An analysis of the governance structure of the Single Supervisory Mechanism
AUTHOR: THOMAS A. DANENBERG 10422137
MASTER THESIS POLITICAL SCIENCE (POLITICAL ECONOMY) GRADUATE SCHOOL OF SOCIAL SCIENCES, UNIVERSITY OF AMSTERDAM
AMSTERDAM, JUNE 2019
SUPERVISOR: PROF. JONATHAN ZEITLIN 2ND READER: PROF. DANIEL MÜGGE
FINDING GOVERNANCE BALANCE FOR
1
Abstract
The Eurozone crisis generated momentum for the implementation of the Single Supervisory Mechanism (SSM), a renewed, centralized form of banking supervision involving cooperation between the ECB, as the coordinator, and the National Competent Authorities (NCA). The aim of the SSM is to eliminate the home bias of the NCAs and the cognitive capture of the national important financial institutions. The thinking is that a level playing field and centralized coordination should lead to a stronger and crisis-prone supervisory system. This study provides a detailed analysis of the new regime’s governance structure and the cooperation between the ECB on the one hand and the NCAs in general and the Dutch Central Bank in particular on the other. The analysis forms the basis for an examination of the appropriateness of the balance between centralization and uniformity in the governance structure. The findings suggest that the SSM is making progress towards achieving the right balance but that it is still facing challenges. First, the playing field is far from level because of the differentiation between the national banking systems. National options and discretions and a lack of trust in the European resolution scheme is leading to Member States ring-fencing their markets. Second, the decision-making process constitutes a problem because the decisions required from the decision-making authority, the Supervisory Board (SB), cannot be delegated without the agreement of the NCAs represented on the SB. For NCAs, delegation means renouncing part of their influence and the result is a stalemate when sensitive issues are involved.
2
Acknowledgements
I would like to thank all people that contributed in any kind to the realization of this thesis.
At first, my supervisor Prof. Jonathan Zeitlin, who guided me through the process with much dedication and very helpful and sharp feedback. Also, my second reader, Prof. Daniel Mügge, for finding time to review my thesis.
Second, my dear friend Nahuel Mercedes who helped me getting into contact with some great and prominent respondents. Without him it would have been a real challenge to speak to such leading actors.
Subsequently, I am very grateful that all respondents were more than willing to have a fruitful discussion with me unless their busy schedules. Their contribution to the thesis really made the research more valuable.
3
Table of Contents
Abstract ... 1 Acknowledgements ... 2 List of Abbreviations ... 5 Introduction ... 7The Single Supervisory Mechanism ... 7
Research problem ... 9 Objective ... 9 Research question ... 9 Relevance ... 11 Reading guide ... 11 1. Theoretical framework ... 12 1.1 Governance ... 12 1.1.1 Hierarchical Governance ... 13 1.1.2 Network Governance ... 14 1.1.3 Experimentalist Governance ... 15 1.2 Conclusion ... 17
2. Literature review on centralization and uniformity in Eurozone banking supervision ... 19
2.1 The SSM’s objectives and advantages ... 19
2.2 Limitations of the SSM... 21 2.3 Conclusion ... 23 3. Methodology ... 24 3.1 Research Strategy ... 24 3.1.1 Research design ... 24 3.1.2 Case selection ... 25
3.1.3 ECB-DNB relationship within SSM... 25
3.2 Methods ... 27
3.2.1 Document and theoretical analyses ... 27
3.2.2 Interviews ... 27
3.2.3 Validity ... 28
3.2.4 Limitations ... 29
4. The design of the Single Supervisory Mechanism ... 30
4.1 Institutional Structure ... 30
4.1.1 Joint Supervisory Teams ... 31
4.1.2 National Competent Authorities ... 31
4
4.2 The decision-making process ... 32
4.2.1 The Single Rulebook ... 34
5. Analysis of the SSM governance ... 36
5.1 The Supervisory Cycle ... 36
5.1.1 Discretion ... 36 5.1.2 Peer review ... 37 5.1.3 Revision ... 38 5.2 Coordination ... 38 5.2.1 NCA influence ... 39 5.2.2 National characteristics ... 39 5.3 Conclusion ... 40
6. Balance of Centralization and Uniformity ... 41
6.1 Uniformity and the level playing field ... 41
6.1.1 Options and Discretions ... 41
6.2 Centralization and delegation ... 44
6.3 Discretion ... 46 6.4 Conclusion ... 46 7. Conclusions ... 48 Bibliography ... 51 Primary sources: ... 51 Secondary literature: ... 51 Appendix ... 55 Interview respondents ... 55
5
List of Abbreviations
BaFin Bundesanstalt für Finanzdienstleistungsaufsicht
BCBS Basel Committee for Banking Supervision
CRD Capital Requirements Directive
CRR Capital Requirements Regulation
DG Directorate General
DGSSB DG Secretariat Supervisory Board
DNB De Nederlandsche Bank
DSTI Debt-Service-To-Income
EBA European Banking Authority
EBF European Banking Federation
EDIS European Deposit Insurance Scheme
EU European Union
EC European Commission
ECB European Central Bank
FSC Financial Stability Board
GC Governing Council
ICAAP Internal Capital Adequacy Assessment Process
ILAAP Internal Liquidity Adequacy Assessment Process
JST Joint Supervisory Team
MoF Ministry of Finance
MPF Macro Prudential Forum
MS Microprudential Supervision
MSD Methodology and Standards Development
NCA National Competent Authority
LSI Less Significant Institution
LTV Loan-to-Value
OND Options and Discretions
SB Supervisory Board
SI Significant Institution
SREP Supervisory Review and Evaluation Process
6
SRM Single Resolution Mechanism
SSM Single Supervisory Mechanism
SSMR SSM Regulation
SSMRF SSM Regulation Framework
7
Introduction
The outbreak of the recent financial crisis revealed the vulnerability and flaws of the financial system in the European Union (EU). As a consequence, numerous banks had to be bailed out by the Member States’ governments. Bank bailouts were paid for with public taxpayers’ money and the common understanding was this should never happen again. During the crisis, Jacques de Larosière, chairman of the High-Level Group on Financial Supervision in the EU, presented a report which served as the first template for the reformation of banking supervision in the Eurozone. In the Financial Times, de Larosière stated that “if the macroeconomic and macroprudential surveillance had really been in place, in the way the report proposes, things would have been profoundly different” (2009). The consequences of the crisis paved the way for a more influential role for the European Commission (EC) and the European Central Bank (ECB). Epstein and Rhodes (2016) explain that the EC and the ECB, together with the crisis-stricken Member States, concluded that it would be necessary to sever political bank-state ties in order to support the common currency’s credibility and sustainability. The ‘cozy’ bank-state relationship resulted in a kind of banking nationalism, with the states focusing more on competing internationally with their national champions than on limiting the financial risks (Epstein & Rhodes, 2016).
The presented solution took the form of the European Banking Union. The task of supervision for banks in the Eurozone was centralized by being transferred to the ECB. To restore public confidence in the common currency, an integrated supranational executive organization was needed (Ioannidou, 2012). The Banking Union is a supranational set of three pillars. It consists of the Single Supervisory Mechanism (SSM), the Single Resolution Mechanism (SRM) and the European Deposit Insurance Scheme (EDIS). In this arrangement, the ECB has been given responsibility for the exercise and coordination of supervision in cooperation with the national supervisory authorities. The cooperation between the ECB and the individual national competent authorities (NCA) within the first pillar, the SSM, is the central focus of this thesis. The SRM is intended to safeguard the orderly resolution of failing banks, and to minimize the associated costs for taxpayers and to the real economy. The same banks that will be supervised directly by the ECB will be covered by the scope of the Single Resolution Mechanism. Deciding on the resolution of a bank at the European level should help to take into account the consequences of the resolution for the whole euro area [BRRD, (1), (2)]. The proposed EDIS is not yet operational. This means that the deposit insurance scheme is still structured on a national basis.
The Single Supervisory Mechanism
The mechanism involves cooperation between the ECB and the NCAs. The ECB is responsible for the supervision of all the banks located in the Eurozone. As shown in Figure 1, the ECB directly supervises the Significant Institutions (SI), including the systemic cross-border banks. That means that
8 it is indirectly responsible for the Less Significant Institutions (LSI)s. The NCAs directly supervise the LSIs, which are small to medium-sized banks.
Figure 1 (ECB, 2018: 8)
An NCA may be the national central bank but it can also be another national authority such as the Bundesanstalt für Finanzdienstleistungsaufsicht (BaFin) in Germany. The SSM pursues a more formal approach in the supervisory framework in order to make close ties between national banks and supervisors less likely. It also aims to make European supervision more uniform and harmonized. Harmonization and uniformity are reached through compliance to the Single Rulebook and the ECB Supervisory Manual (ECB, 2018). Despite this emphasis on uniformity, the SSM was not designed to promote a ‘one-size-fits-all’ approach to banks throughout the Eurozone. The objective is rather, as Supervisory Board (SB) Chair Daniele Nouy explains, to ‘ensure consistency across institutions and supervision tailored to [their] specificities … by balancing uniform supervisory anchor points with constrained supervisory judgement’, thereby accommodating banking diversity, which remains ‘very desirable for financial stability’ (Nouy, 2015). Financial stability is further enhanced through the continuous improvement of the supervisory manual. Best practices from local supervisors complete the supervisory approach throughout the Eurozone.
Supervisory decisions in the SSM are prepared by the SB, which is composed of ECB and NCA representatives, and adopted by the Governing Council under the non-objection procedure that is described later in Chapter 4 (ECB, 2018). The ECB acts as coordinator and decision-maker, but not the sole executor of the supervisory task, because of considerations relating to the distribution of expertise. The NCAs have years of experience with supervising financial institutions and they know
9 their home markets and financial institutions best. This national expertise must be preserved, and the supervisory framework has therefore designed as a mechanism in which the ECB occupies a central position (Moloney, 2014). SIs are not supervised solely by the ECB but also in cooperation with the relevant NCAs. These supervisors are grouped in Joint Supervisory Teams (JST) that supervise the SIs on a day-to-day basis. International cooperation in the JSTs allows for a better oversight of systemic risk in the system. Moreover, the NCAs also sit on the SB of the SSM to represent their national interests. The SSM therefore implies that the ECB is the coordinator with responsibility for the supervision of SIs, but the process is highly cooperative and not exclusively top-down in nature.
Research problem
Objective
The objective of this study is to analyze how the different actors in the SSM interact with one another. The institutional framework is complex but it should, after thorough research, become clear. There may be discrepancies with how interactions actually take shape in practice. The role of the Dutch National Bank (DNB) in relation to that of the ECB is not clearly defined in some respects and it remains open to interpretation. The study aims to identify how governance operates in the SSM and to analyze whether this form is appropriate for the supervision of significant financial institutions.
Research question
The creation of the Banking Union and the SSM implied the centralization and uniformity of banking supervision in the Eurozone. The NCAs operate under the supervision of the ECB and they have given up their autonomous position. However, this does not necessary imply that the NCAs no longer have any influence.
More centralization and uniformity can be formalized in different structures. Centralization normally means a more top-down, hierarchical form of governance. Nevertheless, it may also include some horizontal characteristics of network forms of governance. Local units may, for instance, work together at the supranational level as a self-organizing network. Alternatively, the governance of the SSM may acquire a more intermediate form between centralization and decentralization, as in experimentalist governance.
The new SSM framework is complex and, in some senses, innovative. It incorporates elements from several different forms of governance and a deeper study should produce a nuanced answer to the question of which form comes closest to capturing the SSM. However, the essential question relating to the form of governance is whether the balance of centralization and uniformity in the governance between the ECB and the NCAs is well-distributed and appropriate for banking supervision.
Ioannidou (2012) stated, before the SSM was introduced, that the European supervisor will obviously need to oversee the national supervisors in a clear hierarchical structure and possibly have a representative regularly present in the organizations of the national supervisors. However, not all
10 scholars or stakeholders support a clear hierarchical structure of this kind for supervision. Sijbrand, the former Director of Supervision at the DNB, stated in his farewell speech (2018) that joint supervision has many advantages but the downside is that it is too rule-based.
The issue here is whether Eurozone-wide centralization and uniformity is the preferable solution for the reform of banking supervision. The cooperation between the ECB and the NCAs is shaped in dialogue between the two but has it worked out the way which is most appropriate for the purposes of supervision? Do the preferences of the ECB and NCAs differ on the subject of centralization or uniformity? The JST is the platform where both the ECB and the NCAs meet and act and where this problem is at the core. Do the NCAs have enough discretion to function properly or is there no need for such national discretion? To analyze these issues, the cooperation between the ECB and the DNB has been selected as a case study. This focus of the present thesis is expressed in the following research questions:
“What degree of centralization and uniformity can be identified in the current governance structure of the Single Supervisory Mechanism?”
“Is the balance in the case of the cooperation between the European Central Bank and Dutch Central Bank appropriate for banking supervision?”
The research questions can be broken down into six areas. The first area involves the elaboration of the three governance theories - hierarchical, network and experimentalist governance - that constitute the basis for the analysis of the SSM. All three governance theories include characteristics of the SSM and they can be used to explain how the system operates. The second are is how the SSM operates in terms of the three theories. The third area relates to the degree of centralization and uniformity in SSM governance, which will be addressed in a review of the secondary literature about the possible limitations of the governance structure. The fourth area is the design of the SSM and it will be followed by a fifth area: the perceived functioning of the DNB. Finally, the sixth area will be the balance between the centralization and uniformity identified in the governance structure. These areas can be stated as the following sub-questions:
1. “How can the Single Supervisory Mechanism be characterized in governance-theoretic terms?”
2. “Which forms of governance can be identified in its operations?”
3. “What is the appropriate degree of centralization and uniformity for SSM banking supervision?”
4. “How is the Single Supervisory Mechanism designed?”
5. “How do the different levels of governance in the Single Supervisory Mechanism relate to one another in the case of the DNB?”
11 6. “Does the Single Supervisory Mechanism strike the right balance between centralization and
uniformity in the current governance structure?”
Relevance
Since the SSM has been recently created and many elements have not yet been studied, the search for more transparency in the Eurozone’s banking supervisor is highly relevant. Little is known or published about the internal functioning of the SSM. Recently, in June 2017, the first test for the Banking Union in general was a success. The SSM signaled the risk for Banco Popular, then Spain’s sixth-largest lender, and the SRM handled a run on the bank (Financial Times, 2019). This is, however, not real proof since it involved a single medium-sized bank and it would not have led to panic in the pre-Banking Union era either. How the SSM operates and whether it is effective and well-designed for its supervisory task is still a field of debate. As will become clear from the literature review, centralization and uniformity are still areas where supervisors and scholars have not arrived at a consensus. Moreover, the SSM governance structure has not yet been examined properly. This means that most research results will be ‘new’ information. Apart from the academic relevance, the examination of the SSM governance structure is also very relevant in social terms to understand what new system is required to protect financial stability and, in a broader sense, taxpayers' money.
Reading guide
Following this introduction, Chapter 1 of this thesis provides an examination of the available theory on governance, addressing the theories of Hierarchical, Network and Experimentalist governance. The chapter concludes with an answer to the second sub-question and discusses the theory in relation to the operations of the SSM. Chapter 2 consists of a review of the literature on centralization and uniformity relating to the appropriate degree of banking supervision. The discussion delineates the strengths and limitations of the SSM balance found in secondary literature. The research strategy and the methodological choices are explained in Chapter 3 and further details are given about the case of the DNB. After the methodological chapter, the analysis of the SSM is presented in three chapters. The first, Chapter 4, provides a thorough overview of the SSM design on paper followed by the examination of the functioning of the DNB in practice in Chapter 5. The final chapter of the analysis, Chapter 6, examines the right balance between centralization and uniformity in the SSM. Finally, the thesis ends with conclusionary remarks and recommendations for further research.
12
1. Theoretical framework
The design of the SSM presents a complex web of actors interacting and cooperating in various ways. Governance theory can contribute to analyzing the SSM. With a clear description of the governance structure and form, the SSM’s relation to banking supervision in the Eurozone can be established. Governance is a rather broad concept and variously used in different contexts. To sharpen the discussion about governance forms, first, the concept of governance is operationalized. Thereafter, three forms of governance are elaborated, and their application to the SSM governance structure is discussed.
1.1 Governance
The concept of governance is widely used among a diverse set of research disciplines and has even filtered through the media into public debate. The concept is known for its applicability to fields where a mix of competencies and functions creates complex structures of decision making (Bartolini, 2011). In political science, governance has increasingly become a central concept, especially in analyzing changes in western democratic societies. But governance itself is an overarching concept covering many issues. A single precise definition of its meaning has not yet been formulated and presumably will never be either. There is no consensus about what governance entails and what it does not. David Levi-Faur (2012: 3) is one of many scholars who have tried to define it and kept the definition at such a broad level that it is workable and respected. Levi-Faur defines governance in the Oxford Handbook of Governance as “an interdisciplinary research agenda on order and disorder, efficiency and legitimacy, all in the context of the hybridization of modes of control that allow the production of fragmented and multi-dimensional order within the state, by the state, without the state, and beyond the state”. As this definition indicates, governance is not solely related to the state and should, therefore, not be mixed up with ‘government’. Rhodes (1996) explains in his famous article, ‘the shift from government to governance’ is two modes of control. He refers to those as ‘government’ deals only with state actors and excludes civil society and ‘governance’ can be more of a co-production of norms and public goods by a variety of private and public actors. Governance is not a synonym for government but rather a change in its meaning, referring to a new process of governing or a changed condition (1996: 652-653).
The rise of governance appeared with a widespread consensus that society was changing and that there was even a transformation and paradigm change. The proclaimed shift in governance analysis and the changing nature of authority from a single authority regime towards a system of different spheres of possible institutionalized authorities which can compete, bargain, or coordinate among themselves or ignore each other (Levi-Faur, 2012). The shifts are conceptualized in three different directions (upward, downward, and horizontally) but in this case, the upward direction is relevant. The
13 shift in governance is often explained by a dichotomy, centralized versus decentralized or soft versus hard law (Bartolini, 2011), but such a simplification is insufficient since in most cases, the reality will be somewhere in the middle of that spectrum. In the following sections, the different forms of governance are elaborated which are most applicable to the SSM design. The forms discussed are hierarchical, network and experimentalist governance.
1.1.1 Hierarchical Governance
Hierarchical governance is seen as the starting point in governance literature. It has the characteristics of government to which Rhodes (1996) referred. Hierarchical forms of governance had an important contribution to the successful rise and persistence of nation-states in modern times (Zielonka, 2007). The current literature on governance is, as mentioned above, an explanation of a new set of processes, attributes or other features of governance towards a completely new form or at least a shift away from hierarchical governance (Craig & de Burca, 2015). But before the transition to a new form can be explained, the ‘old’ form must be defined. The term old does not truly mean that it is old-fashioned, because the hierarchical form of governance is still used in many organizations. Hierarchical governance implies that policies come from ‘above’ or from the center, and thus are top-down. These policies are relatively complete and prescriptive and do not leave much room for discretion at the lower level of policy implementation. Policies are obligatory and binding for those to whom they apply and generally allow for compulsory legal enforcement (Ibid). This prescriptive and complete steering helps to reduce diversity and foster homogeneity within its territory. Hierarchical governance is strongly related to the bureaucracy described by Weber, therefore, also to the principle that rules should be applied equally, which requires uniformity and standardization (Levi-Faur, 2012).
However, hierarchical governance policies being top-down does not mean that delegation or decentralization is not possible. Zielonka states that “hierarchical governance can also envisage a considerable degree of decentralization and delegation” (2007, p.191). Here the principal-agent theory is linked to hierarchical governance, namely goals are defined by principals and responsibility for implementing the policy is devolved to lower-level agents. Important here is that hierarchical governance allows for some discretion but always modestly because the scope of the agent is clearly defined and so are its tasks. But as the political science view on principle-agent suggests, this theory is based on the principle of bounded rationality. Consequently, the contract between the principal and the agent is incomplete, which makes some discretion necessary (Héritier & Lehmkuhl, 2008). Hierarchical governance, like Weberian bureaucracy, is based on clear lines of control and responsibility (Zielonka, 2007). Craig and de Burca (2015) also, therefore, characterize the hierarchical form of governance as command-and-control regulation. Although the governance literature is focused on this shift, command-and-control regulation is not only a mode of the past. Zielonka explains that “although with the passage of time states have changed their nature and mode of functioning, the principles of hierarchical governance are still very much in place” (2007: 191). The
14 basic structure of the states is still an easily identifiable center of authority with a built-in hierarchy and division of task. Zielonka (2007: 188) argues that the aim of an “ever closer union” and institutional fusion at the EU level are indications of the persistence of the principles of hierarchical governance in the EU. Yet, increasing challenges of complex society and transnationalization, among other developments, demand more innovative steering capacities by governments and other spheres of authority (Levi-Faur, 2012). This occurs more in a rather horizontal way than in vertical steering through hierarchical forms of governance. The next two forms of governance are two of those ‘new’ forms of governance.
1.1.2 Network Governance
The traditional hierarchical form of governance was not sufficient to explain new developments that emerged during the 1970s and 80s. The state was no longer the single authority in policymaking and there grew a rising importance for civil society to became part of the process. The simple top-down prescriptive policy was not consistent with new challenges which a changing society yields. Social problems such as social dislocation or drug abuse plagued communities despite efforts to resolve it. Such problems pose a more complex challenge because they defy precise definition, cut across policy and service areas, and resist solutions offered by a single agency (Keast et al., 2004). Relations between private and public actors created new opportunities for improved efficiency, enhanced learning, and increased capacity to address complex problems (Provan & Kenis, 2007). Torfing (2005) states that a new interactive form of network governance was according to central decisionmakers in the public and private organizations the most suitable response to the increasing societal fragmentation and dynamism.
Torfing (2005) defines network governance with the commonly known features described by Rhodes and Jessop as: “(1) relatively stable horizontal articulations of interdependent, but operationally autonomous actors who (2) interact with one another through negotiations which (3) take place within a regulative, normative, cognitive and imaginary framework that is (4) self-regulating within limits set by external forces and which (5) contributes to the production of public purpose”. The definition makes no claim to originality but rather aims at capturing the common understanding of network governance. The relation between the stakeholders is not hierarchical and top-down but rather horizontal and based on autonomous actors. Governance is, therefore, more about negotiation and persuasion than about control and steering. Zielonka (2007) adds to this that incentives can result in better outcomes than sanctions and coercion. Participation is, therefore, voluntary and actors are free to leave the network.
Keast et al. (2004) also state that mutual concerns and incentives in a network in which people are loosely linked to each other create more opportunities to work actively together and achieve a rational solution for the shared problem or issue. Network structures are, consequently, distinguished from traditional hierarchical structures because there is no single rule maker. However, this does not mean
15 that no actor can be in charge of the network and establish the formal rules of collaboration. It means that the traditional forms of governance and the way of steering and control are not effective in a network governance structure. Moreover, unilateral power is not effective in a network structure, as all actors should be taken into account (Keast et al., 2004). Provan and Kenis (2007) have found four key structural and relational contingencies for the adoption of network governance: trust, size (number of participants), goal consensus, and the nature of the task (specifically, the need for network-level competencies). These four elements are comparable to those specified in the general cooperation and collective action literature (Olson, 1982).
In some cases, the network does not function under formal hierarchy but under the so-called the ‘shadow of hierarchy’. Policymaking then is delegated to a self-regulating network, but the government or principal still keeps a sanctioning instrument as a backup (Coen & Thatcher, 2008). Heritiér and Lehmkuhl (2008) explain that a governmental principal is willing to delegate because of the extra expertise of a network or a lack of time. They explain the shadow of hierarchy as “Legislators can threaten to enact adverse legislation unless potentially affected actors alter their behavior to accommodate the legislators’ demands” (Heritiér & Lehmkuhl, 2008: 2). So, although the networks operate in a rather horizontal manner, the possibility of vertical hierarchical control could still persist. As Klijn and Koppejan (2012) already stated, these complexities may be such that they cannot be captured in one single theory. The challenges for public administration produce new approaches to governance and network theory. The next mode of governance, experimentalist governance, is such new approach.
1.1.3 Experimentalist Governance
Experimentalism could be explained as a form of governance which combines elements of the first two forms discussed above. Experimentalist governance is neither a purely decentralized nor a centralized form of governance. The theory was introduced by Sabel and Zeitlin (2008) as an answer to the far-reaching transformations in the nature of contemporary governance. In their article ‘Learning from Difference: The New Architecture of Experimentalist Governance in the EU’ Sabel and Zeitlin outline the new and innovative ways of the governance structure that grow out of the EU’s own traditions and allow them to pursue their own best judgements for innovative advances.
Experimentalist governance, also called directly deliberative polyarchy, is as the title of the article by Sabel and Zeitlin (2008) indicates, about learning from the differences within and between organizations. Diversity is a core element in the theory as it makes space for different approaches to the same issue or objective. Sabel and Zeitlin describe the theory as a multi-level governance structure with an iterative process. The process contains four key elements or stages. In the first stage, ‘local’ and ‘central’ units deliberately produce provisional broad framework goals, potentially together with relevant civil society stakeholders. In the second stage, the local units implement the broad objectives following their own approach. With this discretionary freedom, the organization creates the diversity
16 needed for the effectiveness of the system. In the third stage, the local units must regularly report on their discretionary choices and performances through a process of peer review in which their results are compared with results of other local units pursuing the same goals. In the fourth and final stage, the goals, metrics, and decision-making procedures themselves are periodically revised by a widening circle of actors. The resulting objectives and goals then serve as the point of departure for a new cycle which repeats itself (Sabel and Zeitlin, 2012a: 169-170). In a later article, Zeitlin (2016) explains that this four-step architecture does not always match with observed governance structures. In cases like EU chemical policy where integration is concerned with the internal market, a harmonized set of rules is created which give the lower-level unit a limited level of discretion. The difference, however, from a typical hierarchical form of governance lies in the provisional character of the rules. The lower-level units still can report on new information, and implementation experience. Moreover, they can explore alternatives through review processes, while the organizational center is obliged to take account of such local experience in reconsidering and revising the rules (Zeitlin, 2016).
The possible conditions for experimentalist governance are according to Sabel and Zeitlin (2008) quite minimal. They assume that there must be strategic uncertainty, meaning that policymakers cannot precisely define their goals and how best to achieve them in advance; and that there is a polyarchic distribution of power, which means that no single central actor can govern or has the capacity to force its own preferences and approaches without taking other stakeholders into account. De Burca, Keohane and Sabel (2013) make the conditions stricter. They enumerate five conditions which need to be present to label a governance process as experimentalist: openness to participation of relevant stakeholders; articulation of broadly agreed common problem and the establishment of a framework understanding setting open-ended goals; implementation by lower-level actors with local knowledge; continuous feedback, reporting and monitoring; and established practices, involving peer review, for revising rules and practices. This stricter definition of experimentalist governance, however, does not add value to the analysis of the governance structures. The broader view of Sabel and Zeitlin gives the opportunity to signal governance structures which are in transition towards a more experimentalist architecture. The described form of experimentalist governance is an ideal type which can be matched closely but almost never entirely. Actors can, for example, try to achieve a more experimentalist structure when their current structure is hierarchical.
The alternative definition of Experimentalist Governance as Directly Deliberative Polyarchy captures the essence of this form of governance. It is direct because it uses concrete experiences of actors’ and differing reactions to current problems to generate novel possibilities. It is directly deliberative because the novel possibilities are taken into consideration by the use of arguments. The settled practices are open for reconsideration in a cooperative way. And it is polyarchic because no single actor has final decision-making powers; rather it is a system in which the local units learn from, discipline, and set goals for each other (Sabel & Zeitlin, 2008).
17 The last point is an important factor in the difference from the first two described forms of governance. The polyarchic system together with broad open-ended objectives is not suited for the principal-agent model. The principal is not able to provide well-defined objectives which makes it impossible to incentivize other actors – the agents – to achieve them. When there is a strategically uncertain situation the principal’s goals are reformed by the agent’s efforts at implementation (Sabel & Zeitlin, 2012b). The element which makes the deliberative aspect pervasive is the ‘penalty default’ (Börzel, 2012). The incentive for the agents to operate and bargain fairly in a public way stems from the exchange with the hierarchical principal. The ‘penalty default’ is defined as the threat to engage in traditional rulemaking that is disruptive and produces dysfunctional results (Sabel & Zeitlin, 2010: 14). The threat of a possible policy outcome imposed by the central authorities makes the actors willing to compromise and produce an agreement in which they are all better off (Börzel, 2012). The impracticality of solutions or failures of rules, directed by traditional means, makes the mere threat of imposing them so effective for inducing parties to deliberate in good faith (Sabel & Zeitlin, 2008).
1.2 Conclusion
All three governance forms are present in the current European Union governance structure (Zielonka, 2007). It is hard to prove which model is best suited for the Union today as it depends strongly on the context and the policy subject. Different organizational structures are likely to be appropriate for different types of systems. Therefore, the three governance modes described above have a certain level of applicability to the system of the SSM. The suitability of the governance form depends on the policy-making process. Network Governance prevails in a situation where autonomous groups cooperate in a self-regulating network, without a single authority with decisive powers. Just as Experimentalist Governance, the goals are open-ended, and deliberation is an important aspect. For Hierarchical Governance, the objectives are more specified, and a single authority has decision-making power.
In the situation of the SSM, the centralization of banking supervision means a more horizontal structured cooperation between the separate NCAs in the form of the JSTs. The JSTs perform a more uniform assessment of the systemic banks than in the previous fragmented system. The previous system could be explained as a type of network governance. The separate NCAs cooperated as autonomous actors in a self-regulating system. The crisis showed, however, that this governance system was not working well because supervisors had a too close relationship with their banks and the sharing of information was not optimal. For the governability of the SSM, a network form of governance would also not be preferable because it would make the governance very complex and dragging, something that is not desirable in such a fast-moving and complex field as banking supervision. In vertical terms, the relationship between the ECB and the NCAs is based on the mechanism of cooperation. There is a hierarchical distribution in decision making, but it is not purely
18 authoritarian. The decision-making process is deliberative in the sense that all NCAs are represented in the SB, the highest organ of the SSM. A majority of votes is needed for decision making.
SSM governance is, therefore, more leaning towards Experimentalist Governance. First, decision making occurs through a polyarchic distribution of power because a majority of votes are needed for a decision. Second, the objective for the SSM is open-ended since the approach to achieve financial stability is not straightforward. The Supervisory Manual of the SSM remains open for revision and improvement. Besides the two possible conditions, diversity is also a crucial element in Experimentalist Governance. Diversity is, as Nouy (2015) puts it, important for financial stability and is emphasized in the differentiation of approaches to banks. At the same time, it is limited through the compliance to the uniform Single Rulebook. On the basis of theory, an experimentalist form of governance would, therefore, seem most applicable to the SSM governance. Network Governance has proven to be insufficient through the projection of the previous network of supervision. It remains to be seen how much hierarchical features the SSM possesses. The deeper analysis in chapters four and five present an answer to this issue.
19
2. Literature review on centralization and uniformity in Eurozone
banking supervision
The recent financial and euro crises demonstrated the weak spots of the EU’s decentralized banking supervisory architecture. Cross-border banks were supervised through cooperation between NCAs, but this was demonstrated to be insufficient in times of stress and crisis. The failure of the Fortis Bank case is the textbook example of the international resolution regime and shortcomings of the cooperation between national banking supervisors (Kudrna, 2012: 288-290). Fortis Bank was a systemic bank in all three Benelux countries, and it was a banking group which had one of the most developed ex ante cooperation arrangements, even with a specific Regional Memoranda of Understanding between the Netherlands and Belgium. Nevertheless, when the Euro crisis caused trouble for Fortis, the authorities failed to sustain a multilateral resolution.
The solution to overcome such problems has been the SSM and the centralization and uniformity paired with it. In this section, the appropriate degree of centralization and uniformity in the SSM in relation to the task of banking supervision is discussed. Is the new supervisory architecture better designed for the supervision of the banking sector? The risks of over-centralization and under-centralization are possible concerns. The startup phase of the SSM and the Banking Union as a whole was characterized by political competition between member states, but also between the NCAs and the ECB. Gren, Howarth, and Quaglia, among others, (McPhilemy, 2016; De Rynk, 2016; Donelly, 2014) describe the outcome of this political struggle as a compromise: “The European Council agreed to the SSM Regulation in December 2012 which represents a compromise that assigned the ECB responsibility ‘for the overall effective functioning of the SSM’ and ‘direct oversight of the euro area banks’. This supervision, however, would be ‘differentiated’, and the ECB would carry it out in ‘close cooperation with NCAs’” (Gren, Howarth & Quaglia, 2015: 1). The question remains, if the SSM is the outcome of political competition, is this outcome an effective design for banking supervision? Are the NCAs influential enough or should they play a larger role within the mechanism? And is this uniform approach attainable in the differentiated Eurozone banking system? These questions are discussed in this section. First, the objectives and the positive aspects of the SSM design are explained and, thereafter, the critiques on the SSM structure.
2.1 The SSM’s objectives and advantages
The pre-SSM regime of banking or prudential supervision was based on an increasingly detailed harmonization of prudential rules at the EU and national level, while the application of these rules took place at the level of the national supervisory agencies. This led to diversity among the Member States and caused numerous distortions in the supervision of multinational banking groups. Consequently, this prevented a comprehensive view and approach to the European banking system as a whole (Wymeersch, 2014). This increased contagion was risky and the internal financial market remained incomplete and insufficient. Another burden was the administrative element, as the different
20 Member States, with their varying views of supervision, created extra costs as banks were confronted with all the non-harmonized supervision systems (Ibid.). The transformation of the supervision into a centralized system was, however, a rupture by the view on supervision of the past (De Rynk, 2016). The Capital Requirements Directives of the pre-crisis period, agreed upon in 2006, had been based on the principle of ‘minimum harmonization’, which permitted the Member States to impose more stringent, ‘gold plated’, standards if they so choose (McPhilemy, 2016).
The objective of the SSM was to diminish these problems and more importantly, to close the ties and lax supervision in some of the Member States. This argued for central coordination of supervision and followed in complete opposition to the principle of the 2006 Directive, namely the ‘maximum harmonization’ principle (McPhilemy, 2016). The aim was to get rid of banking market fragmentation and the national anti-competitive incentives to prop up failing institutions, deposit flights, and destructive feedback loops between euro-area sovereign debt and the banking system (Moloney, 2014). Another important objective was the mitigation of systemic risk related to financial institutions.
The maximum harmonization principle means in practice that the European legislators and the Commission adopt more regulations instead of directives since these are directly applicable in the national legal systems and do not require transposition. Hence, regulations create more uniform law, because once adopted and promulgated, they are immediately applicable in the national jurisdictions (Wymeersch, 2014). Through this harmonization, the SSM regulation aim at putting an end to diversity by centralizing prudential supervision in the hands of one authority. A single authority is able to ensure a level playing field and subject each financial institution to the same supervisory method. A level playing field provides each financial institution with the same operating conditions and evens the competition. These intend to eliminate regulatory arbitrage and the favored position of ‘national champions’, while the strict supervisory practices hope to restore confidence in the banking system (Ibid.).
National champions are the large banks which are considered important players in their national economies. Such banks were closely associated with the political authorities and have been allowed more flexibility, illustrating a form of moral hazard (Ibid.). Still, differences will continue to apply in national law due to the existing financial structures and national options and discretions (OND) documented in the SSM regulation. ONDs are provisions on the European regulatory framework for credit institutions, most notably the Capital Requirements Regulation (CRR) and the Capital Requirements Directive (CRD IV), which allow the NCAs to decide on specific aspects of their application (Nouy, 2017). The ONDs can be warranted to serve as an additional part of the regulatory framework reflecting the different national systems, or the specific cases of a single institution (Ibid.). A final important advantage of the SSM structure is the introduction of the Joint Supervisory Teams. The coordination of the supervision of Sis, such as cross-border banks, is much more balanced, more
21 efficient, and less biased than in the past (Lannoo, 2014). A situation such as with the Fortis case is out of the question with the supervisory teams because the ECB’s representative is responsible and is always an official from another state than where the bank is headquartered or where it has its largest operation (Regulation ECB, 468/2014). Nouy (2014), formulated the working of the JSTs as: “We will be a truly pan-European supervisor operating without national bias or prejudice.” For example, she indicated that Crédit Agricole’s chief supervisor will be a German national, Unicredit’s from France and ABN AMRO’s from Spain (de Volkskrant, 2014).
2.2 Limitations of the SSM
The general conception of the Banking Union and the centralization and uniformity of supervision through the SSM is positive. A monetary union without a complementary banking union exacerbates systemic risk, precisely what the financial crisis in the European Union showed (Wagner, 2012). The Banking Union and especially the SSM were presented as a completion of the Economic and Monetary Union through which it would be saved (Howarth & Quaglia, 2013). But the SSM design does have some challenges to overcome, which are consequential from the way uniform supervision is centralized. In this section, two challenges for supervision within the SSM framework are discussed. The first is the challenge of systemic risk arising from the maximum harmonization principle. The second challenge is related to the negative consequences of diminishing the responsibilities of the NCAs.
The first challenge is a possible hazard of harmonization of regulation and supervisory functions. The SSM has the objective to harmonize and converge supervisory activities in the Eurozone. It follows the maximum harmonization principle and aims at creating a level playing field for banks across the Eurozone. Wagner (2012) and Romano (2014), however, state that the harmonization principle also poses new systemic risks for the banking system. Wagner states that: “When harmonizing regulation, European supervisors should not fall into the danger of encouraging more similar financial systems across countries. In order to mitigate the risk of systemic crises, we need a diversity of approaches to financial intermediation in the Eurozone” (2012: 121). He argues that there needs to be a diversity of approaches because if all institutions are subject to the same supervisory and regulatory environment, they will tend to undertake similar activities and respond to market changes in similar ways. This increases the risk of joint failure in supervising the banks. It is important to keep a certain level of diversity in institutional structures and strategies while creating a level playing field (Wagner, 2012).
Romano adds to this argument by reasoning that regulators always lag behind the ones regulated in a dynamic and uncertain environment in which financial institutions operate, stating, “In this kind of environment, a more nimble regulatory apparatus that can continually test, learn and adapt to changed circumstances and new information would have decisive advantages over Basel's cumbersome, top-down, harmonized approach” (Romano, 2014: 24). The Basel Accords, she argues, are facilitating ‘herding’ by establishing global standards and encouraged herding particularly on the
22 type of asset that proved to be the most problematic to be held during the financial crisis, namely the Mortgage Backed Security. Herding also increases the probability for financial institutions to be bailed out by the government. The regulator might be willing to let one financial institution fail but certainly does not allow to let nearly all the banks fail (Romano, 2014). Instead of a centralized and harmonized set of regulations for supervision, Romano opts for a mechanism that introduces regulatory diversity within the Basel Accords framework which would incentivize for regulatory innovation. It is, namely, disputed whether one regulatory framework can include all risks when there is no strong consensus about how to best measure systemic risk. To better understand the best way of measuring systemic risk, for instance, a variety of measures would help to determine the best practice (Ibid.).
Gortsos (2015), however, points to the SSM Regulation [Article 1, SSM Regulation] wherein is stated that the ECB must have full regard to the different types of business models and sizes of credit institutions, as well as the systemic benefits of diversity in the banking system for ensuring the safety and soundness of credit institutions. The difference between the two perceptions lies in the scope of diversity. The SSM regulation emphasis the difference in banks’ internal organization, while Romano and Wagner emphasize the national systems in which the banks operate. Zeitlin (2016) explains a comparable system as Romana but more in line with the SSM Regulation. He explains the SSM in terms of the simplified experimentalist architecture, discussed in the theoretical framework, as a hierarchical system with a critical role for local supervisors in revising the Supervisory Manual through local experience and changing environments. Best practices are developed through learning by testing and correcting problems in the methodology (Zeitlin, 2016). But the emphasis in diversity should be found in the different approaches to specific banks. The rules and methods for each bank are the same but the application and assessment of each bank can be completely different. It remains to be seen if the diversity between banks is substantial enough to strengthen and innovate the Supervisory Manual, or that herding still troubles the diversity between banks.
The second challenge is related to the role of the Nasard specifically, the local supervisors in the SSM. The role of the NCAs is critical in the SSM due to their expert knowledge of their home markets and financial institutions. One of the objectives for the SSM was to break capture risk and home bias in the supervision of national champions. Tröger (2014) argues, however, that the close relationship and initiative of the local supervisor also is a valuable resource. The common conception is that it enlarges the potential for moral hazard, but he raises the question that perhaps it could be that the close ties are a key element in the improvement of supervision. Such closer ties offer the opportunity to learn also from the financial institutions and that lax oversight is not necessarily associated with it (Tröger, 2014). Jan Sijbrand, former Director of Supervision at the DNB, also made such a point in his farewell speech. He stated that the joint supervision is a little too rule-based and that the DNB before the SSM worked hard to ensure a more principle-based form of supervision (Sijbrand, 2018). Tröger (2014) advocates for the view that the NCAs are not a kind of black box where the local information
23 automatically comes out. The local supervisors are not robots but are agents who need an incentive to go the extra mile. He does not argue that the local supervisors have a commitment problem with the ECB but suggests that the most valuable advantage of the local supervisor could be lost in this system. Tröger explains this as follows: “If bureaucrats in NCAs are not positively incentivized to voluntarily unveil deficits that their idiosyncratic know-how allows them to detect, even the most plausible advantage of supranational supervision, forestalling forbearance as a function of NCA’s “home bias”, is endangered, because the ECB will simply lack the resources to generally investigate daily supervisory practices of NCAs” (Tröger, 2014: 27). To summarize, the ECB emphasizes the formal cooperation too much and overlooks the professionalism of the local supervisor, which seems to be crucial for the effectiveness of supervision. The home bias of NCA supervisors is generally seen as a flaw of the previous decentralized system of supervision but could also be a key feature of effective supervision. The perception of Tröger explains why the principled-based approach is advantageous but it does not deal with the most criticized part of the approach, the cognitive capture. This remains a flaw in the approach despite the advantages elaborated above.
2.3 Conclusion
The SSM has, as stated above, is an outcome of a political struggle among various actors. The question remains if the outcome would be appropriate for the task of banking supervision. The numerous advantages of centralization and uniformity such as harmonization of regulation and breaking down the capture and home biases simultaneously include some challenges. The harmonization of the supervision regulation creates a pooling of risks in a systemic way of all watching the same risk but miss potential others. This could, however, be overcome through an emphasis of diversity in the internal organization of banks.
Furthermore, breaking down the home bias of NCA could have a negative effect since it is also one of the strengths of NCAs supervision process. The initiative of the local supervisor and the close relationship with the bank could be considered a valuable resource. Whether these potential pitfalls are truly a problem in the system is explored in the interviews.
24
3. Methodology
In this chapter, the translation of theory to practice is elaborated. The empirical part of the thesis takes on the governance structure of the SSM and relates it to the degree of centralization uniformity. The research strategy, research design, and research methods are discussed. The methodology is designed to the answering of the two core research questions:
“What degree of centralization and uniformity can be identified within the current governance structure of the Single Supervisory Mechanism?”
“Is the balance in the case of the cooperation between the European Central Bank and Dutch Central Bank appropriate for banking supervision?”
In order to formulate the research questions, the right strategy and starting point are stated first. Second, the research design which gives the research body is explained and last the methods to gain information are discussed.
3.1 Research Strategy
The research strategy for this study is qualitative inductive research. The relationship between theory and research is inductive because with the data gained through research the theory is broadened. The theory is the guidance for the research direction and, therefore, it is an iterative research process. The theory is used to focus the field of research and is then enhanced with the findings (Bryman, 2012: 26-27). The study is qualitative because it starts from an interpretative point of view. The study of the governance of the SSM is about structures and how people fill in these structures. The behavior of people within the mechanism determines if the cooperation between the ECB and NCAs will work. The complexity of the governance framework also contributes to the choice for qualitative research. With qualitative research, an in-depth analysis can be made of how cooperation is experienced among actors at different levels within the mechanism. Consequently, the ontological position of this research is a constructivist position that asserts “that social phenomena and their meanings are continually being accomplished by social actors. It implies that social phenomena and categories are not only produced through social interaction but that they are in a constant state of revision” (Bryman, 2012: 33). The formal properties of organizations such as rules, regulations, and roles tend to miss the social element of the organization. Social interaction creates informal hierarchies and relationships that are not documented in the regulation or formal rules.
3.1.1 Research design
The research design chosen for this study is a case study. Case studies as a research design enable an intensive study of a single unit for the purpose of understanding a larger or (similar) units (Gerring, 2004). Real-life situations can be investigated in depth, to find out about the relations and processes within the natural context (Denscombe 2010, p.55). In the case study, the focus lies on the individual case which has implications in a wider context and for other phenomena as well. The selection of case
25 studies should be carried out “on the basis of their relevance to the practical problems or theoretical issues being researched” (Denscombe 2010, p.57). They can be supported or specified by known attributes and distinct features found in theory and literature (Denscombe 2010, p.58). The selection of the cases is an important part of the method section where the motivations and justifications for the specific choice of a case study need to be made explicit in order for the results to be valid. This study is focused on the cooperation between the ECB and the DNB within the SSM framework. Therefore, it is not just the case of the SSM but within the mechanism, there is a decision to emphasize the cooperation with one NCA. In the following section, the case selection is further explained.
3.1.2 Case selection
The choice for focusing on one cooperation between the ECB and an NCA is based on a time and resources analyses. The committed timeframe for the thesis project is strictly four months, where all research analysis must be completed. This required choices to be made in the examination of the SSM and whether to do in-depth research on a small part of the mechanism which can be used as a representative case or to conduct a more generalized study of the entire SSM with less eye to detail. This study aims at the former, deciding to conduct an in-depth analysis of a part of the SSM to order to get the most complete projection of at least a part of the SSM.
3.1.3 ECB-DNB relationship within SSM
The case of the DNB is chosen because it represents an exemplifying case. Therefore, it is a representative or typical case study, as defined by Yin (2009). The objective of an exemplifying case study is to capture the circumstances and conditions of an everyday commonplace situation (Yin, 2009: 48). An exemplifying case study is not an extreme or unusual case, but it rather represents a broader set of cases in the same category. With this type of case study, the transferability is higher with a solid and thick description of the case described (Bryman, 2012: 70). The DNB is such a case because the internal processes are much like the processes of the ECB. The methods for small and medium-sized banks are also almost the same as the methods used for the large significant institutions under the supervision of the ECB (Algemene Rekenkamer, 2017). The capital and liquidity requirements are even higher than the minimal norms in CRR & CRD IV. Since most of the processes are similar, it is unlikely that anything related to that will pose an obstacle for the cooperation. The study can, therefore, focus purely on the cooperation in the basic functioning of the system. Moreover, the Netherlands has one institution which is responsible for both on-site and off-site supervision (Thiemann, 2018: 181). The DNB was before the SSM, according to Thiemann (2018), a national bank which was only operating in the interpretive community of banking regulation. The accounting section was something the DNB had no influence in. The ‘interpretive community’ is the regulatory network where actors engage in conversations to interpret the meaning of regulation and compliance (Ibid: 196-197). The DNB was, therefore, very well informed in the developments of the banking
26 sector but because of their disembedded position in the interpretive accounting community, their view was one-sided and easier to cognitive capture (Ibid.).
Figure 4.1 The regulatory networks governing in Germany, the Netherlands, and France (Thiemann, 2018: 197).
The Dutch regulatory and supervisory style was more principle-based than rule-based. This type of style was criticized in the crisis period and the SSM style of supervision is, therefore, more based on the French and to some extent on the German style of supervision (Schoenmaker & Veron, 2016). Consequently, the introduction of the SSM meant the coming of a different supervisory style which was more rule-based. In that sense, it is more an extreme case because the supervisory approach was rather different than the current SSM approach. In the following chapters, the extreme case characteristics become more pronounced. The DNB operates in some respects, such as the methods and processes, just like the ECB but in others like the macroprudential policy more deviant.
The banking landscape of the Netherlands is an interesting system. The banking system is concentrated and harbors three large systemic banks: ABN AMRO, ING Group, and Rabobank. ABN AMRO was formerly part of Fortis and is still 56,3 percent government-owned following the latter’s nationalization during the crisis (Website ABN AMRO). ING Group was divided during the crisis into a banking group which is still called ING and an insurance group which is now called Nationale Nederlanden. The Rabobank is a non-listed cooperative group (Schoenmaker & Veron, 2016). This means that the equity is not publicly traded on a major stock exchange. Along with the largest banks, there are three medium-sized banks which are also directly supervised by the ECB: de Volksbank, BNG and NWB. The banks directly supervised by the ECB cover 88 percent of the Dutch banking system’s total assets (Ibid.). The largest part of the banking sector in the Netherlands is supervised by the ECB, which means that the ECB and the DNB cooperate constantly in JSTs, information-sharing and on-site inspection missions. The close connection between the ECB and DNB makes this an exemplifying case. It is therefore valuable to see how both actors experience the cooperation and how
27 they face the new role the SSM has put them in. The units of analyses for this study are the ECB and the DNB, but also and more important the teams in which they work together, such as the JSTs, and on-site inspection teams.
3.2 Methods
In order to understand and explain the governance structure of the SSM and the degree of centralization in relation to banking supervision through the eyes of individuals within the system, qualitative methods are chosen to investigate this structure in practice. This will be done in three methodological steps where first theory will be an orientation for the basic understanding of the SSM. Second, primary sources of the European institutions serve as an instrument to formulate the institutional framework of the SSM. And finally, interviews provide the missing information and deeper knowledge of the functioning of the SSM. This triangulation of data sources improves the reliability and validity and is discussed later in this chapter.
3.2.1 Document and theoretical analyses
To be able to answer the research questions a document analysis is conducted. Official documents from the European institutions such as the European Parliament, the European Council, the EC, the ECB, and the NCAs serve as the first source for explaining the SSM. These documents include legal documents, discussion papers, statements from institutions, press releases and other publications. Interviews with high ECB officials conducted by quality papers such as the Financial Times are also valuable sources of information to compose a credible picture of the SSM. Besides these primary sources, secondary sources are complementary to the constitution of a description of the SSM. Secondary sources are used for the literature review in which existing research is critically examined in relation to the phenomena and interest of the relevant theoretical ideas (Bryman, 2012: 13-14).
3.2.2 Interviews
A method frequently applied to get insight into how people make sense of their everyday experiences are in-depth interviews. Such technique enables researchers to better understand and describe individual experiences from the interviewee's point of view. Interviews are a useful tool to directly address the participants and gain knowledge about their perception. For this study, elite interviews were chosen as a sampling strategy. Elite interviews were chosen because the interviewees are central in the research field. And since political science research frequently involves developments at the highest level of government, elite interviewees can contribute with relevant knowledge (Tansey, 2007). Elite interviews are commonly used as a confirming informational source. When documents and secondary sources provide an initial overview of the subject under examination, interviews with key actors in the field can corroborate the initial overview (Ibid.). Furthermore, interviews provide the opportunity to gain additional knowledge for the initial overview. Elite interviews can, therefore, confirm the accuracy of the information that has previously been collected from other sources or/and it can sharpen and improve that information. Similar information can be gathered through other method,
