Independence without purpose? Macroprudential regulation at the Bundesbank

Abstract It is commonly assumed that state agencies legitimize themselves via outputs. This paper shows that in situations of organizational crisis, state agencies may adopt new policy areas symbolically to compensate for lost legitimacy. Drawing on an ethnography within the Bundesbank, internal documents, and insider interviews, I trace how the German Bundesbank adopted financial stability as a policy area to compensate for the loss of monetary policy and banking supervision in the early 2000s. By focusing on the relationship between internal organizational struggles over the Bundesbank’s identity and the boundary work it has to conduct to establish its new role, I show that the Bundesbank failed to shift the state-economy boundary post-crisis in its effort to regain its position as autonomous purveyor of macroeconomic governance.


Introduction
Prior to the financial crisis, central banks managed to produce legitimacy towards their audiences by orienting themselves towards a relatively simple measure of performance, namely low and stable inflation (Singleton, 2010). Their public image as guardians of monetary stability, and the mono-functional integration of policy decisions, market interventions, and their rationalization necessitated the exclusion of other policy issues from central banks' purview, such as financial stability, inequality or rampant financialization (Walter, 2019;Walter & Wansleben, 2020). After the global financial crisis, this macroeconomic governance regime, focusing solely on the inflation rate came into question. Central banks engaged in ever more extensive unconventional monetary policy in an attempt to stabilize financial markets and to retain their legitimacy as the purveyor of macroeconomic governance. Meanwhile, post-crisis reforms institutionalized new roles for monetary authorities as financial regulators, as observed in the United Kingdom. As a result of this increasing range of policy areas in which central banks are involved, and the sheer size of their asset purchases, these authorities' image as technocratic guardians of monetary order is facing increasing challenges. Consequent calls for democratization of central banking, monetary financing and more direct intervention in the refinancing conditions of climate-neutral industries via 'green central banking' highlight the shifting position of central banks in their polity and raise profound questions about the organizational forms and institutional arrangements on which contemporary central banking is predicated.
The German Bundesbank is one such case of a central bank that has experienced a radical challenge to its cherished institutional role. Known for its stringent anti-inflationary stance, the Bundesbank was regarded as the central bank of Europe during the 1990s (Marsh, 1992) and served as an international role model for credible, independent monetary policy (Johnson, 2016). By the early 2000s, however, the Bundesbank found itself in a precarious position. First, the establishment of the European Central Bank (ECB) took from the Bundesbank its main task: control over interest rates and inflation. Second, the Bundesbank lost the political struggle to be the single banking supervisor in Germany to the newly created BaFin in 2002. These two developments left the Bundesbank in the awkward position of being a historically powerful, independent central bank without monetary policy authority, and a banking supervisor without administrative power. In its new institutional environment the Bundesbank was left without the ability to legitimize itself via policy outcomes, throwing German central bankers into an organizational crisis (Dyson, 2009).
In this paper, I investigate the Bundesbank case to gain insights into how organizational dynamics internal to central banks, their search for legitimacy, and changes in the broader political and institutional contexts interact in such situations of organizational crisis. I argue that the disempowered Bundesbank engaged symbolically in the production of expertise on the topic of financial stability and macroprudential regulation to produce legitimacy in relation to its environment precisely to compensate for its limited ability to legitimize itself via actual policymaking. In contrast to microprudential regulation, which assumes that the stability of each market actor is sufficient for financial stability, macroprudential regulation assumes that financial markets are unstable on a systemic level; in other words, systemic risks emerge endogenously within financial markets over financial cycles. Prior to the global financial crisis, however, my argument is that symbolic engagement with macroprudential regulation led to 'decoupling' (Meyer et al., 1997;Meyer & Rowan, 1977), a state in which the Bundesbank adopted formal structures and discursive expertise to signal its belonging to an organizational field of macroprudential regulators without changing its core decision-making procedures. However, this symbolic engagement became unexpectedly relevant because of the financial crisis, which raised a demand for actors who could implement macroprudential policies that had been lacking before 2008. A recoupling of the Bundesbank's mandate and its practices was thus provoked by external events. However, the Bundesbank remained in an ambiguous position precisely because macroprudential policy works differently from monetary policy both institutionally and as an instrument. Macroprudential regulation involves visible distributional consequences that do not sit well with the idea of an apolitical, independent policymaker. As a result, the Bundesbank came to share responsibility over macroprudential regulation with the BaFin and the Ministry of Finance. The institutional reforms after 2008 thus have not brought back the widely feared Bundesbank of the 1990s.
The situation after 2008 and the discussion about how to institute macroprudential regulation as a policy programme required finding ways of creating a new boundary between the state as an administrator of macroeconomic intervention and financial markets as an object governable by such intervention. To gain insights into such boundary work, I draw on the sociology of translation (Callon, 1984). More specifically, I argue that creating a new stateeconomy boundary based on instruments adopted from such transnational arenas as the Basel III bank capital requirement standards requires central banks to disentangle initially unclear causal relationships between the means (policy instruments) and the ends (an increase in financial stability). For central banks wishing to legitimize themselves via policy outputs (Scharpf, 1970), unclear means-ends relationships are problematic. In the process of central banks' efforts to create the discursive space they need to justify their macroprudential interventions and internal organizational technologies to facilitate these interventions, lacking material entanglements within financial markets and banks might render these policy interventions inoperable. Banks for example could lack the accounting infrastructure to provide the data necessary for calibrating central banks' policy interventions or market actors themselves are not using the policy rate set by the intervention for their business practices, thereby voiding the effect of the intervention. One possible solution to this is straightforward: if organizations can reconfigure and disentangle their environment to fit their policy instrumentthat is, translate their newly developed problematization of the economyuncertainty could be reduced. If successful, the central bank can then use outcomes produced by an entire network of actors as the effect of its policy intervention, thereby generating output legitimacy. As the introduction of new housing instruments in 2015 and 2020 shows, the Bundesbank has been at least partially successful in this translation of its new policy programme and has therefore been able to make significant strides in producing output legitimacy given its limits within its new institutional environment.
In sum, this paper's analysis adds to this special issue by providing insights into how the new state-economy boundary for macroprudential regulation is formed, institutionalized and made operational. It points towards the crucial role of the central bank as an actor whose concerns for legitimacy in relation to various fields are key drivers in the adoption of macroprudential regulations at the national level (Baker, 2013(Baker, , 2015Goodhart, 2015;McPhilemy, 2016;Mero & Piroska, 2017;Piroska et al., 2021;Stellinga, 2020;Yagcı, 2017). As such this study uses the Bundesbank adoption of macroprudential regulations to provide insights into how new policy programmes for central banks can be used to move from symbolic legitimacy towards legitimization via outputsan area of study so far neglected by organizational sociology (Bromley & Powell, 2012;de Bree & Stoopendaal, 2020)while tracking how divergent early problematizations within the central bank can become consequential in the course of institutional reform. Lastly, it points to the importance of discursive and boundary work by central banks in the production of the state effect presented in the introduction of this special issue.
Studying the Bundesbank's organizational response to the initial shock to its organizational legitimacy and subsequent boundary work post-financial crisis requires an understanding of internal processes and practices. To study these processes and practices, I conducted ethnographic research within the Bundesbank for six months via an internship in 2018. I worked in the coordination department of the financial stability directorate, whose primary task is to prepare upcoming meetings in the financial policy committee, coordinate requests from inside the Bundesbank from other directorates, in addition to facilitating requests from parliament. Furthermore, the department is tasked with preparing Bundesbank representatives for international and national committee meetings and transmit new developments in the political sphere to the more analytical departments, and vice versa. It is therefore the ideal position to observe the organizational processes and boundary work the Bundesbank undertakes post-crisis. Based on my access to the Bundesbank, I use various data sources to track its internal and discursive changes from 2000 until 2020. The first data source relies on internal department plans, which are available to employees for contact information on specific issues. These plans include all employees of the directorates of banking supervision, financial stability and its predecessor the international directorate. Beyond the employees themselves, the plans also include the task descriptions of departments and working groups. I use this data in three ways: first, I track where specific expertise on financial stability has accumulated over time. Second, I establish the degree of decoupling between the expertise and the invested resources measured in terms of the number of employees working in specific working groups over time. Third, I identified and interviewed nine key insiders. Lastly, I analyzed financial stability reports (FSR), parliamentary debates and legal texts to capture the translation process the Bundesbank underwent. A special role among these documents is taken up by FSRs, because they are the prime publication used by the Bundesbank to advocate for its problematization of financial markets. In later periods, FSRs became key components of the Bundesbank's new financial stability mandate, due to their foundational role in decision-making within the post-crisis regulatory framework.
The next section introduces the theoretical framework in more detail, while pointing towards how organizations' legitimacy concerns might interrelate with their boundary work in the creation of a new policy area. At the same time the section will point to the importance of translation processes inside and outside the Bundesbank. In the first analytical step I will show how the early engagement with financial stability emerged from an organizational need for legitimacy, with an emphasis on the internal organizational struggles over resources between the directorate for banking supervision and the international directorate. The second section focuses on the post-crisis period in which a new macroprudential regulatory framework is adopted in the wider environment of the Bundesbank. Crucially, it is here that incongruent problematizations in the early adoption of expertise on financial stability at the organizational level prevented the Bundesbank from taking full control over financial stability. During this period the Bundesbank overcame decoupling by creating a new directorate for financial stability and creating internal organizational technologies to facilitate its new role in the macroprudential regulatory regime. However, the Bundesbank still has to enrol other actors to create the material conditions for its policy interventions. The final analytical section uses a recommendation on new housing instruments to analyze the boundary work the Bundesbank underwent to translate its new policy programme. The last section concludes.

Central bank legitimacy, boundary work and translation
The neo-institutional literature on organizational legitimacy takes organizational survival within a heterogenous environment as its starting point. Organizations rely on their environment for resources, be they material or social. Thus, organizations find themselves subject to an increasingly institutionalized environment making heterogenous demands on them. To cope with these inconsistent demands, organizations might choose to adopt rationalized myths from their institutional environment ceremonially to signal their belonging within an organizational field, in addition to protecting internal processes necessary to fulfil their core tasks. This type of buffering between symbolically adopted formal structures and core organizational tasks is referred to as 'decoupling' (Bromley & Powell, 2012;Meyer & Rowan, 1977). In the neo-institutional literature, organizations therefore produce legitimacy in two ways: by symbolically adopting formal structures and discursive expertise to conform to their institutionalized environment or by producing outcomes via management of relational networks (Meyer & Rowan, 1977, p. 354). The degree to which an organization uses a particular method of legitimization is not static, however, but rather depends on the institutional environment. Organizations might deploy both ways at the same time or predominately use one rather than the other.
While the ceremonial adoption of expertise and norms might be a temporary, albeit risky substitute for output legitimacy, this does not mean that adoption itself has no effect. Rather, once adopted, these myths can be the starting point for the development of new policy programmes, which may allow the organization to legitimize itself via outputs. Establishing a new policy programme is not as straightforward as it might seem, however, as the organization might not be able to overcome its initial decoupling (Bartley & Egels-Zandén, 2016;Lim, 2017;Tilcsik, 2010). A key role in reducing decoupling is played by early adopters within organizations, especially if they manage to gain control over key positions as department heads (Carpenter, 2020;Tilcsik, 2010). Department heads are high enough in the organizational hierarchy to implement their ideal of what the organization ought to be by taking control over the future production of expertise and by solidifying their ideas through hiring decisions and the establishment of new practices. However, even if a department head manages to overcome the initial decoupling within the organization it does not mean that it will be able to produce legitimacy via outputs, because the material entanglements with its environment necessary to make the new policy programme operational might not yet exist.
While the institutional literature has found conditions and processes in which symbolically adopted instruments can become tightly coupled with organizational practices (Bartley & Egels-Zandén, 2016;Glaese, 2020;Hallett, 2010;Lim, 2017;Michelson, 2019;Sandholtz, 2012;Tilcsik, 2010), it has so far failed to come up with a convincing answer to the question of how organizations can overcome mismatches between the conditions outside the organization and their new toolbox (Bromley & Powell, 2012;de Bree & Stoopendaal, 2020). The literature therefore lacks the conceptual framework needed to fully describe how organizations move from producing symbolic legitimacy via the decoupled adoption of new policy programmes to producing output legitimacy via the implementation of said policy programmes. A similar gap can be found in the literature on the adoption of macroprudential regulation as it has focused mainly on policy learning (Yagcı, 2017), idiosyncrasies of macroprudential regulation itself (Goodhart, 2015;Stellinga, 2020), the lack of scientific backing (Thiemann et al., 2018;Thiemann et al., 2020) or national politics (Mero & Piroska, 2017;Piroska et al., 2021). They have often neglected the work central banks must undertake to reconfigure the environment to make their policy instruments operational. It is exactly these material conditions outside organizations, however, which allow for the outcomes suggested by the economic models accompanying new regulatory interventions to materialize, produce output legitimacy and thereby stabilize the organization's position within its polity.
To fill the gap in our knowledge of how organizations can deal with the lack of material relationships with their environment to make policy outcomes materialize, I make use of the conceptualization of the state-economy boundary introduced by Coombs and Thiemann (2022) in the introduction to this special issue. Thiemann (following Mitchell, 1991, 1999), argue that central banks produce the boundary between state and economy. They have a pivotal role in defining and, if successful, creating the governable object of the economy, upon which they can administer macroeconomic policy. This process has two important aspects: First, a rationalized space has to be created, which clearly delineates state from market actions. This delineation must clearly identify means-ends relationships between policy interventions and the macroeconomic outcomes utilizing measures legitimized via quantification, narrative scripts, or appropriateness within the organizational field (Coombs, 2022;Meyer et al., 1997;Sauder & Espeland, 2009;Thiemann, 2022). The rationalized space created by central bank expertise for policy intervention by itself is not sufficient to produce macroeconomic outcomes; concrete entanglements with private actors, which constitute the infrastructure for governing techniques (Braun, 2015;Rose & Miller, 1992;Walter & Wansleben, 2020) are equally important in constituting the state-economy boundary (Eyal, 2013). It is these entanglements which construct the state 'effect'in other words, the attribution of a macroeconomic outcome to policies of the state rather than to the collective of private and public actorsto emerge (Eyal, 2013;Mitchell, 1999).
In this paper I argue that the Bundesbank has to engage in boundary work to establish both the discursive expertise and the material entanglements constituting the state-economy boundary of the post-crisis macroprudential regime (Best, 2022;Gieryn, 1983). The success of this boundary work is a necessary condition for the Bundesbank to produce output legitimacy once again. To better understand how the Bundesbank might disentangle the macroprudential means-ends relationships, I rely on the sociology of translation (Callon, 1984;Callon & Latour, 1981). 1 Viewed from this perspective, becoming an autonomous policy agency in a new policy area means becoming the obligatory passage point in which it can claim the outcome produced by the network of actors within the state-economy boundary for itself. To do so the organization has to reconfigure, enrol and stabilize its surrounding actor-network and its position within the network. That is to say that the organization has to translate its understanding of the economy in an effort to reconfigure the actors within the state-economy boundary to create the conditions under which the means-ends relationships established by their discursive expertise can survive the intrusion of other problematizations (Beunza & Ferraro, 2019;Callon & Latour, 1981).
Following from these considerations I show that there are four conditions for a central bank to regain its ability to legitimize itself via outcomes. First, it must recouple its internal practices with its policy instruments, because without this recoupling policy instruments would remain unused. Second, it must be able to create the discursive space for its intervention. This includes creating the relationships between the means (policy instruments) and a clearly defined and measurable policy outcome (financial stability). Third, the central bank must translate its expertise, that is, to reconfigure the entanglements in the state-economy boundary to produce outcomes. Fourth, within the institutional arrangements between itself and other state agencies, the central bank must become the obligatory passage point to claim the policy outcome for itself. As the case of the Bundesbank will show, until all four conditions are met an organization will have only limited abilities to use policy outcomes in its search for legitimacy. Table 1 summarizes these four conditions for each time period of analysis for the Bundesbank.

The Bundesbank before the euro
Understanding how the Bundesbank managed to assert itself as the expert on financial stability requires a short historical overview of the Bundesbank's fate after the introduction of the euro and the creation of BaFin, the single financial supervisor, in 2002. The following provides a short historical overview of the Bundesbank's initial starting point as Europe's central bank in the 1990s and the subsequent structural changes which started its transition into the Bundesbank of today.
The Bundesbank of the 1990s is a prime example of an organization built on output legitimacy. Following struggles over whether fiscal or monetary policy should take primacy in macroeconomic management after the breakdown of the Bretton Woods system (Rademacher, 2021), the Bundesbank emerged as the main purveyor of macroeconomic governance via monetary policy in Germany. Backed up by its historically strong independence, the Bundesbank in the 1990s found itself in full control of the interest rate. At the same time the goal of low inflation rates became the norm as the goal of monetary policy, due to the worldwide spread of inflation targeting (Wasserfallen, 2019), its scientific backing in mainstream economics (Goodfriend, 2007;Walter, 2019) and its disentangled means-ends relationship between the state and the economy (Walter, 2019;Walter & Wansleben, 2020).
This organizational environment of the Bundesbank was also reflected in a tight coupling between internal structures and practices with the outputs produced. Because of the multiple veto-players in the decision-making process, the Bundesbank was consensus-based, focusing on the core task of controlling the inflation rate. The ideological focus on monetary policy also demoted banking supervision internally to a secondary goal (Dyson, 2009;Marsh, 1992). As a result, the Bundesbank 'was content to formally delegate key activities to an independent agency attached to the Federal Finance Ministry, the Berlinbased Federal Banking supervision office' (Dyson, 2009, p. 141). Crucially, the Bundesbank was also extraordinarily autonomous as it was able to pursue its policy programme even against the preferences of its principal. It therefore fulfilled all four conditions for producing output legitimacy. However, two structural changes within the Bundesbank's environment coincided in the early 2000s, which started the transition from an organization legitimizing itself via output to an organization that symbolically adopts rationalized myths in the form of formal structures and expertise. The first structural change was the establishment of the ECB in 1998 in advance of the introduction of the euro in 2002 and the Bundesbank's resulting loss of control over monetary policy. This change in the conduct of monetary policy downgraded the Bundesbank from the premier central bank in Europe (Marsh, 1992) to only one contributing national central bank, albeit an important one. The introduction of the euro led to questions about the Bundesbank's redundancy in Germany as tensions emerged between banking supervision, monetary policy and central bank, independence. Faced with the question of whether the Bundesbank had become redundant Bundesbank presidents Welteke and Weber attempted to promote a variety of topicssuch as banking supervision and financial stabilityto justify the Bundesbank's size. 2 This self-promotion by the Bundesbank as the new banking supervisor coincided with federal government worries about the competitiveness of the German financial sector and reforms applied to a more financialized pension system, which initiated financial regulation reform debates.
These factors resulted in the second external structural change. In May 2002 BaFin was founded, consolidating the three smaller regulatory agencies. According to Handke and Zimmermann (2012), the federal government decided against establishing the independent Bundesbank as the single supervisor, as recent reforms financializing the German pension system made the abdication of financial regulation politically costly. The Bundesbank was therefore not successful in becoming the single supervisor, but its heavy lobbying did enable it to retain shared responsibilities over banking supervision with BaFin. Importantly though, the Bundesbank is not formally allowed to conduct administrative acts, even today. Hence, it would have to rely on BaFin for all policy interventions.
In the next section I examine the Bundesbank's organizational reaction to the changes in its environment. The first change was to the structure of the Bundesbank itself, which enabled a higher degree of freedom for directorates to adjust to their specific environments. The second internal change is directly related to internal restructuring as it allowed for new problematizations of the economy and financial systems to take hold in two different directorates within the Bundesbank. It is precisely these early problematizations which would become consequential in the Bundesbank's struggles to establish itself as the new financial regulator within the German polity post-crisis.

Internal restructuring towards symbolic legitimacy
The two structural changes in the Bundesbank's environment led to an organizational crisis, which required the bank to seek meaning in a new world where output legitimacy via monetary policy or banking supervision became impossible. In its search for meaning the Bundesbank shifted towards producing legitimacy symbolically. The Bundesbank changed its internal structure to accommodate this new type of legitimacy. An internal structural reform passed in 2002 had two effects. First, it centralized policy decisions in a downsized board of the Bundesbank, rather than a number of state-level boards. Second, it provided individual board members with more control over their respective directorates, thereby loosening the tightly coupled organizational structure geared towards monetary policy. This reform and the subsequent higher degree of freedom coincided with a massive reduction in staff during this time. This scenario led to board members laying claim to as many topics as possible while defending the expertise held in their own directorates in an effort to obtain as many resources as possible. The struggle over resources in a declining Bundesbank therefore took the shape of jurisdictional struggles over policy areas between directorates.
The Bundesbank's directorates, now equipped with more agency, reacted to shifts in its environment by accumulating expertise to lay claim to jurisdiction over upcoming topics within the larger organizational field. One such topic was financial stability, which originally emerged in the IMF and BIS as well as in the central banking community in the early 1990s (Baker, 2013;Čihák et al., 2012;Thiemann et al., 2018Thiemann et al., , 2020. The IMF was confronted with financial stability concerns because of its experience with the Asian financial crisis, while the BIS began developing macroprudential ideas at the end of the 1990s. Both institutions focus heavily on financial markets, which led them to be forerunners in thinking on financial stability (Interview with Bundesbank official 7). Two directorates began to accumulate expertise on the topic of financial stability within the Bundesbank as a result of this emerging norm of central banks as purveyors of financial stability: the banking supervision directorate and the international affairs directorate. Both directorates were restructured and received responsibilities for financial stability during the structural reforms in 2002. Both directorates reacted to different environments, while accumulating their new expertise, leading to different problematizations of financial regulation, the financial system and the Bundesbank's role in it.
The banking supervision directorate changed more drastically during the structural reform (Interview with Bundesbank official 9). A new department for 'microand macroprudential analysis' was created in 2002 to perform analyses on systemically important banks and later on performed stress tests and designed early-warning indicators. The department focused on determining the systemic relevance of banks, systemic risk within the banking sector and the stability of the banking sector on the 'macro' side, while the 'micro' side focused mainly on estimating the effects of new capital regulation based on each bank's respective balance-sheet. The directorate itself was further charged with implementing the Basel II reform package. The key components of problematizing the financial system in the banking supervision department were based on the perception of banks as atomized actors whose balance-sheets are independent of each other. Here systemic risks stem from exogenous shocks to the financial system or the bankruptcy of a big bank, while the regulator's task consists of guaranteeing the resilience of each individual bank. The role of the Bundesbank in this problematization comes down to that of a rather apolitical banking supervisor, whose main responsibility consists of monitoring the compliance of banks with regulatory standards rather than proactive interventions in financial markets. This problematization led to specific analytical tools for the supervision of systemic risks in the form of stress tests, which allowed for an estimation of effects of macroeconomic and idiosyncratic shocks to the entire banking sector or single banks. Changes within the banking supervision directorate can be traced to two causes: Welteke's strategy of establishing the Bundesbank as the key provider of expertise on financial stability and the creation of BaFin as an institutional competitor to the Bundesbank's authority. The heavy focus on systemic risk can also be interpreted as a strategy to carve out a distinct field of banking supervision, in contrast to the focus of BaFin.
The international directorate, by contrast, focused on the topic of financial stability with only a few changes to the structure of the directorate. The international directorate was tasked with representing the Bundesbank in international working groups and committees such as BIS, IMF and ECB working groups, and with analyzing the impact of international financial and currency systems on Germany. Financial stability concerns entered mainly through the prominent figure of Vice President Jürgen Stark (Interview with Bundesbank officials 6 and 7), who represented the Bundesbank on the international stage and became the head of the international directorate during the structural reform in 2002. Stark, as Vice President, emerged as the representative of a strong Bundesbank and an expert in international financial systems, especially considering the weak standing of then Bundesbank President Welteke. It was therefore the emergence of financial stability topics in the IMF and BIS that prompted the adoption of financial stability topics in the international directorate (Interview with Bundesbank officials 6 and 7). 3 As the IMF and BIS both focused heavily on endogenous risks emerging from financial markets, rather than risks stemming from single market participants, the expertise accumulated in the international directorate differed starkly from that of the banking supervision directorate. From a market perspective the inclusion of systemic risk allowed for the perception of risks as endogenous to financial markets, conceptualized as a chain of interdependent balance-sheets. Exemplary for the endogenous risk perspective are over-the-counter derivatives markets, in which even slight price adjustments can cause feedback effects within derivatives markets due to counterparty risks, subsequent withdrawals of market liquidity and margin calls (Deutsche Bundesbank, 2005, pp. 28-31). This problematization of systemic risks as endogenous to financial markets requires a different type of intervention by financial regulators. The regulator's role in this problematization is anti-cyclical; in other words, regulators have to intervene in financial markets before feedback effects in financial markets escalate into a full blown financial crisis. Post-crisis this type of intervention would often involve policy tools such as credit constraints in housing markets, counter-cyclical capital buffers or extra capital requirements for systemically important banks. As such, the prospect of endogenous risk emerging from financial markets stemmed largely from the transnational epistemic community of central banks and was not built up for policymaking as the instruments for its prevention would be introduced only after the crisis. Consequently, the expertise built up in this perspective served mainly as a way to legitimize the Bundesbank in the international community of central banks, rather than actual policymaking intentions.
Even though the accumulated expertise described above would suggest a form of policy learning for future policy action, one has to keep in mind that the Bundesbank did not have any administrative power over policy instruments. Only the banking supervision directorate conducts onsite visits to banks and could, at least in principle, demand a more detailed look into the accounts of banks. Table 2 provides evidence for decoupling on the topic of financial stability by the Bundesbank in the pre-crisis era. As the table shows, the number of non-administrative staff working on systemic risks in the banking supervision directorate rose more quickly than in the international directorate because it was tasked with implementation of Basel II. In the period between 2000 and 2007 the total number of employees in the banking directorate working on systemic risk rose from one to 37, while in the same period in the international directorate the number rose from seven to 20. The lack of expanded resources on macroprudential regulation are in line with the argument that financial stability expertise in the international directorate was mainly for show.
Further evidence of decoupling can be found in the adoption and institutionalization of FSR in 2003. The establishment of FSRs was influenced by the promotion of financial stability as a norm for central banks by the Bank of England and the ECB (Interview with Bundesbank official 7). The strong focus of the Bank of England on financial markets precipitated their early development of financial stability concepts. The Bank of England's publication of the first FSRs was seen as especially important and helpful in convincing board members to publish FSRs (Interview with Bundesbank official 7; Čihák et al., 2012, p. 6). Publication of the report by the Bundesbank thus falls into line with the adoption of formal structures for the purpose of producing symbolic legitimacy (Meyer et al., 1997).

The crisis, regulatory reform and competing problematizations
The previous section has outlined how the Bundesbank reorganized itself to accommodate the production of symbolic legitimacy rather than output legitimacy. Its organizational changes, invested resources, adoption of the FSR and expertise itself indicate decoupling between the purported expertise on financial stability and actual practice. The initial ceremonial adoption would soon come to bear fruit as the rise of macroprudential regulation with the emergence of a new regulatory paradigm after the financial crisis (Baker, 2013;Coombs, 2020) and the subsequent passing of the financial regulatory reform the Finanzstabilitätsgesetz (FinStabG) in 2012. The following section outlines how the Bundesbank reacted to the great financial crisis on an organizational level. It also shows how the initial competing problematizations of financial markets and the role of the Bundesbank in them prevented the Bundesbank from taking full control of the post financial regulatory framework.
The Bundesbank reflected on its organizational shortcomings after the immediate crisis was resolved. This resulted in the eventual transformation of the international directorate into the financial stability directorate in 2009. The crisis therefore served as an external shock to the organizational structure of the Bundesbank, in which decoupling itself became problematized. As already described, the banking supervision department focused on the stability and systemic contribution of individual institutions, while the international department took on the perspective of risks emanating from financial markets. Crucially, these two perspectives were not brought together, leading to a perceived institutional blind spot, in which the banking supervisors failed to judge the risk stemming from international derivatives markets, while the international department was not sufficiently able to estimate the internal weakness of banks (Interview with Bundesbank officials 6, 7 and 1). As a result, the Bundesbank worked towards bringing their practices closer in line with its perceived future tasks of taking full control of macroprudential instruments.
The recoupling process, however, progressed only slowly, as jurisdictional struggles between directorates persisted. The banking supervision and international directorates remained largely the same, prior to the transformation of the international directorate into the new financial stability directorate in 2009. The new directorate was now solely responsible for financial stability within the Bundesbank, but was still largely structured like its predecessor, with the addition of a department for macroprudential analysis and a secretariat focused on handling upcoming topics in international committees. The macroprudential analysis department was staffed mainly with members of the banking supervisory department for macroprudential analysis, even though the old macroprudential analysis department within the banking supervisory directorate still existed. The structure of the financial stability directorate largely stayed the same until 2012 as the banking supervisory directorate continued to be in charge of systemically relevant banks, stress tests and early warning indicators.
These changes, although incomplete, already point to a recoupling of expertise and the newly established tasks the financial stability directorate took on during the crisis (data collection, full responsibility for the increasingly important FSR and coordination of the German rescue fund). Only after the financial stability reform passed in 2012, however, did the new directorate integrate the remaining tasks of the banking supervision directorate. Furthermore, it created new formal and informal organizational structures to facilitate full use of policy instruments and thereby recouple their internal practices with their newly received policy instruments.
The lack of organizational recoupling previous to reform, however, led to persistent jurisdictional struggles between directorates and thus to competing problematizations of financial markets, the nature of systemic risk and the role of the Bundesbank in the future regulatory framework. The financial stability directorate had a clear idea of the new identity of the Bundesbank as intervening in financial markets in an effort to curtail systemic risks, while other directorates of the Bundesbank differed in their problematizations. These directorates raised concerns over the political consequences for the Bundesbank's cherished independence because of the more salient distributional effects of macroprudential regulation (Engelen, 2011), especially in the realm of credit allocations and banking supervision. Furthermore, they argued that direct control over macroprudential instruments would conflict with the Bundesbank's role in the European System of Central Banks and largely retained pre-crisis problematizations of financial distress stemming from exogenous shocks, rather than endogenous pathologies of the financial system.
Post-crisis, a multitude of proposals for regulatory reform were considered by a variety of industry representatives and political parties, ranging from a UK-type system (Engelen, 2011), in which the Bundesbank would have full control over macroprudential instruments, to more mixed forms of shared responsibility between the Bundesbank and BaFin. Even though the financial stability directorate pushed for the Bundesbank to become the single supervisor in Germany in its FSRs during this time, its problematizations failed in trials of strength with other problematizations within the Bundesbank and thus it was unable to enrol and mobilize actors within and outside the Bundesbank. As a consequence, when the time came to decide on a new regulatory setup the inconsistent actor-network of the Bundesbank led to an impasse, in which no clear decision on the new institutional framework for financial regulation in Germany was reached (Handke & Zimmermann, 2012).
The prolonged political struggle ended only when the ESRB published its recommendation to EU member states on the structure of national regulatory frameworks (ESRB 2011). 4 This recommendation was, with minor tweaks, fully adopted by the German parliament. In parliamentary debates, the Bundesbanks' expertise was never called into question; rather it was emphasized that the Bundesbank already possesses the expertise needed to deal with the analysis of financial stability (Deutscher Bundestag, 2012, p. 22659). The eventual post-financial regulatory reform, the Finanzstabilitätsgesetz (Fin-StabG) was passed in 2012. It established the Committee for Financial Stability (Auschuss für Finanzstabilität -AFS), whose main tasks include coordinating discussions between BaFin, the Bundesbank, the finance ministry and Germany's Bank Rescue Fund in an effort to make recommendations on macroprudential interventions or the creation of new policy instruments.
The resulting reform package put the Bundesbank in an ambiguous position. On one hand, it failed to become the obligatory passage point in the stateeconomy boundary for financial stability, due to internally conflicting problematizations. Thus, its ability to claim financial stability as an outcome solely of its policy actions is limited. On the other hand, it was directly tasked with providing problematizations of financial markets in the future in the form of expert opinions communicated via FSRs. It could therefore still have the ability to produce output legitimacy, albeit in a limited fashion, if it manages to recouple its internal practices with its new environment, provide a discursive space for future interventions, and enrol its environment in its policy programme. The next section turns to the Bundesbank's attempts to fulfil the remaining conditions for output legitimacy.

The creation and occupation of the new state-economy boundary by the macroprudential Bundesbank
The post-regulatory institutional settlement established by the Financial Stability Act (FinStabG) reorganized Germany's financial regulation system. In this new settlement the Bundesbank has been given a mandate to carry out economic analysis for the work of the AFS, thereby giving them the authority to provide the main problematizations of the economy. What actions did the Bundesbank take to reorganize itself based on its new environment? More importantly, to what degree were they successful in enrolling other actors in their new policy programmes? Lastly, given their institutional framework, were they successful in legitimizing themselves via outcomes once again? To establish how the Bundesbank reorganized itself requires an understanding of the new regulatory setup. The AFS has the power to issue warnings and recommendations, but BaFin is still the only institution that can perform administrative acts. Thus, after the financial crisis in the last instance the finance ministry is still in control of financial regulation in Germany, given that BaFin remains a subordinate institution to the finance ministry. With the Financial Stability Act the Bundesbank is for the first time directly mandated with maintaining financial stability and macroprudential regulation at the national level. It is tasked with identifying possible risks to financial stability and reporting them to the AFS. It must also publish an FSR once per year, issue warnings and recommendations, and review the proper implementation of recommendations.
In accordance with its new environment the Bundesbank reorganized itself to produce the expertise needed to analyze the state of the financial system and markets, as well as to develop formal and informal organizational structures to enrol BaFin and the finance ministry, in addition to political actors in their policy programme. After some trial and error in the early years of the new regulatory framework the directorate's eventual structure in 2016 would revolve around specific financial markets and actors. The current directorate includes a secretariat tasked with coordinating various political and international committees, a department tasked with tackling basic issues of financial stability and macroprudential supervision (developing methodology, data management and so on), a department for supervision of the banking sector and one for supervision of the non-banking sector, a department for systemic risk stemming from international financial markets and finally a department for international currency markets. 5 The intended effect of these changes was the merger of the political process of recommendations and warnings within the AFS with the economic analysis necessary for these recommendations. Consequently, the Bundesbank's knowledge production is now tightly coupled with its official task of recommending policy interventions.
Beyond formal changes to the financial stability directorate, changes to the procedure of macroprudential regulation were implemented to accommodate the quarterly meetings of the AFS. For example, a new committee called the Financial Stability Coordination Committee (Koordinierungsausschuss Finanzstabilität -KAF) was created. The KAF is headed by the financial stability directorate, but it also includes the macroeconomics and banking supervision directorates. The KAF aims to coordinate and prepare the Bundesbank directorates for meetings in the AFS. Furthermore, it is tasked with facilitating analysis in other directorates as they become necessary. Lastly, the KAF is the committee in which the different directorates discuss topics and analysis for the FSR, such as upcoming risks and evaluate when and if further actions need to be taken. In short, the KAF serves as an organizational structure designed to ensure the enrolment of the other directorates of the Bundesbank in the policy programme of the financial stability directorate.
The key role of departmental heads in the transition towards a macroprudential Bundesbank becomes apparent in the practices of preparing for KAF meetings within the financial stability directorate. KAF meetings are prepared with the help of an internal quarterly meeting of all departmental heads. These meetings utilize a new risk-matrix developed by the directorate in which potential systemic risks are mapped onto each market section under analysis by the Bundesbank. If, for example, a new risk is detected by analysts in a department, it is tested against the risk-matrix to identify how it would impact each market, what possible feedback effects could emerge and what further analysis is required. In addition, these meetings serve to discuss a possible recommendation to activate a macroprudential instrument. Within this newly created structure departmental heads play a crucial role as it is their hiring practices that help to stabilize the expertise within each department. At the same time, their position at the meso-level within the organization allows them to manage the problematizations of financial markets within the Bundesbank, while at the same time incorporate the possible enrolment of the organizational environment into their decision-making (Carpenter, 2020;Tilcsik, 2010).
With the establishment of the KAF and the risk-matrix the Bundesbank was able to overcome its initial decoupling and create the organizational technologies necessary to fulfil its role in the new regulatory setup. It is thereby able to act upon its new mandate and satisfy the first condition of producing output legitimacy, although the two conditionscreating a discursive space for its intervention and reconfiguring the material conditions in the stateeconomy boundaryremain.
To analyze this reconfiguration process, I now turn to the recommendation for new macroprudential tools for the regulation of housing markets in 2015. This recommendation is one example in which this enrolment process and the consequent boundary work between state and economy has had mixed results. As macroprudential instruments for housing markets adopted via Basel III were not initially developed for the German financial system, they lacked a legitimate discursive justification for their use, as well as the material conditions for their administration. In 2014 the Bundesbank therefore began to problematize the relationship between macroprudential housing instruments and financial stability for an eventual recommendation to German lawmakers on the creation of new macroprudential instruments. In 2014 it reported on results of survey data on mortgage lending in 23 cities in Germany. This analysis showed no significant risks or excessive rises in house prices at the current time, however through the use of stress tests the Bundesbank was able to argue that future downward scenarios might reveal vulnerabilities in the banking sector. The downward scenario is used to emphasize the feedback effects between increasing housing prices, general deterioration of macroeconomic conditions and possible financial stability risks to the banking sector (Deutsche Bundesbank, 2014, pp. 61-66). Crucially, the downward scenario is also used as a narrative of the future state of the economy, thereby shaping a mechanism through which the administration of macroprudential tools by the state in relation to banks could improve financial stability.
However, the analysis of the Bundesbank identified two obstacles to the use of macroprudential housing instruments. First, German banks use sustainable LTV ratios to evaluate a mortgage risk-profile, a variation of the standard instrument 6 suggested by Basel III. This makes the widespread administration of LTV ratios difficult. Second, accounting and data collection standards in Germany do not include sufficiently fine-grained information on borrowers and mortgages to calibrate possible macroprudential interventions. The AFS therefore recommended the creation of new accounting standards and four new macroprudential instruments (LTV ratios, debt-to-income ratios, debtservice-to-income ratios and amortization ratios).
While the enrolment of the Ministry of Finance and BaFin in the AFS was successful, the enrolment of parliament and market actors was only partially successful. In consultations on the recommendation, industry actors decried the new reform as prohibitive in relation to their lending operations and profits. They also objected to the high costs of creating the accounting infrastructure for future data requests. These lines of argument were picked up by parliamentarians, who mainly argued against the distributive effects of new housing instruments, the political salience of denying credit to lowincome borrowers and lastly its effect on the provision of affordable housing within Germany. As a result of these electoral considerations the eventual law only introduced two new instruments, LTV ratios and amortization ratios, discarding the two more politically perilous borrower-related measures (debt-to-income, debt-service-to-income). They also did not create new accounting standards for German banks, which severely hinders the Bundesbank in its efforts to rigorously analyze housing markets or even apply the few new instruments it has been given. Only after the Bundesbank repeatedly criticized the federal government for its lack of action and the ESRB issued a warning on housing markets in 2020 did lawmakers react and create new data requirements for banks. This created the material conditions for the administration of housing instruments. However, German lawmakers have once again failed to create borrower-related instruments, despite the ESRB's recommendation.
As the recommendation on housing instruments shows, the Bundesbank is well able to problematize the economy in line with its envisioned policy programme and even enrol BaFin and the Ministry of Finance in its programme. However, at crucial moments it struggled to enrol several actors. In other words, problematizations of financial and political actors intruded the problematization of the Bundesbank, therefore preventing a fully successful translation process. Despite the mixed outcome of the recommendation itself, the Bundesbank was still successful in establishing itself as the key authority on financial stability in Germany as it is their analysis and problematizations of the economy that function as the foundation for the work of the AFS.

Conclusion
This paper has documented how the Bundesbank reacted to a crisis of legitimacy in the early 2000s. It showed the internal transformation the Bundesbank underwent to compensate for its loss of monetary policy and banking supervision powers by adopting financial stability as a new policy area, at least 'ceremonially'. However, this ceremonial adoption of new policy areas was not without effect as the financial crisis created a demand for macroprudential interventions. Subsequently, the Bundesbank attempted to position itself as the new macroprudential regulator within the emerging German regulatory regime, although internal struggles over the role of the Bundesbank led to an institutional setup in which it shares responsibility for macroprudential regulation with BaFin and the Ministry of Finance. The Bundesbank's efforts to create a new state-economy boundary to legitimize itself via policy outputs was thus limited by its new institutional context. As its attempts to establish new macroprudential instruments for the housing market has shown, however, it could still be partially successful in translating its policy programme. These findings provide insights into the emergence of national macroprudential regimes, which so far have been explained mainly by policy learning (Yagcı, 2017), idiosyncrasies of macroprudential regulation itself (Goodhart, 2015;Stellinga, 2020), the lack of scientific backing (Thiemann et al., 2018(Thiemann et al., , 2020 or national politics (Mero & Piroska, 2017;Piroska et al., 2021). While these aspects play an important role, this study has shown how path-dependencies within central banks can become consequential in the institutionalization of macroprudential regulation at the national level, if they prevent the central bank from seizing macroprudential regulation for itself. Additionally, it showcased the importance of tracing where, when and how macroprudential expertise accumulated in central banks before the crisis because this can influence how macroprudential regulation is problematized and consolidated.
At a more general level this study points to the importance of legitimacy concerns at state agencies for the creation of new policy areas. Understanding why and how central banks can legitimize themselves via outputs, rather than via symbolic adoption of discursive expertise is crucial in explaining the emergence of new policy programmes. This study has proposed four conditions for this legitimization to be successful: an institutional arrangement which allows the organization to differentiate itself from other agencies in the polity; the availability of policy instruments; a tightly coupled organization; and sufficient boundary work to disentangle the discursive and material relationship to make policy instruments operational. The Bundesbank has failed to create an institutional arrangement in which it could claim financial stability as a policy output for itself, but it has managed to establish itself as the authority on financial stability in Germany. As a result, the Bundesbank managed to stave off its organizational crisis and regain a stable position in the German polity.
Future research should pay attention to areas outside monetary policy, in which central banks do not occupy the same central position at the heart of the banking system, because organizational processes and concerns might play out in an unusual fashion. In such cases, central banks might not be able to perform the boundary work necessary to create a state-economy boundary amenable to their policy interventions. Lastly, if they are unable to meet the abovementioned conditions central banks implicate themselves politically by engaging in these policy areas, thereby risking their organizational autonomy.
3 The international directorate cooperated closely with the IMF in its financial sector assessment programme (Deutsche Bundesbank, 2003, p. 166). 4 Unsurprisingly, the recommendation heavily suggested the involvement of national central banks in new financial stability regulatory regimes, as these banks played a major role in the transnational policymaking process (McPhilemy, 2016). 5 An analysis of department heads in the period 2009-2018 reveals that all but one were early adopters of macroprudential thinking within the Bundesbank. Furthermore, only two out of seven are former banking supervisors. 6 Exact sustainable LTV calculations can vary between banks, as well as regionally.