because DNB policy advisors present at a G20 Green Finance Study Group meeting proposed his name (R7). During the summit, the NGFS was launched, with Elderson as chair and Banque de France as secretariat (D11). Since its creation, more than 80 central banks and supervisors have joined the network (D12).

Positions, policy views and interests
Within the central bank: 'We are not Greenpeace' Before DNB published its first major report on climate change, respondents report that there were only a few protagonists in the central bank (R1-R11). Respondents link the motivation of protagonists to contribute to the issue to a functional need to perform tasks well, for example, being able to contribute to national and international policy discussions (R2-R4), but first and foremost to personal beliefs (R4-R11): 'This subject resonates […] with people's conviction that this is important and because of this we had a tribe within DNB that wanted something with it' (R8). 9 Respondents also note that there were quite some young people who were 'not stuck yet' (R10) and were enthusiastic about pushing the topic (R5, R6, R9, R10).
The overwhelming majority of the organization was not interested and it was an 'uphill battle' (R10) to get these people on board: 'The majority at the coffee machine, they thought: "Come on, just get to work! What are you doing?" Because it takes a lot of work, such a study' (R3). Respondents give many reasons that they or their colleagues were not in support of taking up the issue. To start, it is not part of the (legal) mandate and organizational process of most people's jobs (R3, R7-R10). For instance, climate change was not perceived as a financial risk (yet), which made it difficult to fit it into the standard policy-making process of measuring risk, analyzing drivers and developing mitigating measures (R3, R4, R7, R8). More generally, for lowerlevel employees, if subordinates do not care, it is not an attractive option to pay attention to a topic: 'If you are in a division with certain division plans and a manager saying what you need to do, then there is only so much space to mess around apart from that' (R10).
Furthermore, sustainability 'was not part of your frame of reference' (R7). This had an epistemic component: sustainability more generally is not part of standard economics, because there is not a first best: 'It is not our standard neoclassical theory of […] leaving it to the market, […] there is a lot of coordination between different interests, […] but you do not see that in your economic models' (R2). Another respondent explained: 'There are people that look more at the traditional, historical risks […] and feel supported by data, by academic theory and also maybe that it is generally accepted' (R7). The second component is normative: central bankers were not supportive because they considered sustainability to be 'idealistic', 'left' or 'political' in general, which has a negative connotation (R1, R2, R6, R8, R10). 'We do not have time for this, we are not Greenpeace', summarized a lot of reactions (R8).
Lastly, there was the fear of reputation loss. This was strongly connected to concerns about critique that DNB would overstep its mandate (R2, R7-R9, R10), but also to the role of DNB as specialist (R2, R10): I think we were also careful because it was for ourselves not clear if we had a lot to add to this discussion, that we did not want to say things that are not right and could undermine our reputation as knowledge institute. (R2) The financial sector: 'Completely surprised' In the eyes of the respondents, the overwhelming majority of the financial sector was not proactive on the issue (R3, R5-7, R11, R12). At the time, the sector was in favor of lowering capital requirements for sustainable investment, the so-called green premium, asset managers offered the first green products, and large pension funds started reallocating some portfolios (R3, R5-7). However, especially in the beginning, financial institutions were not convinced they were subject to any climate risks, nor had they sufficient insights into their exposures (D16): A lot of these institutions were completely surprised that this was on the agenda […]. And then the first reaction was: 'no, we have a graph here from the IEA [international energy agency] […] it is no problem, there will be too much demand for oil and gas'. (R7) However, all financial umbrella organizations joined the Platform for Sustainable Finance mentioned above and some individual organizations contributed actively to writing reports (R7, R10, R11).

The government: 'Not active themselves'
The government had a clear consenting role according to the respondents: it did not actively push for the topic on the agenda of DNB, nor did it oppose anything (R2, R3, R5, R6, R9-R11): 'Nowadays it is there booming and big too, but they have not been active themselves' (R5). The Ministry of Finance did some things it did not have to do, for example, asking DNB a second time to check if the Dutch financial sector was at risk from a carbon bubble after questions from an opposition party (see above), or informing DNB if something came up: We had someone at the Ministry of Finance, […] he knew from this topic and was interested in what we were doing and he sent me the message that there was going to be a sustainable finance working group in Europe. (R7) This working group would be the HLEG (see above).
To sum up, the positions of the financial sector and the government were largely consenting in the beginning: they did not push for the topic nor were they opposed. This allowed the protagonists to start setting the agenda. I now turn to the strategies they used.

Mobilizing supporters
Protagonists mobilized supporters by venue shopping and modification, internally as well as externally. The external shopping started with road shows and round tables: before and during the start of writing the first report, the responsible authors visited a wide array of potentially important stakeholders to explore their positions and test the waters (R1-R4, R7). This included the usual suspects within the financial sector, but also oil companies, NGOs, think tanks, and other governmental agencies on the subject (D16). The positive reaction of (most of) the stakeholders fueled internal support as it abated reputation concerns: 'The reactions were very positive to start, no one said "cobbler, stick to your last"' (R2). Another example of this feedback effect is the external speeches given by board members: board member Elderson would accept speech requests on the topic even though there was no material, which then had to be written in the remaining time, or senior policy advisors would embed the topic in speeches even if the topic did not necessarily require this (R2, R10, R11).
Next to venue shopping, DNB also co-initiated new venues. Horizontally, DNB took the initiative for the Platform for Sustainable Finance (see above) to mobilize the financial sector in the Netherlands (R3, R6, R8, R10, R11): 'As a Baron von Münchhausen we initiated this Platform, without anyone asking us, without a higher committee to report to, we just did it' (R10). The Platform had all national financial umbrella organizations on board, a deliberate choice to guarantee 'buy in' (R6), but the working groups that did the actual work were designed as small coalitions of the willing in which the front runners could 'shine [and] show how good they are' (R6). Vertically, DNB was co-initiator of the NGFS, a literal coalition of the willing (see above). Two design choices were crucial for the mobilization of other central banks: first, low entrance requirements in terms of experience or expertise on climate goals. And second, high commitment requirements: to join the NGFS, the highest level would need to sign (R8). The international agenda-setting then boomeranged to the national level, as a lot of DNB work 'is driven by the international level' (R9): DNB employees in the strategy department started working on NGFS reports and conferences (R7, R8). The opposite happened in the financial stability department: here, the issue received less attention because international counterpart FSB did not have climate change high on the agenda (yet) (R1, R4).

Arousing interest
Respondents identify three distinct big words that helped to arouse interest internally as well as externally for the subject. First, the Paris agreement of 2015 (R2-4, R6-8). Although the preparations for the first report, Time for transition, started before the agreement, the authors used the Treaty to stress the significance of the topic: 'with the link to Paris, that was very powerful. There is going to be an enormous change that will affect the economy […], that strengthened the message, the story and that helped with the success of the report' (R7). Together with the use of the terms carbon bubble and stranded assets (D16), this story helped with media attention: Klaas [Knot, DNB president]  Lastly, the report Waterproof connected the issue of climate change to a very strong Dutch symbol: dykes and flood risk (D14). Furthermore, many respondents identify the speech by Bank of England governor Carney as an important focusing event that drew attention to the topic (R1-7, R9-11) 'You need people that are unsuspected in the eyes of the internal staff that are still sceptical. Those people are green wolves in sheep's clothes, […] who do not act green, but act financial' (R10).
At the same time, small steps took place. As described above, the process started with three studies within DNB. Respondents describe how they used generic terms and looked for common ground (R1, R3, R4). This is also one of the reasons why the first reports do not talk about specific measures, such as higher capital requirements: 'That is one of the reasons behind the generic conclusions, we were not there yet, on that issue it was internally a struggle to make clear that this is important' (R1).

Capacity building
Building capacity took place in three reinforcing ways: via people, prowess, and policy. Firstly, the number of people working on the subject increased. Whereas the first authors started as a temporary project team with representatives of different departments that had sustainability as one of their responsibilities, slowly but steadily specialists started to work full-time on the issue (R2, R5, R7). Secondly, these people, in cooperation with the above-mentioned stakeholder venues acquired prowess, distinguished expertise and skills. The first data on climate risk exposures was collected at Dutch financial institutions, economic models were adjusted, and climate stress tests developed (D14, D16, D18). That this prowess was new was, for example, important for gaining media attention (see above). The last step was the implementation of the topic in internal policies, such as departmental year plans and supervision priorities and guidelines (R1, R3, R9). In this way it is 'not just a small project and then set aside again […]. In my department one of the spearheads is sustainability and then it comes back every week' (R9).

Claiming authority
Claiming authority over the issue and connecting climate change to the existing mandate and tasks of DNB was at the core of the agenda-setting strategy and took place in two ways: mission adjustment and issue linking. The mission of DNB, as told above, had contained 'sustainable welfare' since 2011. A deliberate act of board member Elderson, the words provided fertile ground (R3, R5, R9, R10): 'He managed to get something on sustainability in the mission, […] using it thereafter as anchor [for further steps]' (R5). Most importantly, however, protagonists were able to connect the issue to two existing roles of DNB: that of independent economic advisor and financial supervisor. In the first major report, Time for transition, DNB states that the energy transition is 'one of the greatest challenges that the economy faces in the long term' (D16, p. 9). This link with the economy enabled the research department to do research and advise the government on policy for a climate-neutral economy (R1, R2, R4): 'Energy, inflation, it makes sense that you talk about that. So I think we were careful at the beginning, but during the process it turned out a lot of people were waiting for our opinion on this' (R2).
Because of these demands, DNB was able to position itself as a neutral facilitator that hears everyone, from civil society to industry, and can distinguish facts from fiction (R1-6). 'We heard very often that there were only little independent voices in the energy and climate debate and everyone represents an interest, but we do not' (R2).
The second frame enabled DNB to actually use its own instruments: it framed climate change as a source of (macro-prudential) financial risk, and therefore a job for the financial supervisor. This frame was not new: it made its entrance on the international stage in June 2014, when the United Nations Environmental Programme (UNEP) published the first results of its Inquiry into the 'Design of a Sustainable Financial System', and is also visible in the early work of the Bank of England (D1, D13, D17). Protagonists came into contact with this work through their personal international network (R1, R2, R8-10). Furthermore, the frame was promoted by a Dutch network of academics advocating for a sustainable financial sector, who wrote working papers and organized meetings between members of the UNEP Inquiry and DNB early in 2015 (D15, R8, R10, R11). DNB developed the narrative further by applying it to the Dutch financial sector and quantifying the exposure of the financial sector to climate-related risks. Quantifying was an important step, because it enabled protagonists to 'speak the language' of central banks (R8).

Conclusion and discussion
Becoming a climate activist The paper provides an important puzzle piece in understanding central banks' (early) engagement with climate change by studying why and how the Dutch central bank changed from an organization without mandate and expertise on climate goals to an organization that is leading the NGFS, a worldwide coalition of central banks to green the financial sector. The results show that the initiative mainly originated with a few individual, policy-seeking central bankers within the Dutch central bank. Above all, these protagonists used (newly created) external venues to build up support and expertise internally and framed climate change as a financial risk to legitimize the involvement of a central bank. On the international level, furthermore, the same mechanisms were at play. The results reveal that the NGFS was deliberately set up as a one-issue coalition of the willing to circumvent opposition from opposing governments and central banks within existing institutions.
Moreover, the paper contributes to future research on central banking by providing a theoretical and methodological framework for studying the internal organizational politics within central banks. It can be used to analyze the (in)action of other central banks on climate issues (while taking into account institutional and time context differences). I expect, for example, that the protagonists of the European first movers used the four strategies to push for climate change on the agenda of the ECB. The framework can also be used to explain the uptake of other (surprising) topics by central banks, such as biodiversity, gender, or economic inequality.

Climate policy, agencies and other unexpected climate activists
More generally, the results of this paper have implications for the study of climate policy, independent agencies and the politics of the climate-neutral economy, which I discuss in turn. To begin with, for climate policy scholars, studying central banks' activism highlights the importance of the financial sector in the transition to a climate-neutral economy. Scholars of climate policy should therefore pay significantly more attention to this funding side of the economy; analyzing governmental policy, but also the role of more hidden actors, which may includein addition to central banksprivate investment activists or public developments banks (e.g., Mertens & Thiemann, 2019).
There are three implications for the literature on European and national independent agencies. Firstly, the paper confirms and deepens our understanding of the trend that, in order to legitimize their actions, 'agencies are increasingly turning towards building support and allies from their environment' (Busuioc & Jevnaker, 2022, p. 171), by showing that agencies not only create general stakeholder bodies to legitimize policies externally, but also explicitly one-issue bodies to legitimize policies internally. Secondly and related, this paper directs attention to individuals as an explanatory force by pointing to the role played by policy-seeking central bankers, a factor that has so far been given scarce consideration. The results of this paper, thus, call for more research on individuals within agencies, their preferences and their impacts on policy. Next to intra-agency politics, the necessity of establishing a coalition of the willing draws attention to inter-agency politics: national agencies are not always in agreement with their peers from other EU member states. This motivates new avenues: do we see the same trend in other (regulatory) domains in the EU? How does interagency politics influence policy-output? And what does this mean for the power of EU agencies vis-à-vis their national and European government(s)?
The central banks discussed in this paper are just a few among many public actors that are (increasingly) involved in the politics of the transition to a climate-neutral economy. Ministries of Finance from all over the world have made pledges to implement climate goals into their fiscal policies, a court in the Netherlands has issued a revolutionary verdict stating that Shell has a responsibility to prevent climate change and the UK's National Health Service has adopted a plan to become carbon neutral by 2040 (Coalition of Finance Ministers, 2020; NHS, 2020; The Hague District Court, 2021). Based on the dynamics studied in this paper, I expect the advocacy of these public actors to have considerable implications: besides implementing climate change concerns into their own policies, they can push for the issue among their peers and parent governments, enlarging the coalition of climate protagonists. By showing how a fundamental U-turn on climate policy came about in a central bank, this paper contributes to our understanding of how these and other public actors that at first glance have nothing to do with climate change in particular, suddenly turn into climate advocates.
Finally, the activism of central banks and other public actors brings to the forefront new questions of accountability: can these steps outside their mandates and expertise be legitimized by the argument that they are making policy for the common good? And what are the long-term effects of their activism? Will they use their experience with climate to start advocating for other policy issues? These questions will become even more relevant as governments continue failing to meet global climate targets. Indeed, we may see even more unexpected climate advocates turn up, reshaping the politics of the climate-neutral economy. Notes 1. R10. 2. See for a more extensive overview of what central banks are doing or could do, also globally, Baer et al. (2021) and Dikau and Volz (2021). 3. As the issue diffused and started to appear on the agenda of more central banks, as well as the larger political and public agenda, potential reputation gains and career interests increased. Indeed, the incentives probably even swapped: not paying attention to climate change became a reputational risk. 4. See for a full overview of DNB's tasks https://www.dnb.nl/en/about-us/missionand-tasks/. 5. Banque de France: Art L. 141-1 Code Monetaire et Financier; Bank of England: Art. 11(b) Bank of England Act 1998. 6. For the first contact with DNB I was able to draw on my personal network, as I worked at DNB as trainee policy advisor in 2018-2020 before taking up employment at Goethe University Frankfurt. During my employment at DNB I was not directly involved in sustainability-related projects. 7. See for document (D) references Section 2 of the online appendix. 8. Elderson was officially a member in his role as member of the EU Single Resolution Board. 9. All quotes are translated from Dutch into English by the author.