Systemically relevant banks : determinants of bailouts : the case of Dutch bailouts between 2008 and 2009

SYSTEMICALLY

RELEVANT

BANKS:

DETERMINANTS

OF BAILOUTS

The case of Dutch bailouts between 2008 and 2009

Master thesis Political Science

Name: Floor ter Bogt Student number: 11058897 Date: 21-06-2019 MSc: Political Science Track: Political economy

Thesis group: The changing global economic order: rising powers and growing risks Supervisor: dhr.dr. J.G.W. (Jasper) Blom

Second reader: dhr.dr. F. (Farid) Boussaid Words: 18342

Table of content

3.1 Different theories on determinants of bank bailouts ... 12

3.2 Conclusion ... 21

4. Background ... 23

4.1 What happened in the Netherlands? ... 24

4.2 The Dutch banking system ... 28

4.3 Conclusion ... 30

5. Analysis of case study: the Dutch bailouts ... 30

5.1 The political organization (government-bank relations) prior to the crisis ... 31

5.2 The political organization (collective negotiation and burden-sharing) and bank lobbying ... 32

5.3 The threat of leaving the country ... 39

5.4 The political side of the bailouts: Political ideology and politicians’ political considerations ... 40

5.5 Economic development ... 43

5.6 The real determinant: system-bank ... 44

5.7 New propositions? ... 47

6. Conclusion ... 50

7. Bibliography ... 52

Appendix 1 ... 56

1. Introduction

In 2007, the financial crisis started in the United States with the burst of the housing market bubble. Citizens were no longer able to repay their loans and financial markets quickly got into trouble. This eventually led to a series of bank failures (e.g. Northern Rock in September 2007 and Bear Stearns in March 2008) and on September 15, 2008, it became a disaster when the government decided not to save Lehman Brothers. By the end of 2008, the crisis did not only affect the US but it had spread to Europe and Asia. It touched many countries but caused an immense crisis in countries such as Greece, Latvia, Spain, Ireland, Iceland, Italy, and Portugal. These countries went into recession and some even faced bankruptcy. European Union GDP fell by an estimated 4 percent in 2009, which marked the beginning of the first recession since the early 1990s (van Ewijk & Teulings 2009: 33-51). During the crisis, several national banks came into trouble and where no longer able to pay out loans or savings. They faced severe financial problems and were standing on the edge of bankruptcy. These banks had to be saved by national governments in order to survive. The emphasis of this paper lies on the bailouts of these banks. By the end of 2008, many large banks reached such a high level of insolvency that only massive government bailouts could keep these ‘zombie banks’ alive (Crotty 2009: 574). The financial crisis of 2008 was not only a global crisis, but it was also a crisis of extreme proportions (Orlowski 2012: 121). “The economic crisis of the 1930s was seen as peanuts compared to the global financial crisis of 2008” (Interview: Knot)1_.

Besides, it was of such proportions that no one had experienced a crisis like this, it was something entirely new2_{. Furthermore, the interdependencies in the financial system, between}

banks all over the world, became visible (Salden in Financieel Dagblad 2017). The financial crisis of 2008 still has an enormous impact on the world. The flaws in the global financial system, exposed by the crisis, need to be taken into account when making decisions nowadays.

Financial support, provided by the state, can prevent bank runs, improve the stability of the financial system and reduce uncertainty. Nevertheless, if bailouts are large and infer a challenge to the sustainability of public debt levels, they may essentially worsen the outlook for the economy and lead to extreme risks for the financial system (Fratzscher & Rieth 2018:

1 _{All interview quotes are quoted based on my own translation of what was said during the interviews.} 2 _{Confirmed by Commission de Wit 2011: 363.}

2). In the United States, bank bailouts during the financial crisis provoked long-running and bitter recrimination, for good reason, Tooze claims (Tooze 2018: 43). He argues that hundreds of billions of taxpayers’ funds had to pay to rescue avaricious banks. In the United States, these choices led to deep rifts within the Republican Party (Tooze 2018: 43). Bailout costs even exceeded 1 trillion dollars in the summer of 2009 (Grossman & Woll 2014: 579). In Europe the crisis caused panic, led to many bank bailouts and debt levels of some European member states exceeded 100% of GDP (Candelon & Palm 2010: 83). Extraordinary amounts of public funding were thus made available to commercial banks during the financial crisis of 2008, dwarfing the budgets of many other policy domains” (Grossman & Woll 2014: 574-575). The bailouts resulted in political discussions, discontent, and anger among the

population (Tooze 2014: 376). It is clear that designing bank rescue packages and making the decision to bailout (or not) is thus an extremely hard task.

Even though every OECD country was hit by the global financial crisis, governments

responded in various ways. Governments were afraid that individual bank failures turned into general financial crises and responded in various ways. They provided liquidity support to banks, recapitalized them, issued state guarantees to reassure depositors and provided mechanisms to relieve financial institutions of impaired or ‘toxic’ assets. While some countries undertook all of these measures, others only undertook some, or none, of them (Grossman & Woll 2014: 579). The emphasis of this paper lies on the responses of the banking crises emerging in 2008. The Netherlands initially responded with high capital injections, while the United Kingdom mainly responded by providing guarantees. On the other hand, countries like Germany and France exerted very limited liquidity measures and responded in other ways that only took up small percentages of their GDP. So, how is it possible that governments responded in different ways whilst being hit by an equivalent crisis? It is very interesting to see why and to what extent a government decides to financially support a bank during the financial crisis.

This thesis aims to analyze the determinants of governments’ decisions to bailout banks. It contributes to the debate on the main determinant of bailing out banks by national

governments in OECD countries. This is done by doing a case study of the determinants of the measures taken by the Dutch government between September 2008 and January 2009 to save the financial system and restore trust in the system. The stability of the financial system

and the protection of consumers with savings led to an approach including, among others, the recapitalization of vulnerable systemically relevant financial institutions (Commission de Wit 2011: 84). There are many different theories on the determinants of bank bailouts that often contradict each other but at least propose different arguments for governments to perform bailouts. It is crucial to IPE research to find out which factor truly determines the choice of executing bailouts.

This brings us to the research question: What is the main determinant of governments’

decision to bailout banks in OECD countries? The research question follows in several

sub-questions:

o What is the structure of the Dutch banking sector?

o Which actors played a role during the decision-making process? o What happened during the Dutch bailouts?

To conclude, in depth analysis of the Dutch bailouts between 2008 and 2009 demonstrates that whenever a bank is systemically relevant to the financial system (‘too-big-to-fail’

principle) it must be saved. Therefore, the too-big-to-fail principle/systemically-relevant-bank is the main determinant of bank bailouts by governments. Furthermore, the fear of contagion also highly influences the choice to perform a bailout.

2. Methodology

2.1 Introduction

In order to answer the research question: “What is the main determinant of governments’

decision to bailout banks in OECD countries?”, I have chosen the Netherlands as case study.

The literature emphasizes different possible causes that have determined governments to bailout banks. Therefore, I will assess the different propositions in the case of bailouts in the Netherlands (N=1), to contribute to this debate. It is important to note that each European country is allowed to decide on its own rescue package design. In some way it was

bailouts nor the design of the rescue packages were coordinated or put into a certain framework that had to be implemented.

In this paper I will make use of process tracing. To clarify, I want to define what the determinant of implementing a bank bailout is (X) in relation to the specific bank bailout (Y) (Gerring 2001: 163-179). My most important data consists of 1. the parliamentary hearing performed by the Commission de Wit in 2011 on the financial crisis in the Netherlands, and 2. interviews with private and public policymakers and banking officials. The parliamentary hearing that I analyzed is available online and made noticeable/available in the bibliography of my thesis as well as the statistics I used (banking concentration, capital injection per institution, revenue and costs of supporting financial institutions and types of government intervention, political ideology). Moreover, except for several anonymous interviews, the interviews I have done are transcribed and recorded.

The parliamentary hearing is systematically analyzed and gives a clear picture of what occurred during the Dutch banking crisis. All interviewees have been asked similar questions (except for some slight adjustments with regard to the different interviewees). By analyzing the process of Dutch bailouts in 2008 and 2009 I have reached different conclusions

respecting the propositions tested.

2.2 Case selection

Case studies have made a central contribution to IPE, especially in rare cases like the financial crisis. My case study is a research based on the in-depth empirical investigation of bailouts in the Netherlands in 2008 and 2009. With the results of this case study, I aim to elucidate features of a larger class of phenomena, namely bailouts in OECD countries, by testing different theoretical propositions (Vennesson 2008: 223-226). Consequently, this research considers the bailout of banks in democratic countries.3

The Netherlands is used as a case study and the emphasis of this thesis lies on the crisis measures taken by the Dutch government between September 2008 and January 2009 (with some exceptions) to save the financial system and restore trust in the system. The Netherlands

represents a good example of an OECD country since it is a democracy, it is well developed and industrialized. Every factor that I want to test with my propositions is present in the Netherlands. Obviously, there are multiple banks present in the Netherlands, it has a strong government and it is seen as a wealthy state. Moreover, banks like ABN AMRO and ING were internationally connected and parts of their banks existed in foreign countries. In 2007 ING, with a balance of more than 1300 billion euro’s, was number 15 on the list of biggest banks of the world. Furthermore, ABN AMRO desired to become one of the five biggest investment banks in the world. The Netherlands, even though it is considered to be a small state, was seen as a significant player in the financial market and it has an important financial system (Teulings et al. 2011: 94). Furthermore, the Dutch state performed great bailouts, in which multiple banks were financially supported. Nevertheless, there are some particular factors present in the Netherlands.

First, the Netherlands has been one of the first and few countries that performed capital injections in 2008, which makes it extra interesting since they have performed these injections at the very first stages of the financial crisis. However, this does not have a significant effect on the results of this paper since I have not compared Dutch bailouts to for example a state that was very late with performing these capital injections. The emphasis lies on 2008 and 2009, representing the time in which nearly every state has performed bailouts.

Second, the Netherlands has a multi-party system, meaning that there are not only multiple political parties but, as a consequence, a coalition government will in most cases be formed. The fact that there are so many parties in the Netherlands does influence the results of this paper. It is very unlikely that politicians’ political considerations determine the decision to perform bailouts since there is not ‘one’ politician that makes this decision. Multiple parties and politicians felt the necessity to work together and to form a coalition.

Third, the Netherlands has a large and concentrated banking sector. The concentration of an economy is measured by measuring the market share of the five biggest banks. Before the crisis (and still), the five biggest banks (ING, Rabobank, ABN AMRO, Fortis and SNS Reaal) together owned more than 80% of total assets. The other 20% is divided over many other and much smaller financial institutions (Banken.nl). The concentration of the Dutch financial system might make it more likely that the too-big-to-fail logic occurs but in order to find out whether this logic is supported, I will analyze the recent bailouts and find out what was the main determinant. In contribution to this, there are several other countries that have a

high concentration of the banking sector (Belgium, Denmark, Portugal, Finland, Greece, and Sweden) (Boonstra & Groeneveld 2006: 25).

Operationalizations

In this paper, I conceptualize a bailout as either a capital injection, recapitalization or nationalization of a bank. I thus not specifically look at banks that made use of guarantees provided by the government. This paper concentrates particularly on the following financial institutions: Fortis, ABN AMRO, ING, SNS Reaal, and DSB Bank. The first four banks all have significant market shares and are crucial to the Dutch economic system. However, Rabobank was also relevant to the financial system but was not a publicly traded bank and eventually did not receive a bailout, therefore the role of Rabobank is not relevant for this thesis. Furthermore, Fortis, ABN AMRO, ING, and SNS Reaal have each received financial support from the government. Finally, I have decided to include DSB Bank, a smaller and less crucial bank that eventually deliberately was not saved by the Dutch government. Including DSB Bank enables me to compare the bigger and more important banks to a small one. Furthermore, DSB Bank went bankrupt while the others ultimately survived. Second, I take the bailout of AEGON (an insurance company) into account since they received a significant capital injection and is crucial to testing my propositions at the same time.

To operationalize how I will be measuring the different propositions I argue that most of my operationalizations are based on what the Commission de Wit and the interviewees conclude. However, in the case of economic development, the size of banks, the political organization and the political ideology of the state I also include some independent sources (GDP per capita, concentration of the banking sector, ProDemos and profits on loans). Thus, in several cases I base my conclusions on multiple measurements to strengthen the argument.

2.3 Data

This thesis is based on two main sources: the parliamentary hearing and interviews.

Parliamentary hearing

The Dutch parliamentary hearing is used as one of the most important sources of this paper and provides me with all the ins and outs of the process of bailing out banks in 2008 and

2009. In 2011 the final report of the parliament was published. The parliamentary hearing was established to determine what precisely happened during the crisis. To be more precise, their research assignment was to come to an understanding and judgment of the crisis measures taken by the Dutch government between September 2008 and January 2008 and the public financial support given in that time, to gain more information for the future (Commission de Wit 2011: 68). The hearing is set up by six politicians from different political parties, with J. de Wit (SP) as chairman. It is also known as Commission de Wit4_{. Their report consists of}

709 pages in which every step of the Dutch financial bailouts is discussed. It starts with the months following the fall of Lehman brothers, up to the role of the parliament after 2009. The parliamentary hearing describes in detail how the process worked, who was involved and who called the shots. It is therefore crucial for my research to determine which factors played a role in de choice to perform a bailout. The parliamentary hearing is a crucial part of my thesis and is used to determine if the different propositions were applicable in the Netherlands. The hearing is seen as trustworthy; therefore, I assume that every important step and factor is taken into account. Still, it is important to notify that the parliamentary hearing has not done my research yet. Their emphasis lies on what happened, but especially, if the decisions made at that time, were justified. It does not examine which factors determined the decision to perform a bailout. Moreover, after analyzing the parliamentary hearing I conclude that the interviews I have performed have given me additional information on the determinants of bailouts. For instance, the parliamentary hearing does not give any information on the factors of political ideology and bank threats of leaving the state.

Interviews

Interviewing is a crucial part of my thesis. Since there were only a small number of people involved during the decision-making on bank bailouts, it is crucial to speak to these persons. Nearly every conversation, negotiation and decision was made in private to avoid the fact that people would lose their trust in the market and that a bank run would happen as a

consequence. Therefore, speaking to the persons who were involved gives me essential insight into what happened at that time, concentrating on the dynamics of the policymaking process behind closed doors and the considerations of the decisionmakers.

I have employed interviews with financial market participants, members of the Dutch Central Bank (DNB), top officials of the Dutch Ministry of Finance, a former member of parliament and a representative of the NVB, the Dutch Association for Banks. I have decided to

interview decisionmakers that came out of the Commission de Wit report. This list was supplemented with suggestions from other interviewee’s (snowballing effect). To have a broad spectrum, I have interviewed both private and public actors. See Appendix for more information on the interviewees and interview questions.

I contacted politicians, policymakers and bankers that were in 2008 and 2009 highly involved in the decision-making process surrounding the Dutch bailouts of Fortis, ING, ABN AMRO, SNS Reaal and DNB Bank. I contacted potential interview subjects with an interview information sheet, including a project description and arranged interviews with those willing to participate in the study. In some cases, I have mentioned the name of the person who helped me to get in contact with them. To be more precise, I have spoken to the current president of the Dutch Central Bank and former director financial markets at the Ministry of Finance (Klaas Knot), former director financial markets (Bernard ter Haar), a former member of parliament (anonymous), a representative of the Dutch Association for Banks

(anonymous), the former agent of the Treasurer-General (Erik Wilders), two other

policymakers of the Ministry of Finance (Rens Bröcheler and an anonymous interviewee) and a former director of banking supervision at the Dutch Central Bank (Rudi Kleijwegt). More than 80% of the contacted interviewees agreed to do an interview with me, the other 20% did not respond. I have conducted 7 interviews in May 2019, each lasting from 30 to 45 minutes.

The interviews consisted of a set of open-ended questions regarding the decision-making process of bailouts in 2008 and 2009. The decision for open-ended questions is based on the desire for new insights and enables me to go deeper into the subject. Questions

centered on the course of the crisis, the role of banks, negotiations, politicians’ political considerations, the color of government, bank lobbying, the international position of banks and the size of banks. I have used a similar set of questions for each interview but slightly adjusted them in some cases to the interviewee. The interviews provided crucial insight into the policymaking process. Furthermore, the answers that were given by the interviewees were in line with each other.

Ultimately, I have used some statistics, I made use of capital injections per institutions (Rijksoverheid), revenue and costs of supporting financial institutions (Algemene

Rekenkamer), types of government intervention (parliamentary hearing). Finally, economic development is measured based on GDP per capita of OECD countries (OECD.nl).

2.4 Process Tracing

This type of methodology/mechanism fits in the best possible way to my research since I will be looking at the determinants of bailouts, therefore it is important to look at what precisely happened before the government decided to financially support a financial institution.

Therefore, I use the parliamentary hearing of the Dutch government, that has been made after the financial crisis, to trace the process of bailouts. By using process tracing as a method, I can find out which elements determine the decision for a bailout.

Even though process-tracing is often seen as one singular method, according to Beach and Pedersen it has 3 variants: theory-testing process-tracing, theory-building process-tracing and explaining outcome process-tracing. For my thesis, I concentrate on a combination of the former and the latter. The main goal of process-tracing is to explain a particularly interesting and puzzling outcome, so the why/how/when, etc. Explaining outcome process-tracing is defined as seeking the causes of a specific outcome in a single case (Beach & Pedersen 2013: 18-19). In my case, I look at the factors that determined the Dutch state to perform bank bailouts. However, I also use theory-testing process tracing since I test several existing propositions and trace how the process in the Netherlands worked. In theory-testing process tracing, different propositions are tested. The goal is to determine whether evidence (in my case policy documents and interviews) verifies that the hypothesized causal mechanisms linking X and Y were present and that it functioned as theorized (Beach & Pedersen 2013: 11-12).

2.5 Conclusion

To summarize, to define what the main determinant of bank bailouts is I analyze the Dutch bailouts between 2008 and 2009. I will do so by making use of the parliamentary hearing conducted by Commission de Wit in 2011. Additionally, I have performed interviews with policymakers, central bankers, a former member of parliament and a representative of the

Dutch Association of Banks (NVB) to get a deeper understanding of the policy-making process of bailouts.

It is important to note that this thesis is based on the parliamentary hearing, completed by Commission de Wit, and interview statements. Measurements are thus mainly based on their findings instead of independent sources. Their perception and conclusions of the financial crisis and the bailouts that arose out of it are considered to be leading. This implies that the results of this paper are in fact (almost entirely) based on what the Commission de Wit and the interviewees perceive as the main determinants for bailouts. The Commission and the interviewees therefore, determine whether the propositions are supported or not in the case of Dutch bailouts.

In the next chapter, I will discuss the literature on the determinants of bailouts. This follows in multiple propositions that will be tested in chapter 5. Chapter 4 consists of a

background on the Dutch financial crisis, the bailouts that were performed by the government, a description of the most important actors and an overview of the structure of the Dutch banking system.

3. Theoretical framework

This chapter discusses the existing literature on the determinants of bank bailouts. Many scholars have been interested in researching bank bailouts. Most previous research has been done on the effect of bailouts (e.g. Faeh et al. 2009, Bordo, M.D. and Schwartz, A.J. 2000, Tooze 2018, Sapienza 2004, Brown and Dinc 2005) or for example the moral hazard problem that coincides with bailouts (e.g. Lyons 2009, Dam & Koetter 2012, Antzoulatos & Tsoumas 2014, Ashraf 2017). But more importantly, research on the determinants of bailouts has similarly been rather large. It has often been a critical and sometimes even sensitive subject amongst officials and administrators that worked and had to perform these bailouts during that time. Therefore, it is very difficult to determine what it is precisely, that determines these bailout packages. For starters, it is difficult to know how much and what kind of emergency aid is necessary for a given situation (Grossman & Woll 2014: 575). It is thus not only

extremely hard to determine the strength of the capital injection, but also what then eventually determines the act to do so for governments. Grossman & Wol (2014), Cukierman (2013),

Fernandes et al. (2016), Rosas (2006), Bian et al. (2016) and Blau et al. 2013 have studied the causes of why governments decide to bailout banks. This chapter analyzes their contributions.

In this chapter, I discuss the theoretical debate concerning the determinants of bank bailouts. A variety of (sometimes contradictory) determinants are proposed. For example, some argue that lobbying activities determine bailouts (Blau et al. 2013), while others claim that lobbying power is irrelevant (Culpepper & Reinke 2014, Woll & Grossman 2014). Therefore, the debate develops several propositions that will be tested in a case of bailout schemes in the Netherlands. In this way, I aim to determine which one of these theories is, in fact, the main determinant of bank bailouts.

3.1 Different theories on determinants of bank bailouts

In this paragraph, I set out the theoretical debate and discuss the literature concerning the determinants of bank bailouts. Each sub-section follows into a proposition that is tested with the case of the Dutch bailout packages. After each proposition, I included a short

operationalization in which I specifically describe how I perceive these propositions when examining them.

3.1.1. Bank-government relations

Grossman and Woll have researched how much leeway governments had in designing bank bailout schemes. Grossman and Woll claim that bank-government relations is an important determinant of bailing out banks. To elaborate, their main finding is that the strategies used by governments to cope with the instability of financial markets do not depend on economic conditions alone (see section 3.2.5 below). In contribution to this, they are dependent on the institutional and political setting of each country. The variation of bailouts is caused by the different types of business-government relations banks were able to entertain with public decision makers (Grossman & Woll 2014: 574). Through their four case studies (France, UK, Ireland and Sweden) they demonstrate that “countries with close one-on-one relationships between policymakers and bank management tended to develop unbalanced bailout packages, while countries in which banks negotiated collectively developed solutions with a greater

burden-sharing from private institutions” (Grossman & Woll 2014: 574). They thus give great value to the connection between bankers and public officials (whether the focus lies on

corruption, lobbying or banking systems) and state that the tighter their relationship, the more likely bailouts occur. This argument is supported by others who state that politically

connected institutions are more likely than others to receive financial support (Tahoun and van Lent 2010, Duchin and Sosyura 2012, Blau et al. 2013). Furthermore, the way in which the banking system is build up highly influences the strategies used by national governments. Eventually, the relationship between the banking sector and the government do matter, but neither capitalism nor lobbying per se capture this variation. What thus matters are the political organization of the banking sector and the collective banking action. The most successful bailouts consisted of the participation of the banking industry in finding the best policy solutions (Grossman & Woll 2014: 579-594). To conclude, the political organization of a state determines the likeliness of bailouts.

Proposition 1

The political organization of a country determines the likeliness of bailouts.

Operationalization

The political organization of a country is measured in two distinct ways. First, the relationship between bankers and policymakers is operationalized in the following way. I consider it a strong relationship when interviewees and Commission de Wit conclude so. If interviewees and Commission de Wit argue that there was not a good relationship, I consider it to be a weak relationship.

Second, I inquire whether bank representatives and governments worked closely together to reach a common solution. I measure this in the following way: if bailouts are negotiated individually with banks (and not with multiple banks at the same time), I consider it to be a collective solution (an arrangement made by the bank and the government) and therefore bailouts are more likely to occur (Grossman & Woll 2014: 576-593). I base this on the conclusions made by Commission de Wit considering the negotiations and what the interviewees argued when asking them whether banks participated in negotiations. Moreover, if banks participated in burden-sharing, thus if there were also negative consequences for the private sector (i.e. high interest rates on the loan, fine premiums, cut of bonuses, the board had to leave, etc.) I argue that bailouts are more likely to occur. Furthermore, I examine what

banks had to pay for the bailouts they received. If interviewees and Commission de Wit conclude that the relationship between bankers and policymakers and collective negotiations and burden sharing determine the decision to bailout a bank, it supports this proposition.

3.1.2. Strategic structural power: “too-big-to-fail” and “the threat of leaving

the country”

According to Culpepper and Reinke (2014), the main determinant of implementing bank bailouts is the use of strategic structural power by national banks. Banks’ structural power is their deliberate use of their role in the economy as a resource in bargaining with the

government (Culpepper and Reinke 2014: 447). Research verified that strategic use of structural power was the main determinant of variation in the form of bailouts chosen.

Culpepper and Reinke even state that using strategic structural power as a bargaining resource is necessary if a bank wants to deter a policy that is implemented by the government

(Culpepper & Reinke 2014: 432). To clarify, I differentiate the two most important factors of Culpepper and Reinke’s structural power explanation: the too-big-to-fail principle (the size of a bank) and the threat to leave the country.

1. Too-big-to-fail

First, the role of the bank in these negotiations coincides with the ‘too-big-to-fail’ idea. To be more precise, in some situations financial institutions are so large and/or so strongly

interconnected to the rest of the economy that their failure would result into a disastrous domino effect for the entire economy (Stern and Feldman 2004 & 2009, Morgan and Stiroh 2005, Mishkin 2006, Stern 2009). Therefore, if the size of the banks becomes larger, the need for government intervention will become more intense (Laeven & Valencia 2010). Culpepper and Reinke’s research demonstrates that “large banks are central to the functioning of

financial systems, and when their failure risks bringing down the entire financial edifice, the structural position of these banks makes a bailout the most likely outcome” (Culpepper & Reinke 2014: 446-447). The hypothesis of the too-big-to-fail principle thus consists of the idea that the larger the bank, the higher the probability of public intervention in case of default. Additionally, it has been argued that the too-big-to-fail issue may weaken market discipline since it pressures the governments to take action (Antzoulatos and Tsoumas 2014:

170). Their research emphasizes the size and power of banks, which can make them too-big-to-fail. Fernandes et al. support this argument and state that large banks have a higher chance of receiving government support when faced with financial distress (Fernandes et al. 2016: 724).

Besides, they argue that the size of the bank inherent to the concentration of the banking sector. Fernandes et al. argue that the higher the concentration of the banking sector, the higher the chance of the government to perform a bailout (Fernandes et al. 2016: 742). It is argued that concentration is a factor that can give an indication of the economic importance of the banking sector. The concentration of the banking sector also raises bank fragility and so, bank distress (Boyd & De Nicolo 2005). Most importantly, policymakers are more

concerned about bank failures and are more likely to bailouts banks, when the banking sector consists of a limited number of great banks (Fernandes et al. 2016: 725). This argument is supported by Culpepper and Reinke, who argue that countries like the United Kingdom, with a higher concentration of the banking system and thus a smaller number of banks, might be able to coordinate more easily to resist state pressure than a large number of smaller banks, like in the United States (Culpepper & Reinke 2014: 443).

Proposition 2

The size of banks coincides with the likeliness of bank bailouts.

Operationalization

The concentration of the banking sector is calculated as the fraction of assets owned by the five largest banks in the country (European Central Bank 2017). I consider banks too-big-to-fail if the concentration of the banking sector is higher than 60% (Boonstra & Groeneveld 2006: 25). If we look at the formal definition of too-big-to-fail it states “TBTF refers, in short, to financial institutions of such a scale that politicians and regulators are broadly agreed that their failure would cause deep damage to the economy more broadly” (Christophers 2014: 430). Besides, if the Commission de Wit argues that letting a bank go bankrupt would have disastrous consequences for the financial system it is also seen as too-big-to-fail. Similarly, such statements of the interviewees on these banks are simultaneously considered as too-big-to-fail. Moreover, banks that are seen as system-banks by the Commission de Wit and the interviewees, are also seen as too-big-to-fail. If interviewees and Commission de Wit

conclude that the high concentration of the banking sector and ‘too-big-to-fail’ banks determined their decision, it supports this proposition.

2. The threat of leaving the country

In contribution to the size of the bank, the international position of a bank plays a significant role in the structuralist view. To clarify, the power banks can use in negotiations with the government, is a function of the dependence of a bank on the domestic market. Structurally powerful banks are not entirely dependent on the national government since they earn a large share of their revenue abroad and thus have an alternative option. This outside option is not necessarily a threat to exit but makes them more powerful vis-à-vis the government since they will be able to ignore the threat of regulatory sanctions in one jurisdiction (Culpepper & Reinke 2014: 448). This will eventually make these banks more likely to engage in bailouts (Rosas 2006: 184). Moreover, the ability of politicians to carry out public policies, like a bailout, can be constrained by the possibility of capital flight (Rosas 2006: 179). Thus, governments designing their bank rescue schemes are influenced by this threat of leaving the country, used by these banks. The costs imposed by governments on structurally powerful banks will thus be lower and their bailout plans will be even more beneficial for these banks (Culpepper & Reinke 2014: 432). To conclude, threatening the state to leave the country, determines the decision of a state to perform a bailout. If the banks are internationally connected, they have more alternatives and using this as a threat increases the change of engaging in a bailout.

Proposition 3

The threat of leaving the country, exercised by banks, determines the choice for governments to bailout banks.

Operationalization

Initially, in order for banks to threaten to leave the country, they have to operate

internationally to have an alternative domicile in the first place. This information is gathered throughout the report of the Commission de Wit. Furthermore, I have operationalized ‘threat’ as arguments brought up by the banks to leave the country. This is measured by statements of interviewees. If they claim that banks brought up these threats, it supports the proposition.

Moreover, if these arguments are brought up in the report of the Commission de Wit they are also seen as banks exerting the threat to leave the state. If the banks indeed threatened to leave the country, I will decide if that influenced the choice for a bailout, based on the statements of interviewees. If interviewees admit and state that these threats determined their decision, it supports this proposition.

3.1.3. Political ideology

Another frequently discussed cause for bank bailouts is the ideology of governments. Cukierman argues that since bailouts are financed with taxpayers’ money, the decision to bailout a bank or not involves a normative consideration such as the resource distribution within a country and are therefore usually made by elected officials rather than by

bureaucrats. Therefore, the decision to provide a bailout depends on the political ideology and political consideration of the government. He argues that the more conservative a state, the less willing it is to spend a substantial amount on bailouts (or at least in the beginning)

(Cukierman 2013: 376-381). Moreover, research demonstrates that capital injections are about 18 percent less likely when the bank chairman is a member of the conservative party, which corresponds to the conservative ideology of limited state interventions (Bian et al. 2016: 16). Furthermore, according to literature in comparative political economy, we would expect countries with a liberal market tradition (conservatives/right-wing countries) to not use extensive government aid to save banks, while more interventionist countries (left-wing) should be more proactive in saving banks (Siaroff 1999). Cukierman’s argument coincides with other explanations that have been given by various other scholars (Moschella 2011, Tufte 1978, Schmidt 1996, Garrett 1998, Cukierman 2013). They call their determinant of bailouts the ‘Partisan explanation’. The partisan theory, or hypothesis of partisan influence, argues that a major determinant of variation in policy outputs and choices in constitutional democracies is the party structure of the government (Schmidt 1996: 155). Partisan theory studies have researched the role of different parties in determining what a government’s policy is based on. They have found that center-right and conservative governments tend to adopt fewer interventionist policies in the economy than left and social democratic

To conclude, one can primarily see two blocks emerging. On the one hand, the

conservatives/right-wing parties that have a liberal market tradition and who are restrained

when it comes to government intervention. And on the other hand, we have the

liberals/interventionists/social-democratic/left-wing parties who are more proactive when it

comes to intervention.

Proposition 4

The political ideology of a state determines the likeliness of bailouts.

Operationalization

I operationalize political parties in the Netherlands in the following way: when political parties are in favor of state interference for the purpose of achieving greater equality in society, they are considered as ‘left-wing’ (SP, GroenLinks, and PvdA). On the other hand, parties that consider inequality in individuals’ personal income and power as justified (to some extent) and inevitable, are categorized as ‘right-wing’ (VVD, SGP). At the same time, economically, GroenLinks and PvdA can be seen as progressive parties. While, D66, VVD, CDA, SGP and ChristenUnie can be considered to be more conservative. Some parties are not particularly right- or left-wing and belong to the political ‘center’ (CDA) (ProDemos 2013: 21-25). The political ideology of a state is thus based on ProDemos’ research. To

operationalize the role of political ideology as a determinant of bailouts I will examine the political composition of the parliament and which parties governed. Furthermore, I analyze if certain actions of policymakers have been in line with their political color. If Commission de Wit and interviewees argue that the political ideology has determined the decision to provide a bailout and if the actions of politicians were in line with their political party, it supports this proposition.

3.1.4 Economic explanation

Rosas argues that the economic development of a state matters for performing bailouts. He argues that richer countries are more likely to engage in bailouts. Richer governments have larger resources at their disposal to afford these bailouts meaning that if resources are bigger, a choice to perform a bailout will be made more easily (Rosas 2006: 176-188).

Proposition 5

Economic development determines the choice to perform bailouts.

Operationalization

Similar to Rosas (2006), I measure economic development by GDP per capita. Furthermore, if interviewees argue and Commission de Wit concludes that sufficient resources determine bailouts, it supports this proposition. Besides, I analyze the role of having sufficient resources during the negotiations by looking at necessary resources for a bailout and whether the state argues that they have (or do not have) sufficient capital for these bailouts.

3.1.5. The power of lobby

Literature on the determinants of bank bailouts suggests another factor: the role of lobbying. Blau, Brough, and Thomas (2013) argue that lobbying determines the likeliness of bailouts, the amount of money invested through a bailout and the timing of bailouts. Their research is based on the TARP (Troubled Asset Relief Program) support that was implemented in the US in 2008. Economists have persistently acknowledged lobbying firms receive a variety of economic benefits in return (Richter et al. 2009, Igan et al. 2009). Evidence shows that firms that lobbied had a 42% higher chance of receiving financial support than firms that did not lobby. Additionally, firms that received financial support, invested up to four times more money on lobbying than firms that did not receive financial support (Bleu et al. 2013: 1-11).

Proposition 6

Bank lobbying determines the likeliness of bailouts.

Operationalization

Since data on bank expenditure of lobbying is impossible to find I operationalize bank

lobbying in the following way. I rely on the interviews I have done with policymakers and ask them about the influence of lobbying. Furthermore, I analyze what Commission de Wit

concludes on the process prior to the bailout and the process of negotiations. If they argue that the banks exerted power to financially support them and searched for alternative options for financial support, I consider that there was indeed bank lobbying. On the other hand, if DNB or the Ministry of Finance needed to convince the bank to accept the bailout or demand

financial support, I consider that there was no actual bank lobbying. Moreover, if banks were ignorant about their own financial problems, I argue that there was no lobbying since they could not lobby for financial support if they were unaware of their financial issues. I consider bank lobbying as a determinant of bailouts if the interviewees and Commission de Wit argue that due to the lobbying and demands of the banks, they performed a bailout.

3.1.6 Politicians’ political considerations

Bian, Haselman, Kick and Vig (2016) concentrate on the electoral cycle and politicians’ political considerations in designing bank bailouts packages. They claim that politicians have a strong influence on the likeliness of bailouts. Similar to Grossman and Woll they believe that the relationship between politicians and banks is important. Bian et al. claim that local politicians have more knowledge about the local banks in distress, which can potentially improve the decision-making process, leading to better-designed bailout packages. Yet, bailout decisions made by politicians who are closely related to the banks in need, tend to be distorted by personal considerations (Bian et al. 2016: 1-23). Additionally, politicians may use their influence to pursue non-value-maximizing objectives to realize political objectives, such as re-elections (Sapienza 2004, Brown and Dinc 2005). Bian et al.’s research

demonstrates that interventions in the form of bank bailouts by politicians are about 30% less likely in a year before an election (Bian et al. 2016: 1-23). The authors argue that this is in line with the personal interest explanation. “Voters exert more discipline if the political

competition is more intense. Although a politician might want to prevent the restructuring of a distressed bank in order to keep it under her control, she cannot do so if this will be perceived as a waste of taxpayers’ money and hence be punished in the next election. The more intense the political competition, the more severe the threat of punishment” (Bian et al. 2016: 16). In short, personal incentives such as re-election determine the likeliness of a bailout.

Proposition 7

Politicians’ political considerations (namely re-elections) determine the likeliness of bailouts.

Operationalization

For this proposition I look at political considerations in the following way: 1. the desire to get re-elected and 2. whether politicians’ actions were in line with their political party. Similar to measuring the determinant of lobbying, the determinant of politicians’ political considerations is rather hard to measure. It is not likely that politicians would admit that they have decided to implement a bailout (or not) because they wanted to get re-elected. And if a politician does not admit this themselves, there is nearly no evidence that this was truly the case. Moreover, based on what interviewees and Commission de Wit conclude on whether the actions of politicians were in line with their political color, I argue that political considerations have played a role. Therefore, I operationalize political considerations in the following way: if interviewees and Commission de Wit state that politicians were influenced by personal considerations (such as re-election or based on the political party they represent), it supports this proposition.

3.2 Conclusion

To summarize, it is clear that there are different opinions on what precisely determines the choice of governments to perform a bank bailout or not. In this chapter, I have set out the theoretical debate and revealed which possible determinants are existent. However, as described before, these determinants cannot all be the main determinant of the choice for the government to perform a bailout. This debate has led us to come up with 7 different

explanations for governments to perform bailouts and thus 7 different propositions. These propositions are tested in chapter 5.

In table 1, a short overview of all propositions is demonstrated.

Table 1. Overview of propositions

Proposition Literature

1. The political organization of a country determines the likeliness of bailouts.

Grossman and Woll (2014), Tahoun and van Lent (2010), Duchin and Sosyura (2012), Blau et al. (2013)

2. The size of banks (too-big-to-fail) coincides with the likeliness of bank bailouts.

Culpepper and Reinke (2014), Stern and Feldman (2004) & (2009), Morgan and Stiroh (2005), Mishkin (2006), Stern (2009), Antzoulatos and Tsoumas (2014)

3. The threat of leaving the country, exercised by banks, determines the choice for governments to bailout banks.

Culpepper and Reinke (2014), Rosas (2006), Oatley (1999)

4. The political ideology of a state determines the likeliness of bailouts.

Cukierman (2013), Bian et al. (2016), Fernandes et al. (2016), Moschella (2011), Tufte (1978), Schmidt (1996), Garrett (1998)

5. Economic development determines the choice to

perform bailouts.

Rosas (2006)

6. Bank lobbying determines the likeliness of bailouts.

Blau et al. (2013), Richter et al. (2009), Igan et al. (2009)

7. Politicians’ political considerations (namely re-elections) determine the likeliness of bailouts.

Bian et al. (2016), Sapienza (2004), Brown and Dinc (2005)

In the next chapter the background of the Dutch bailouts is discussed. It also provides a clear overview of the Dutch banking and governance structure and discusses the most important actors concerning the financial crisis and bailouts that followed.

4. Background

First of all, since bailout programs in the Netherlands are used as case study for the

determinants of bank bailouts, it is important to have some background information on the Dutch banking system. This chapter enables the reader to get a grasp of what broadly happened during the crisis and which banks were bailed out. This chapter demonstrates that the crisis exceeded the established borders rapidly. Besides, it becomes clear that the Dutch state was unprepared for a crisis of such proportions. Banks did not expect this and their financial situation deteriorated rapidly. Eventually nearly every Dutch bank needed to be bailed out. I will first discuss the crisis broadly.

On July 18, 2007, Bear Stearns, an American bank acknowledged that they had faced severe financial trouble, which resulted in stress on the American market. Only a month later the European Central Bank (ECB) decided to provide unlimited loans to European banks. Before that, European banks used to loan money to each other, however, that system had collapsed and was no longer working in an effective way. Therefore, the ECB took over.

On September 15, 2008, Lehman Brothers fell. This point in history is often seen as the factual start of the financial crisis. The bank faced a debt of $613 billion and other banks were no longer willing to provide Lehman with loans, trust in the bank had vanished. In the fall of 2008, the US came up with a rescue scheme, or in other words, bailouts, to help the financial sector. They made $750 billion available to help banks in financial distress, either in the way of a capital injection, or by buying up ‘bad’ loans (taking over toxic assets).

In October 2008, European banks received more money to keep the economy up and running. In 2010, countries like Greece, Portugal, and Ireland experienced severe crises and the credit crisis turned into a debt crisis. From 2012 onwards the economy indicated signs of recovery. Nevertheless, the moment and level of recovery varied across countries. Countries like Greece and Italy are nowadays still struggling with the consequences of the financial crisis.

4.1 What happened in the Netherlands?

Timeline Dutch financial crisis

Since American banks had sold worldwide mortgages (subprime mortgages), the crisis that emerged in the US also spread to the rest of the world, including the Netherlands. In 2007 Fortis took over several parts of ABN AMRO (Private banking, asset management, and Business Unit Nederland). However, the required capital was nearly impossible to find. Even though their financial solvability was quite strong in the beginning, after the takeover of ABN AMRO their financial position worsened drastically. In 2008 Fortis tried to sell ABN AMRO again but no buyer was found. To protect financial institutions, the Dutch government decided to make €20 billion available to support financial institutions. Each bank and insurance

company that was assumed to be healthy and viable could apply for financial support. Gita Salden, former head of financial stability at the Ministry of Finance, claims that in September 2008 the financial crisis began in the Netherlands when the government financially supported Fortis (Salden in Financieel Dagblad 2017). On September 28, 2008, the Dutch, Belgium and Luxembourg governments invested €11,2 billion euros to support Fortis.

Nevertheless, on October 3, 2008, the Dutch government revealed that all Dutch parts of Fortis group (including ABN AMRO) were bought by the Dutch government. Meaning that the government now also owned Fortis Verzekeringen Nederland (currently ASR) and Fortis Corporate Insurance. Thus, Fortis/ABN AMRO was separated and nationalized

(Rijksoverheid). Four days later, the Minister of Finance revealed that the guarantee on deposits of 1 year would be increased to €100.000. The Icesave-crisis (an Icelandic bank that was unable to repay consumers savings, including Dutch savings) led to severe distrust amongst consumers. In October the Dutch government, together with DNB, promised that consumers that were holding savings at Icesave would get their savings back up to an amount of €100.000. Moreover, in October and November, the Dutch government provided ING, AEGON, and SNS Reaal with capital injections (Table 2).

Table 2: Capital injection per institution with repayments and returns in euro’s

Institution

Capital injection Repayment

Return on

repayment

AEGON

3 billion

3 billion

18,5%

ING

10 billion

10 billion

12,7%

SNS Reaal

750 billion

185 billion

8,5%

Source: Rijksoverheid: aanpak financiële sector

Without this support, these banks would have probably faced bankruptcy comparable to Lehman Brothers. Nevertheless, the Dutch state deliberately decided not to bailout the DSB Bank. The bankruptcy of DSB and the Icesave-crisis still led to severe distrust amongst consumers.

Primarily, the government had no desire in taking over parts of ING or owning parts of any other bank in fact5_{. However, the €10 billion capital injection in ING did not have the desired}

effect. It turned out that recapitalization of ING did not improve trust in the financial market and their solvability continued to be highly risky. On January 26, 2019, the government took over the Alt-A mortgage loans of ING and provided them with a back-up facility. In the second half of 2009, the economic shocks slowly stabilized. Even though the financial crisis was not over, the Dutch Ministry of Finance and DNB felt that the biggest crisis days had passed. The capital injections provided to ING, AEGON, and SNS Reaal had rescued these banks (Rijksoverheid).

From 2011 on the Dutch government worked on economic recovery but this did not go easily. Unemployment increases, families face financial trouble, inflation increases and eventually the Dutch government/administration falls. On the 1st _{of February, 2013, the Dutch state}

becomes the owner of SNS Reaal. From 2014 onwards the economy then finally indicates some signs of economic recovery. Finally, in 2016 similar growth levels were reached as in 2008. Eventually, the Dutch government became the owner of ABN AMRO, SNS Reaal and insurance company ASR. The Dutch state financially supported ING twice. Liquidity

facilities were inducted which enabled financial institutions to borrow up to €200 billion (Financieel Dagblad 2017, Rijksoverheid: Aanpak financiële sector).

In graph 1, an overview of revenue and costs of supporting each financial institution during the crisis is demonstrated.

Graph 1

Source: Algemene Rekenkamer

Moreover, the Netherlands is argued to be one of the first (and strongest) countries of the EU to provide liquidity support to banks. As you can see in graph 2, the Netherlands not only already started with capital injection in 2008, they also provided the largest share of GDP to capital injections (liquidity) compared to other European countries. The Dutch state preferred

Back-up facility ING

Pre-financing deposit guarantee scheme

Guarantee facility

Capital injection facility

Fortis/ABN AMRO

SNS Reaal

Revenue Costs Negative/positive

to support the banking sector by implementing capital injections instead of actually buying up shares of banks or taking over entire banks. They wanted the Dutch financial system to be stable again, instead of owning majority shares in the banking sector. In the case of Fortis/ABN AMRO, and eventually SNS Reaal, this was impossible.

Graph 2: Types of government interventions in European banks: 2008-2013 (% of GDP)

4.2 The Dutch banking system

Important actors

There were three important actors during the Dutch bailouts of 2008/2009. The Ministry of Finance (and the parliament), the Dutch Central Bank (DNB) and of course, the commercial banks and institutions that were in need of rescue. Potentially, the Dutch Association for Banks (NVB) could have played a role. However, this role was rather small during the crisis.

Initially, the Dutch Ministry of Finance is responsible for the Dutch financial system, decisions, the budget, debt, and taxes. Bernard ter Haar, who was at that time director

financial markets of the Ministry of Finance, together with Wouter Bos (Minister of Finance), made the crucial decisions to financially support a bank or not. Nevertheless, in the end, Wouter Bos and Jan-Peter Balkenende (former prime minister) were responsible. After ter Haar, from 2009 till 2011, Knot was assigned as director financial markets. Thus, on the policy side the decisions were taken, but on the financial side bank ratings and the practical issues took place. The Ministry of Finance had the final say in whether or not to support banks since they were the only ones with the authority to decide what happened with tax money. Besides, the Minister of Finance (Wouter Bos) is the only one authorized to spend public resources and therefore the appointed crisis manager. Thus, when liquidity problems can no longer be solved by DNB or the financial institutions have severe solvability problems, the Ministry of Finance steps in (Commission de Wit 2011: 429). Moreover, it is important to note that the Ministry of Finance is an executive body and their decisions have to be approved by the parliament, which has the legislative/controlling power. Thus, whenever the Ministry of Finance spends public money it needs to justify these expenses to the parliament. In the next chapter, the outcome of this practice is explained more specifically. Finally, the Ministry of Finance is dependent on information provided by DNB and consequently assesses the financial position of banks with that information (Commission de Wit 2011: 49).

The second actor was the Dutch Central Bank (De Nederlandsche Bank, DNB) who played an advisory role. Dutch banks were and are in general supervised by DNB, who aims for financial stability and sustainable wealth in the Netherlands6_{. They verify if financial}

institutions are financially solid, so that in times of financial distress, they can still fulfill their

6 _{Since 2014 the ECB has control over large-sized banks and DNB only supervises small- and medium-sized} banks (DeNederlandscheBank.nl).

obligations (DeNederlandscheBank.nl). When financial institutions are in need of liquidity, they turn to DNB, however, when they have solvability problems, DNB refers them to the Ministry of Finance. While DNB supervised the banks and their financial status, they were thus also expected to communicate with the Dutch government when things got out of hand. Furthermore, the highest officials of DNB (among others Wellink, Knot and Kleijwegt) were also involved in the negotiations on bailouts. Still, DNB solely played an advisory role. The Minister of Finance was concerned about DNB’s opinion on the bailouts, whether it was a wise decision or not, but eventually the decision was made by Wouter Bos, whether DNB desired it or not was irrelevant (Bos in Commission de Wit 2011: 225).

Finally, the third important actor were the commercial banks. This thesis concentrates on ING, Fortis/ABN AMRO, SNS Reaal and DSB Bank. The banks were, as previously mentioned, regulated by DNB but in the end, nor DNB neither the Ministry of Finance could force them to accept a capital injection or any other kind of help. This meant that the banks needed to agree with a bailout. Besides, banks are private institutions and have their own responsibility. Efficient functioning, management, and governance is the responsibility of the bank itself, and not the responsibility of DNB nor the Ministry of Finance (Commission de Wit 2011: 420).

In principle, The Dutch Association for Banks (NVB) aims to represent the interests of the banking sector. They are a branch-organization and they represent banks collectively. They solely have an advisory role to the banks; they do not supervise them nor can they impose any rules. They assess the situation of the Dutch banking sector and through meetings with the CEOs of the banks, they discuss the financial stability. Potentially they could have organized a crisis management in which they collectively decide what to do and how to approach the Ministry of Finance for financial support. Nevertheless, this collective approach did not take place during the crisis. It was the first time the NVB experienced such a crisis, meaning that they were not prepared for any collective solution (Interview: Representative of NVB). Moreover, every bailout was negotiated and assessed individually, NVB was thus not one of the important actors during the decision making on bailouts.

To conclude, who made the decision to bailout a bank? While the Dutch Central Bank advised and informed the Ministry of Finance on the financial stability of the banks, the Ministry of Finance had the final say. Yet, a bank had to accept support, they could not be forced. Still, in

the next chapter it becomes visible that DNB and the Ministry of Finance did exert pressure on some of the banks to accept financial support.

4.3 Conclusion

This chapter described the Dutch financial crisis and bailouts that were performed. Furthermore, it clarified the way the banking system is structured and in what way the government in theory deals with banks in distress. The most important actors during the bailouts are also discussed. The next chapter demonstrates the policy-making process of the bailouts that took place. It discusses the negotiations, the decision-making, the role of the Ministry of Finance and DNB, and demonstrates which factors determined the performance of bailouts. Finally, the propositions that arose from the literature are tested.

5. Analysis of case study: the Dutch bailouts

When the Dutch banks first got into trouble, in 2008, DNB did not expect this. The promises the banks made led the bankers believe that they could cope with lower capital reserves, and thus lower buffers since their prospects of possible risks had improved. This view on sophisticated risk mitigation and management was incorporated in the bank

capital adequacy regulation (Basel II). Knot argues that while Basel I (1988) was blunter and rawer but with quite high capital requirements, Basel II (January 2008) was refined and model-based but demanded lower capital requirements. DNB had put trust in the banks, but afterwards this turned out to be a major mistake. When in 2008 banks faced

financial distress, it turned out that their risk management was far from efficient.

Basically, DNB had agreed to measures which led to a reduction in capital buffers. Thus, when the buffers were necessary, they were not set in place. At that point, DNB was seen as a lender of last resort and needed to provide liquidity to support the banks7. DNB’s perception

7 _{DNB can provide banks with liquidity support if a financial institution is no longer able to fulfil its obligations} on the short-term. In that case, DNB is ‘lender of last resort’. Whenever a financial institution faces solvability problems however, it has to ask for support from the state. DNB cannot fulfil this role. (Commission de Wit 2011: 425).

of the stability of the financial system was wrong, but not misplaced since initially, it was believed that some banks were simply hit by an idiosyncratic wrinkle. Nevertheless, a global crisis had deeply overrun the Netherlands. DNB approved the takeover of ABN AMRO by Fortis bank. This purchase had to be approved by DNB, which they did. In fact, this was the beginning of the end for Fortis8_{. It turned out that while it seemed like Fortis was quite stable,}

in fact, they were unable to buy ABN AMRO and when they did, they were nearly unsavable (Interview: Knot 2019). During the period after this, Fortis and ABN AMRO were

nationalized and ultimately ING, AEGON, and SNS Reaal received capital injections. What factors led to the decision to do so? This chapter aanalyzes the Dutch bailouts from 2008 until 2009 and discusses which determinants of bailouts played a role in the case of the

Netherlands. Besides, each proposition is tested. Furthermore, each paragraph discusses fractions of the Dutch bailouts and which factors played a role in bailing out the Dutch banks. Each paragraph opens with the proposition that will be tested in that section.

5.1 The political organization (government-bank relations) prior to the

crisis

In this section the relationship between the government and bankers (as part of the political organization of the state) is discussed.

Proposition 1

The political organization of a country determines the likeliness of bailouts.

The relationship between the government and bankers is arguably important for the likeliness of bailouts. So, what was this relationship like in the Netherlands? All my interviewees

indicated that during the crisis, the relationship between bankers and politicians was not good, to say the least. Negotiations were rough10_{, and government officials felt no pity towards}

bankers, who claimed to be always thinking about the big money (Interview: ter Haar). “We always said, one of the conclusions you should draw from the financial crisis, is that the

8 _{Statement of non-objection given by DNB has led to severe consequences (Commission de Wit 2011: 109).} 9 _{All interviews have taken place in May 2009. All references thus refer to the interviews I have performed in} May 2019.

relationship between the government and the banks was simply not good enough. It was just not good enough for the banks to trust the government and admit that they saw a problem arising. There was no mutual confidence” (Interview: Bröcheler)11_.

Furthermore, banks were quite arrogant and condescending. Since banks like ABN AMRO and ING were internationally big players, they felt like the Dutch state was just a small, meaningless factor in their business. Basically, they believed that they were greater than the Netherlands (Interview: a former member of parliament, Knot, Ministry of Finance policymaker, Wilders). Additionally, “during the crisis banks discovered that they had

severely neglected their relationship-management with the government in the years before the crisis” (Interview: Knot). Thus, the attitude of banks towards the government was not

something that worked in their favor when all of a sudden, they needed the government to save them from facing bankruptcy. As a consequence, the government wanted to have a serious talk with the banks and rough negotiations followed (Interview: Knot). Negotiations were intense up to the point where banks either had to accept the conditions given by the Ministry or accept the risks of bankruptcy (Interview: ter Haar). Besides, there was no regular contact between the Ministry of Finance and the CEOs (Interview: ter Haar).

To conclude, there was not a tight relationship between policymakers and bankers. Yet, nearly every bank was still bailed out. The political organization (bank-government relation)

proposition is thus not supported and is not the main determinant of bank influence.

5.2 The political organization (collective negotiation and burden-sharing)

and bank lobbying

In this section, I discuss the process prior to the bailouts and the negotiations that took place. Did the banks request money, or were they so strongly monitored that the government decided to intervene by themselves? To what extent was a collective solution reached, including burden-sharing for the private sector? The collective solution mechanism and burden-sharing variable (as part of the political organization of the state) will be tested. The role of lobbying is also discussed. In general, all banks were involved in the negotiations and