Non-performing loans in Ireland and The Netherlands. The political
economy of promoting private debt: a comparative process-tracing study
Name: Mark van der Woude
Student number: s1474243
Supervisor: Dr. N.A.J. van der Zwan
Date: 07-08-2020
Contents p. 3
Chapter 1: Introduction p. 5
Chapter 2: Theoretical framework p. 8
2.1 Economic theory on NPLs p. 8
2.2 Varieties of Capitalism p. 10
2.3 Historical Institutionalism p. 13
2.3.1 General developments within historical institutionalism p. 13
2.3.2 Exogenous institutional change p. 14
2.3.3 From exogeneity to endogeneity p. 15
2.3.4 Path dependency p. 16
2.3.5 Some remarks on Pierson’s path dependency p. 17
2.4 Expectations p. 18 Chapter 3: Methodology p. 20 3.1 Research design p. 20 3.2 Case selection p. 20 3.3 Data collection p. 21 3.4 Data analysis p. 23
3.5 Validity, generalizability and reliability p. 23
Chapter 4: Case study The Netherlands p. 25
4.1 Deregulation in the 1990s p. 25
4.2 Deregulation enters the financial markets p. 27
4.3 Laissez faire in credit supply p. 27
4.4 Political motivation for credit p. 29
4.5 From fiscal incentives for homeowners to incentives for bankers p. 31 to sell mortgages
4.6 Supranationalization and technocracy in banking regulation p. 32 4.7 Securitization: making profits from mortgages on the financial p. 33 markets
4.8 Post-crisis: re-regulation of financial markets p. 34 4.9 Back to theory: increasing returns and final path deviation p. 36
Chapter 5: Case study Ireland p. 38
5.1 The ‘Celtic Tiger’ and economic transformation p. 38
5.2 Institutional transformation p. 39
5.3 Credit growth as political hobbyhorse p. 40
5.4 Risk seeking: increasing profits on mortgages and higher p. 41 LTV ratios
5.5 Loosening lending requirements p. 42
5.6 Political incentives for Fianna Fáil p. 43
5.7 The rise of non-performing loans after the crisis p. 45 5.8 Little attention for private debt and regulation p. 45
Chapter 6: Conclusion and discussion p. 49 6.1 Theoretical framework p. 49 6.2 Results p. 50 6.3 Limitations p. 51 6.4 Last remarks p. 51 Literature p. 53
Chapter 1 Introduction
The financial crisis of 2007/2008 had enormous political and economic implications for the entire European Union. Various markets collapsed, the economy started to enter a long period of stagnation, national governments responded with austerity policies, and the European Central Bank initiated the policy of quantitative easing1, which was noted as a particularly far-reaching measure. Much has already been published about the political and economic causes, the course and consequences of the financial crisis (e.g. Crotty, 2009; Foster & Magdoff, 2009). One of the characteristics of the crisis was a rapidly rising ratio of non-performing loans. These are loans from lenders to credit recipients where the payment obligation has not been fulfilled for more than 90 days, or there is the impression that the loan cannot be repaid within a reasonable period (Basel Committee, 2016, p. 8). A high NPL ratio leads to increasing financial uncertainty among lending institutions, and will ultimately have a multiplier effect on, for example, private investments (Klein, 2013, p. 6). An increased NPL ratio is therefore an important economic outcome of the financial crisis.
Despite the economic downturn in Europe in the aftermath of this crisis, there are noticeable differences between European countries, particularly at the levels of non-performing loans. In the heyday of the financial crisis, Ireland had an NPL ratio of 25 percent, while the Netherlands had an NPL ratio of 3 percent (CEIC Data, 2020). This significant difference between the two countries is striking if you look at this difference from the well-known North-South contrast; gross differences would rather be expected between a Northern European and a Southern European country. Nölke (2011, p. 1) states, however, that Ireland also belongs to the economic periphery, and a country like the Netherlands belongs to the economic core.
The Netherlands and Ireland are similar in many areas. For example, both countries use the Euro, knew a significant economic boom in 1990s, and can be characterized as coordinated market economies (Hall & Soskice, 2001). These equal circumstances would initially not lead to the assumption that there are rigorous differences in NPL ratios. My research question is the following:
Why has Ireland a higher non-performing loans ratio than the Netherlands after the financial crisis?
1 Quantitative easing can be understood as the central bank buying assets on financial markets. The aim of
quantitative easing is to boost economic activity by money creation that leads to lower borrowing rates (Blanchard, Amighini & Giavazzi, 2016, p. 494).
However, this study is less economical than a reader would expect. The main focus lays on the undercurrent with reference to the political-institutional context.
This research will be conducted as a comparative process-tracing study. The main contextual factor here concerns institutions. A handy framework could be Hall and Soskice's (2001) Varieties of Capitalism model. This model deals with how different capitalist systems, on the one hand a liberal market economy and, on the other hand, a coordinated market economy, can achieve comparative advantages over each other. One problem is that the variables involved in this model only refer to macroeconomic, institutional factors, and disregard private household debt, the crux of the NPLs. Also, the more "typological" way of theorizing is not sufficient in linking institutional divergence to a specific institutional outcome (Witt & Jackson, 2016, p. 780). In addition, Shih (2004), for example, states that a particular economic policy is the result of both political and economic considerations, which seems to be analogous to the reasoning of Acemoglu and Robinson (2012), that political institutions determine the character of economic institutions. Strech (in Witt & Jackson, 2016, p. 782) also states that normative restrictions and social obligations play a role in economic institutional arrangements.
As briefly outlined above, the question arises whether the Varieties of Capitalism framework of Hall and Soskice (2001) will be sufficient for answering the research question. As Witt and Jackson (2016) indicated, in the Varieties of Capitalism framework, institutional dynamics have received little attention in explaining specific outcomes. For example, Thelen (1999) argues that an institutional outcome is determined by intersection and interaction of various economic and political processes. "Interaction of different processes" fits in with the criticisms mentioned above about the almost exclusive focus on economic institutions in the Varieties of Capitalism framework. One way to, at least partially, fill the theoretical vacuum is to use other historical institutionalism as a theoretical frame. The more critical Political Economy scholarship will also receive the necessary attention. The rationale for 'deviating' from the Varieties of Capitalism framework is therefore to stay away from one-size-fits-all explanations, and to attribute a more important role to the broader, and more critical, context of private debt.
Pierson's (2000) propositions about path dependence and increasing returns will be central to this historical institutional frame. Path dependence means ‘that what happened at an earlier point in time will affect the possible outcomes of a sequence of events occurring at a later point of time’ (Sewell in Pierson, 2000, p. 252, quoted). Path dependence is crucial here, because of the process-tracing method. In both cases, the starting point is to identify whether
there is financial regulation or deregulation (in terms of credit supply). After that, we will try to identify whether there is path dependence or deviation (and how to account for that) and what consequences those developments have for NPLs. More explanation will follow in the theoretical framework.
In chapter 2 the theoretical framework will be discussed, which will look at the Varieties of Capitalism framework, focusing primarily on the relevant criticisms. Historical institutionalism will be discussed in general terms, with a specific focus on institutional change. Finally, the relevant propositions and theoretical expectations are discussed. In Chapter 3, the methodology will be explained, which includes looking at the research design and data collection and analysis. Chapter 4 deals with The Netherlands, in which the period from 1990-2013 is central to this process-tracing study. The Irish case follows in chapter 5, according to the same periodization. Subsequently, in the last chapter, conclusions will be drawn, and a review will be held with the aim of stimulating the discussion.
Chapter 2 Theoretical framework
The ‘societal’ reason for this research is that the difference in the NPL ratio between Ireland and the Netherlands is remarkable, with Ireland being an outlier. Therefore first, an explanation will follow on non-performing loans, based on economic theory (e.g. Nkusu, 2011). The Varieties of Capitalism framework of Hall and Soskice (2001) is discussed below. The main propositions will be discussed. This framework has two shortcomings for this study; private debts are not considered, and the focus is mainly on equilibria, while the underlying political dynamics are insufficiently explained (Pierson, 2000). Historical institutionalism, based on propositions of Pierson (2000) and Capoccia (2016), will fill this theoretical vacuum, especially in view of the importance of the institutional dimension in NPLs (Nkusu, 2011). The theoretical discussion about historical institutionalism will be broadly explained, after which we will zoom in on relevant propositions.
2.1 Economic theory on NPLs
As mentioned in the introduction, the NPL ratio is the outcome in this study. There is a so-called default if the payment obligation is not met for more than 90 days, or if there is a suspicion that the loan can no longer be repaid within a reasonable period (Basel Committee, 2016, p. 6). The next issue that looms up is which parts of a bank balance sheet are included in the NPL value. The Basel Committee (2016, p. 8) states that all financial instruments that banks use in calculating the required capital requirements must be counted as non-performing exposures. There is no universal conception of an NPL concerning which parts of the bank balance sheet should be used in counting the total exposures (Basel Committee, 2016, p. 16). However, this is the case in a European context, since all European Member States must comply with the regulatory requirements of the Capital Requirement Directive (EU regulation), that flows from the recommendations of the Basel Committee.
There are roughly three trends in the economic literature that treat NPLs. First, macroeconomic scholars have focused on the relationship between the NPL ratio and the risk of a crisis, which appears to be a positive correlation (Nkusu, 2011, p. 5). The intuition behind this is that an increased NPL ratio has an impact on several macroeconomic variables (such as disposable income and investments) that are suitable for predicting a crisis. Klein (2013, p. 6) illustrates this by arguing that an increased NPL ratio can put downward pressure on the investment drive of financial institutions, which also has negative multiplier effects on the economy.
A second area of scholarship focuses on analyzing aggregated NPL ratios. Macroeconomic indicators, such as the housing market and total private household debt, are compared with the aggregated NPL ratio (Nkusu, 2011, p. 5). Messai and Jouini (2013, p. 853) show that the NPL ratio positively correlates with variables such as inflation and money supply, while there is a negative correlation with GDP growth. From an economic intuition, the negative correlation between the NPL ratio and GDP growth can be expected. In a booming period, or rising economic growth after a recessionary period, disposable income will be higher than in a recession, resulting in, for example, employment levels above natural unemployment levels (Blanchard, Amighini & Giavazzi, 2016). Then, it is also easier for a borrower to meet the corresponding payment obligations.
The third trend in the literature touches the main question of this study the most, since this trend focuses more on the institutional domain of NPLs. Macroeconomic performance is linked to the quality of banking management and its policy choices (which can be proxied by the operating expenses-to-operating income ratio). The latter two variables are then linked to financial regulation and the incentive structure (Nkusu, 2011, p. 4-5). For example, Podpiera and Weill (in Messai & Jouini, 2013, p. 854) stress that management performance is important for maintaining financial stability. Jiminez and Saurina (in Messai & Jouini, 2013, p. 854) add that banks that lend more than responsible and charge a high interest rate (which flows from financial regulation) are more likely to face payment problems from the borrower. The latter actually comes down to moral hazard, if you consider the goal of financial stability as a collective good (for more information on moral hazard within banks, see Dowd (2001)).
Inefficiencies within banking management may result in moral misconduct. The microeconomic intuition would say that financial stability is a common-pool resource, and thus is non-excludable and rival in character (Olson in Ostrom, 2008). Assuming that a banker is rational and thus makes a cost-benefit assessment in economic decisions, this banker will choose to provide loans until the marginal costs (MC) equal the marginal benefits (MB). Until the financial crisis, there were too few regulatory arrangements to discourage or eliminate perverse incentives (Roberts, 2019; Agarwal & Wang, 2009). By granting loans until the point MC = MB is reached, also loans are provided that have too little collateral (i.e. have a high chance of default) and can thus end in NPLs. Due to the fact that the social costs of non-performing loans (i.e. a deterioration of financial stability) are not included in the cost of a loan (interest flows to the lender), a negative external effect is created (Nas, 2016, p. 47).
From a theoretical point of view, it is more convenient to approach NPLs with marginal social costs (MSC) and marginal private costs (MPC), because these concepts consider societal
consequences rather than solely the economic ones. MPC consist of the capital and labor input for a specific service, while the MSC includes, besides the MPC, the external costs associated with the MPC. The producer does not pay for these external costs (Nas, 2016, p. 48). When the NPL ratio is high, or even present, the MSC is higher than the MPC, since a high NPL ratio has macroeconomic consequences, as shown above. The common pool resource, in this case financial stability, will be "exhausted" if the status quo is maintained, since the producer (theoretically) has no incentive to consider the external, societal costs of its service.
In economic theory, there are roughly two approaches that deal with resolving (internalizing) a negative external effect. First, there is Coase's theorem. In this theorem, Coase (in Nas, 2016, p. 50) states that an external effect can be internalized through negotiation. Taking into account that there is a small number of cases and that the present transaction costs are low, the property rights of the negative externality can be assigned to a market party (Coase in Nas, 2016, p. 50). This does not apply in the case of NPLs, as it involves a large number of diverse actors. A second way to internalize the negative external effect is through government intervention. According to the classical view, this can easily be achieved by introducing a Pigouvian tax (Nas, 2016, p. 52). The amount of this tax must be set so that the MPC equals MSC. As a result, the price of the negative externality is included in the market price.
The above overview mainly focuses on some fundamental theoretical perspectives with regard to NPLs, with the aim of clarifying the underlying economic intuition. It would not be convenient to use this framework to answer the research question. The above economic theory is prescriptive, and its accompanying one-size-fits-all approaches could be useful in a quantitative cross-sectional study into, for example, the relationship between NPLs and the policy rates. This research focuses on the causal mechanisms, and in specific on the underlying political-institutional foundation of NPLs. The following section will therefore focus on the Varieties of Capitalism framework of Hall and Soskice (2001), shifting the focus from economics to political economy.
2.2 Varieties of Capitalism
Considering the main question of this research, it seems reasonable to approach the research question from a comparative capitalism perspective. Indeed, there is a different outcome (in NPL ratio) between two capitalist economies; namely, Ireland and the Netherlands. Hall and Soskice (2001) would approach this outcome in terms of comparative advantages, thus the advantage in this context implies a lower NPL ratio. The core of Hall and Soskice's (2001)
framework is that different sets of institutional arrangements lead to different comparative advantages (Schneider et al. in Witt & Jackson, 2016, p. 779).
Hall and Soskice (2001) argue that radical innovation takes place in a liberal market economy, which will lead to comparative advantages. Liberal market economies are characterized by top-down relations inside firms, equity-based capital markets, short-run profits and shareholder dominance (Kesting & Nielsen, 2008, p. 25). However, in a coordinated market economy, incremental innovation takes place, which will lead to comparative advantages. Furthermore, a coordinated market economy is characterized by bargaining relations between firms and unions, stakeholder dominance, and long-term and bank-based capital (Kesting & Nielsen, 2008, p. 25). The central variables in Hall and Soskice’s (2001) thesis are of an institutional nature, which are complementary and mutually reinforcing (Witt & Jackson, 2016, p. 780).
An important theoretical shortcoming in the framework of Hall and Soskice (2001) is that little attention is paid to the political dynamics of institutions. For example, Kang (2006) indicates that there has mainly been a reorientation with regard to government intervention (instead of a reduction). It is therefore not the case that there has been a reduction in government intervention after the waves of privatization in the 1980s. Governments are therefore not sidelined in national economies.
The institutional variables Hall and Soskice (2001) use in their thesis are mainly apolitical. However, Amable (in Kang, 2006) argues that the political dimension is essential at the macro level of these institutions. Institutions therefore always have a political function and Amable (in Kang, 2006) emphasizes that comparative advantages should therefore always be seen in the light of this political dimension. Kesting and Nielsen (2008, p. 26) indicate that the formal institutions in the framework of Hall and Soskice (2001) are supported by informal institutions, such as informal rules, that are of a political nature. Kesting and Nielsen (2008) therefore argue that unjustifiably little attention is paid to political dynamics in the framework of Hall and Soskice (2001).
Fundamental criticism mainly concerns the mechanisms behind the propositions that Hall and Soskice (2001) propose. For example, Goodin (in Kesting & Nielsen, 2008, p. 27) shows that political discourse can indeed influence how a country converts from a CME to an LME. Also, the "nature" of politics is not fully crystallized. Because Hall and Soskice (2001) assume that coordination takes place in a political economy, political conflict and crises are wrongly ignored (Howell in Kesting & Nielsen, 2008, p. 29). The institutional focus therefore appears to be too static, while political factors imply more dynamism.
By not considering political conflict in the mechanisms, institutional dynamics are analyzed in a relatively 'old' way. Haggard (in Kesting & Nielsen, 2008, p. 28) states that institutions are looked at from stickiness and functionalism. The foundation for change cannot be explained for. In response to these criticisms, Hall and Soskice (2003, p. 245) argue that political processes are indeed relevant for explaining institutional change. At the same time, however, the authors argue that political processes are generated by many factors and should therefore be analyzed accordingly (Thelen in Hall & Soskice, 2003, p. 245).
Witt and Jackson (2016, p. 782) also indicate that little attention is paid to the causal mechanisms behind the institutional equilibria. Pierson (2000, p. 264) argues that national economies are highly path dependent. Varieties of Capitalism looks at a certain economic equilibrium and explains this from the institutional status quo. So, in such a study t1 would be the employment miracle and t2 the financial crisis. Therefore, by using a process-tracing approach, Hall and Soskice's (2001) framework does not necessarily seem convenient to describe a process from a beginning to an end.
A second criticism of the framework of Hall and Soskice (2001) is that private debt is not looked at (which the Y mainly focuses on). Fernandez and Aalbers (2016, p. 73) argue that this is because the VoC literature is based too much on the stickiness of institutions, and therefore cannot accommodate the dynamics of the mortgage and capital markets (and their associated private debts). Despite the fact that private debt seems to be a purely economic variable, according to Metz (2018), debts have a political element. In a more classic approach, private debt is seen as a relationship between a debtor and a creditor or as the result of a rational choice of an individual, while private debt carries an intrinsic power element (Metz, 2018, p. 78). Roberts and Soederberg (in Metz, 2018, p. 79) argue that debts are asymmetrical and exploitative. When an individual enters into debt, there is a chance that he will become inextricably linked to certain life choices and must move into unbalanced power relationships (LeBaron & Roberts in Metz, 2018, p. 79).
In financial markets and services, there is overall trend of deregulation since the 1980s. (Fernandez & Aalbers, 2016, p. 71). Liberalization does not mean, however, that political institutions left the financial markets untouched. Metz (2018, p. 79) illustrates this with the European discourse of encouraging and promoting securitization, while securitization has also been one of the causes of the global financial crisis (Blanchard, Amighini & Giavazzi, 2017, p. 178; Crouch, 2009, p. 392). According to Kohl (2018, p. 178), financial deregulation is politically motivated, because deregulation could ensure that the housing market could remain accessible to all layers of the population. Crouch (2009, p. 396) even argues that low- and
middle-income groups can only maintain their consumption level if they have access to credit. Governments therefore feel compelled (from a political motive) to keep lending as accessible as possible, with the aim of keeping consumer confidence at the desired level.
In conclusion, it therefore does not seem appropriate to take Varieties of Capitalism by Hall and Soskice (2001) as the sole theoretical starting point. First, too little attention is paid to the political dynamics of institutions. Secondly, Hall and Soskice (2001) also do not consider private debt and, as explained above, that also has a political dimension. It would therefore seem more reasonable to take other historical institutionalist scholarship as a starting point, which could fill the "political-institutional vacuum" of Hall and Soskice (2001). By being guided by historical institutional propositions, causal mechanisms with the associated political dynamics can receive more attention, which are relevant around NPLs, as argued by Nkusi (2011).
2.3 Historical institutionalism
Historical institutionalism will serve as a theoretical basis for this research. First of all, broad developments within this theory (the so-called state of the art) will be explained, after which the most important theoretical propositions relevant to this research will also be outlined. 2.3.1 General developments within historical institutionalism
The historical institutional approach is a result of the behavioral revolution of the late 1950s and early 1960s. Before this revolution, there was an institutional movement that was mainly focused on formal institutional structures and arrangements (Thelen & Steinmo, 1992, p. 3). Thelen and Steinmo (1992, p. 2) indicate that historical institutionalism focuses on how political battles are mediated in the institutional context in which these battles take place. An important implication of this conception is that institutions by themselves do not by definition determine the (political) outcome. Institutions structure the underlying political struggle and can therefore indirectly influence the outcome (Steinmo & Thelen, 1992, p. 3). Where Steinmo and Thelen (1992) emphasize political conflict, Lowndes (2010) emphasizes the implications of institutions for the individual. According to Lowndes (2010, p. 65), the core of historical institutionalism is the way in which previous institutional choices influence the future decision-making of individuals.
The first question that arises is what exactly is meant by an institution, since the new institutionalism rejects the old institutional conception and therefore no longer only looks at almost "physical" forms of institutions. For example, Hall (quoted in Steinmo & Thelen, 1992,
p. 2) defines an institution as "the formal rules, compliance procedures, and standard operating practices that structure the relationship between individuals in various units of the polity and economy". In contrast to Hall's more formal definition, North (1991, p. 97, quoted) argues that institutions are "humanly devised constraints that structure political, economic and social interaction".
The phenomenon of institutional change has been central to the ontological discussion within historical institutionalism for decades (Ikenberry in Immergut, 2002, p. 242). At the same time, a solid explanation must be given both for stability and for change. Immergut (2002, p. 243) indicates that the historical institutional lens requires a historical view, but that this historical view makes it inevitable to observe the unpredictability of human nature. Conran and Thelen (2016, p. 4) add that the historical view leads to institutions being seen as stable and exogenous (due to the inherently long lifespan of these institutions).
Due to the narrative way of analysis, a factor such as contingency can indeed be important, as it can be of explanatory value within the causal mechanism (Immergut, 2002, p. 243). Immergut (2002, p. 247; 254) indicates that institutional change cannot be approached singularly, since causality runs through multiple channels. According to Immergut (2002), there are always many contingencies and subjective observations that can explain a certain institutional change, but at the same time they cannot be systematically classified (Immergut, 2002, p. 254).
2.3.2 Exogenous institutional change
Contingency is one form of an exogenous change. This is also reflected in historical institutionalist terms such as critical junctures. Illustrative are the propositions of Acemoglu and Robinson (2012), that also assign a certain role to contingency, and argue that institutions can only change through exogenous shocks. And vice versa, institutional stability is guaranteed if no exogenous shocks occur (Acemoglu & Robinson, 2012).
As discussed above, exogenous shocks were seen by traditional historical institutional scholars as the cause of institutional change. Institutional stability is based on the idea that the situation has reached point X, and institutions structure political reality from there. Mahoney (in Conran & Thelen, 2016, p. 5) would argue that these institutions are then path dependent, and therefore are the result of choices made earlier. Pierson (in Conran & Thelen, 2016, p. 5) would explain this stability through institutional consolidation, because there are increasing returns from the existing institutional arrangements. Conran and Thelen (2016, p. 6) argue that viscosity as a concept provides too little foundation for explaining stability. This brings Conran
and Thelen (2016) to their central thesis: research within historical institutionalism has changed. Besides these exogeneous factors, endogenous factors have become more important in explaining institutional stability and change.
2.3.3 From exogeneity to endogeneity
Conran and Thelen (2016) identify four endogenous gaps that historical institutional scholars consider when it comes to institutional change. First, there is the inherent limitation of institutional design (Conran & Thelen, 2016, p. 7). The crux lies in bounded rationality, as posited by Simon (1990). Within an institution, individual actors have too little insight into the consequences of their actions and therefore have too little insight into the functioning of the institution itself. Second, Conran and Thelen (2016, p. 7) argue that there is a discrepancy between the de iure and de facto functioning of the institution, because political preferences try to penetrate into the institutional arrangements. Third, institutions are being fought for; the winners and losers of, for example, an institutional rule will always try to strike back to change the institution (Conran & Thelen, 2016, p. 7). Fourth, there is the time factor. An institutional rule is formed in a specific temporal context. When this context changes, it can lead to a reinterpretation of these rules, and thus institutional change (Conran & Thelen, 2016, p. 7). The latter factor cannot be called fully endogenous, but rather balances on the interaction between the exogenous environment and the endogenous action that might follow.
The role of agency seems to have been undervalued for a long time by ‘old’ historical institutionalist scholars, although this is unjustified. Capoccia (2016, p. 6) therefore emphasizes that institutions are formed on the one hand by political actors, but on the other hand those same political actors are limited by the institutional arrangements they have formed. Hall (2016, p. 1) wonders whether an institution determines behavior or the institution itself is the object of strategic behavior. This theoretical ambiguity is opposed to rational choice institutionalist propositions, that mainly state that an institutional equilibrium is caused by a Pareto optimal situation for the actors involved (Lowndes, 2010, p. 75).
Capoccia (2016, p. 3) argues that noncompliance with formal rules (and therefore is endogenous due to the focus on agency) may on the one hand lead to reinterpretation of these rules with resulting incremental changes. The exogenous variant of noncompliance is institutional layering. In institutional layering, new rules are added to the existing rules, with the aim of generating a different outcome (Schickler in Capoccia, 2016, p. 5). On the other hand, noncompliance may also entail a maintenance of the status quo, as institutional incumbents resist change (Capoccia, 2016, p. 3).
The fact that incremental change does not necessarily have to be beneficial for structural institutional change, is due to the composition of agency within the institution. Streeck and Thelen (in Capoccia, 2016, p. 5) therefore state that underlying processes of coalition formation are important to understand institutional change. Institutions would therefore be more dynamic than self-reinforcing (Mahoney & Thelen in Capoccia, 2016, p. 5).
Within historical institutionalism, the presence of coalition formation and its agency is now considered as an important variable in endogenous change, but historical institutionalists assume that coalitions will always be fought for by institutional actors that want to maintain the status quo (Hall, 2016, p. 19). Thelen (in Hall, 2016, p. 21) also emphasizes that reinterpretation of formal rules by agency is not necessarily straightforward. Streeck (in Hall, 2016, p. 24) even argues that attempts to bring about institutional changes are experimental in nature. In the current theoretical state within historical institutionalism, it therefore appears that no linear transition can take place from a reform coalition to a new institutional reality. This is also diametrically opposed to the polished shifts in coalitions within the equilibrium thinking of rational choice institutionalists.
2.3.4 Path dependency
A widely used concept within historical institutionalism is path dependency. Sewell (quoted in Pierson, 2000, p. 252) describes path dependency as ‘that what happened at an earlier point in time and will affect the possible outcomes of a sequence of events occurring at a later point of time.’ According to Sewell, temporality and sequence are important factors in determining an institutional outcome. Pierson (2000) makes the link with increasing returns. That is, as time progresses, changing from alternative A to alternative B becomes more costly (Pierson, 2000, p. 251). Hence, temporality is important here for the analysis, as it affects the path to be chosen. When increasing returns are present, it is not surprising that the path already taken continues to be followed, as the marginal returns of the current path increase over time, compared to the not-chosen alternative (Pierson, 2000, p. 252). Arthur (in Pierson, 2000, p. 254) adds that actors will only have incentives to continue on the path already set when increasing returns are involved. Therefore, in the case of increasing returns, opportunity costs do not seem to exist.
Arthur (in Pierson, 2000, p. 254) states that there are four factors that can lead to increasing returns: first when there are significant fixed costs, second when there are learning effects, third when there are coordination effects and fourth when there are adaptive expectations. When there are increasing returns, improvements in core economic activities can have spillover effects towards related economic activities (Pierson, 2000, p. 255). Especially
when these activities are knowledge intensive, specialization will take place and comparative advantages over other countries may emerge (Pierson, 2000, p. 255). In this research context, comparative advantages are not only seen as a result of economic-institutional factors, but more attention is also given to the political dimension.
Pierson (2000) shows that increasing returns are politically relevant due to the following factors. First, because of collective action problems. Pierson (2000, p. 258) notes the assumption that individuals adjust their behavior in view of the expectations of the behavior of other actors. Why does a collective action problem make increasing returns politically relevant? If individual A has the expectation that individual B will free ride, then individual A will also start free riding, which is in line with adaptive expectations, thereby empowering increasing returns (Pierson, 2000). If a possible 'path' is initiated as a result of free riding, this path will be continued as long as the marginal costs are lower than the marginal revenues.
Second, Pierson (2000) argues that institutional density enhances increasing returns. Policy arrangements are the incentives and resources of political actors (Pierson, 2000, p. 259), from which the political pushing and pulling begin. A political actor will always deal with arrangements within various dimensions. It therefore seems difficult for a political actor to disconnect from institutional reality. In addition, Pierson (2000, p. 259) adds that actors involved adapt their actions to existing institutional arrangements, which would result in very high costs when moving from that arrangement (i.e. path). Path dependence therefore arises from the frequent presence of institutions (the so-called density), which entails increasing returns.
2.3.5 Some remarks on Pierson’s path dependency
Pierson (2000) indicates that in case of increasing returns, the political environment is characterized by the following factors: multiple equilibria, contingency, time and sequence, and institutional inertia (Pierson, 2000, p. 263). There are some things that stand out in view of the current state of art. First, that contingency may matter. Within historical institutionalism, the statement ‘coincidence’ is no longer considered sufficient, as explained by Capoccia (2016). In addition, institutional inertia indicates the absence of an explanation for change. Capoccia (2016) already pointed out the importance of coalition formation within institutions, and that these can subsequently explain institutional dynamics. Peters, Pierre and King (2005, p. 1277) concretize this by stating that the complexity of the policy formation should become much more prominent in research. Political conflict and consensus determine this complexity. Peters et al.
(2005, p. 1288) therefore argue that the role of ideas and political preferences is underestimated, while at the same time the importance of existing institutional arrangements is overestimated. As previously indicated, layering can also lead to institutional change (Schickler, 2001; Capoccia, 2016; Thelen & Mahoney, 2010). The rationale for introducing layering is that this concept is also a further elaboration of the complexity of policy formation mentioned by Peters et al. (2005). Layering refers to adding new rules to existing rules and amending and revising existing rules (Thelen & Mahoney, 2010, p. 16). On the one hand, layering offers opportunities for institutional losers, who are unable to change the institutional status quo, but can potentially change the rules from the outside (Thelen & Mahoney, 2010, p. 17). On the other hand, the actors within the institutional status quo are prone to gradual change imposed from the outside, and as Thelen and Mahoney (2010, p. 17) argue, those gradual changes can have a significant impact in the long term.
In this research the proposition of Pierson (2000) will be used as theoretical starting point. However, the notions of Capoccia (2016), Peters et al. (2005) and Schickler (2001), among others, will be added to the analytical proposition of path dependency and increasing returns. By adding some theoretical elements, an attempt is made to overcome some fundamental blind spots within historical institutionalism. The earlier critiques on the Varieties of Capitalism framework of Hall and Soskice (2001) are mitigated, since these propositions consider political dynamics to be important, which was exactly the main shortcoming of the Varieties of Capitalism framework. Thelen (1999) and Pierson and Skocpol (in Thelen & Mahoney, 2010, p. 9) also indicate that institutional outcomes should be seen in the light of various economic and political processes. A single look at chance and institutional stability (the unitary equilibrium according to Pierson (2000)) would then in all likelihood not be sufficient to expose the causal mechanisms.
2.4 Expectations
From the theoretical framework outlined above, it is possible to derive some expectations with regard to our case.
If a government chooses to aim to support the supply side of the economy, financial deregulation may be the result. From the economics literature as described in 2.1, I derive the expectation that financial deregulation may entail negative incentives for banks to structure their lending strategy irresponsibly for society, but profitable for banks themselves. Financial institutions will therefore make every effort to maintain the status quo of deregulation. At the
same time, that status quo may be supported by the discourse of political actors, since extended lending can be a political motive to maintain consumer confidence.
Based on the mechanism of increasing returns, one could expect that the chosen path of financial deregulation and the easy access to credit is not attractive for the political actors involved to deviate from this path, because this is simply too expensive. This choice could result in the amendment of existing legislation and the addition of new laws regarding the financial system (so-called institutional layering). In this way, choices made earlier will continue to shape later choices. The status quo is maintained.
Alternatively, the chosen path of financial deregulation may be resisted institutionally. It could be that the political majority focuses more on consumer protection and therefore does not see rampant lending as the best possible alternative. Political discourse may focus more on the dangers associated with credit.
Theoretically, it is likely that there four different trajectories within the same mechanism. In the beginning there either regulation or deregulation. In the years following, the path of either regulation or deregulation can be sustained of deviated from. When the path of regulation is sustained, there is continuity (i.e. institutional stability), which leads to an outcome of regulation. Second, when deviating from regulation, there is institutional change, that ends up in deregulation. The path may also go in opposite direction. First, in the beginning, deregulation may be witnessed, which will continue over the years (i.e. stability) and ends up in continuity: more deregulation. Second, when there will be deviated from the initial path of deregulation, a certain form of institutional change will lead to regulation.
Chapter 3 Methodology
In this chapter the methodology of this study will be explained. We will start by describing the research design; which research method will be used to answer the research question. The operationalization will then be made clear, in which it will be examined how an attempt is made to witness the individual components of the discussed theoretical mechanisms in the case. An elaboration of the case selection follows below; i.e. why are Ireland and the Netherlands scientifically interesting within the theory and empirical puzzle? The data collection will also be discussed, looking at how the observations will be dealt with and how these observations are of course achieved. Finally, it will be elucidated how attempts are made to draw conclusions from the data, and to what scale these conclusions can then be generalized. This section will also reflect on the validity and reliability of this research design.
3.1 Research design
A process-tracing approach is chosen to answer the research question. Given that both Ireland and the Netherlands have the same starting point, but a divergent end point, it is interesting to look at the trajectory between the start and end points. Within causal process-tracing research, it is common for the research question to be centered around the Y, and to be reasoned from there (Blatter & Haverland, 2012, p. 84). Process tracing makes it possible to uncover underlying causal mechanisms by focusing on several factors that may be important in understanding and uncovering the relevant mechanisms (Blatter & Haverland, 2012, p. 81). Since this research includes both a longitudinal and a comparative element, process tracing is a handy method to look at divergent paths. The process-tracing method can reveal the different trajectories of the present case and is therefore convenient to overcome the comparative aspect of this study. The time delineation chosen covers the period from 1990 to 2013, making this study longitudinal. The chosen theoretical approach, historical institutionalism, forces the design to adopt a historical focus. Process-tracing seems appropriate here as a methodological interpretation.
3.2 Case selection
Blatter and Haverland (2012, p. 84) indicate that an empirical puzzle is a logical starting point for a process-tracing study. This is also the case in the present puzzle, since it is not obvious why Ireland has a significantly higher NPL ratio than the Netherlands. It is also not necessarily in line with expectations that a Northern European ‘employment miracle’ (i.e. Ireland) has a
high NPL ratio. Indeed, a high level of employment is not directly associated with high private debts (Rinaldi & Sanchis-Arellano, 2006, p. 8). The reason for choosing a comparative process-tracing study has to do with the fact that the present cases are comparable, but may show the biggest difference within an important causal factor; namely the institutional domain.
Blatter and Haverland (2012, p. 102) also indicate that accessibility is an important criterion for case selection. The investigator should have access to all relevant documents and information to a large extent in order to expose the causal mechanisms. This is the case for the present cases. The Netherlands and Ireland are both transparent countries, which means that it is plausible that the required government documentation is sufficiently available and can therefore be used for this study (Transparency International, 2020).
In addition, social and practical relevance is important for the case selection in process-tracing (Blatter & Haverland, 2012, p. 102). Since private debts have an impact on both a household and macro-economic level, it is socially relevant to investigate how the institutional foundation can lead to a certain level of debt. Precisely because that institutions are difficult to change (which flows from our theoretical assumptions), it is necessary to expose how these rigid institutions function. This is important for both political actors and civil servants, in order to provide more tools in the search for a solution for a high level of private debt.
3.3 Data collection
From the outlined expectations, some factors follow that will be central (which follows logically from the theory). However, these factors require explanation. Theoretical abstractions need to be distilled in order to be able to operationalize the follow-up. An important part of the theoretical mechanism is, firstly, institutional layering, as described by Schickler (2001), among others. Since the core of institutional layering involves the amendment of existing regulation, and the addition of new rules to these existing rules, it is useful for this research to operationalize institutional layering as financial regulation. The political dynamics of institutional layering are mitigated by alternative factors, which will be discussed below. Financial regulation is reflected in legislative proposals. These will therefore be considered, including the corresponding amendments. The focus is on the keywords regarding ‘borrowers and lenders’, ‘mortgage/lending standards’ and ‘private debt’. For the Irish case,
https://www.oireachtas.ie/en/bills/ is used, while for the Dutch case
https://zoek.officielebekendmakingen.nl/uitgebreidzoeken is used.
Second, the factor ‘coalition formation’ follows from the discussed mechanisms. Within historical institutionalism '2.0', for example, Capoccia (2016) focuses on political coalitions
that can bring about institutional change. Coalition formation could be observed in governmental agreements and parliamentary debates, which will therefore serve as operationalization. In coalition agreements, the focus will be on the passages that deal with the mortgage market, private debts and lenders. In debates, the focus is on the discourse of political actors, both cabinet members and parliamentarians; were attempts made to forge a political coalition to bring about changes regarding financial regulation that pertains to the mortgage market? It is examined to what extent there is a change of point of view of political parties, and whether that point of view is then put forward in another way. The underlying semantics are important for being able to analyze these debates and, where possible, to draw conclusions from coalition formation. In addition, cabinet changes will also be taken into account; in other words, whether opposition and coalition parties have ‘moved’ politically.
A third factor that is largely related to both institutional layering and coalition formation is political conflict. First, this can be operationalized by looking at parliamentary debates about relevant bills and acts. Second, policy documents may be used. These documents can be written by government officials, ministers or even independent actors, like the Dutch and Irish central banks. Third, parliamentary questions can be used, since those may represent the politics stance towards to daily affairs (only those that have a clear link to this study, see the elaboration in the following). Search terms such as ‘private debt’, ‘loans’, ‘mortgage market’ and ‘loan-to-value ratio’ will therefore be used here. In this political dimension an attempt will be made to expose the degree of disagreement / like-mindedness. Because our theoretical framework prescribes a historical focus (and the chosen periodization also includes this), political conflict will be viewed in the long term, however following the periodization of 1990-2013.
A final factor relates to the incentive structure. Incentives are not easy to operationalize, since only incentives created in public space can be observed, while it is obvious that interest groups and stakeholders also maintain contact with relevant political actors in more private spheres, from which incentives can also arise. One way to observe incentives is to analyze roundtable discussions. In such a conversation, relevant stakeholders are invited by politicians to discuss a specific topic. Roundtable discussions on the financial system, macroeconomic stability, banking, the mortgage market and the housing market will be considered in this study. Loan providers should have a relevant position in these roundtable discussions. Discourse is also central to the analysis of this type of debate. Parliamentarians (from both opposition and coalition parties) and ministers may send out certain signals, express expectations and make promises, which could be seen as ‘credible commitments’.
3.4 Data analysis
The last point to be addressed is how inferences will be drawn from this process-tracing study. In addition to the many comparisons of the pros and cons between qualitative and quantitative research, there are points that are specifically relevant to process-tracing in drawing conclusions. First, there is the ‘variable problem’. This means that the researcher is never sure whether all relevant factors can be observed (Bennett, 2010). In quantitative methods this is also referred to as the omitted variable bias. When this is the case, the dependent and / or the independent variables are influenced by a variable that is not included in the researcher's model and thus the results may contain a bias (Studenmund, 2017, p. 176). The other side of the coin with regard to the ‘variable problem’ is that the researcher might as well come across factors that are relevant to uncovering the causal mechanisms, but which have not been hypothesized about (Bennett, 2010). Because the case description will be of a narrative nature, it offers the researcher opportunities to search for factors that have a large explanatory value, so that more robust conclusions can be drawn.
Beach (2016) points out some conditions that must be met before causal inferences can be made. First, it is important to check whether the actual observations correspond with the expected observations (Beach, 2016, p. 8). It is also possible that expected observations are not found. However, this only has consequences for causal inference if, after thorough research, it appears that there is no evidence for a certain expectation (Beach, 2016, p. 8). Finally, Beach (2016, p. 8) indicates that the evidence found must be examined very critically, so that bias does not penetrate into the causal inference. The analysis in this study revolves around observing and explaining the trajectories that Ireland and the Netherlands have taken. An attempt will therefore be made to identify the most important changes, within the chosen periodization, that may be of explanatory value for the divergent trajectories. Subsequently, evidence will be gathered to explain these changes. In order to draw valid conclusions about the causal mechanisms, it seems useful to follow the advice of Beach and Pedersen (2013, p. 129) to always remain suspicious in process-tracing research; any evidence should be approached critically, and ‘too good’ observations are likely to be false.
3.5 Validity, generalizability and reliability
In order to promote the reliability of process-tracing research (and thus avoid systematic error), Beach and Pederson (2013, p. 128) argue that multiple, independent observations should be collected. It is necessary that the sources of the observations are also completely independent of each other, so that no bias can arise. For example, in this study, when looking at cabinet
plans regarding loan standards, evidence attempts to be collected from a cabinet member (which also prevents you from approaching the civil service and coalition parties in parliament, as they are related to each other ), a journalistic source (which has no direct affiliation with the minister concerned) and a scientific source (which has no affiliation with both political and journalistic actors). This is a brief elaboration of triangulation. The observation that we would like to find is then searched in three different, independent sources, in order to increase the validity of the research.
The reliability of this study is guaranteed because only sources that are freely accessible are used. Also, the search terms for primary sources are specified in the above, and in combination with the periodization used, which makes this research replicable. The reliability of this study is also enhanced by the consistent presentation of the analysis. Empirical results are linked to the theoretical mechanisms discussed, all within the described theoretical and methodological framework.
The last issue to be addressed revolves around generalizability. Unlike quantitative research, qualitative research, such as process tracing, emphasizes the importance of context. The ambitions within process tracing with regard to generalization mainly relate to the specific case and associated context, and not to a larger set of populations (Munck in Beach & Pedersen, 2013, p. 172). In this study too, it would be too ambitious to generalize conclusions, since the causal mechanisms in this study are specifically based on the narratives of Ireland and the Netherlands. Beach and Pedersen (2013, p. 157) do state that process-tracing studies can serve as building blocks for further research, in which potentially generalizable causal mechanisms can be created.
Chapter 4 The Netherlands
In this chapter, the Dutch case will be central. It will be discussed how the underlying institutional dynamics functioned from the early 1990s to 2013 (the heyday of the crisis). The structure of this chapter is as follows: first, the starting point will be examined, so what did the status quo of the Netherlands look like in terms of financial regulation. This will be explained, among other things, by referring to concrete legislation and showing how the political status quo was structured. The analysis below will show that a trajectory of deregulation was followed from the 1990s to the financial crisis of 2007. Increasing returns were present, making it too expensive for political actors to deviate from this trajectory. However, the financial crisis of 2007 functioned as a critical juncture, after which path deviation took place. Deregulation of the financial markets gave way to a political attitude that was more focused on regulation and protecting consumer interests. The expectations, as outlined in section 2.4, come true for the Dutch case.
4.1 Deregulation in the 1990s
Economically, the Netherlands prospered in the early 1990s. Real wages rose by an average of 0.9 percent in the first few years, while unit labor costs did not rise until 1995 (Becker & Schwartz, 2005, p. 13). Wage moderation was considered as the means of boosting employment, following a Keynesian intuition. This increasing employment strengthened the economic position of consumers (Salverda, 2005, p. 39). This strengthened position (due to real wage growth) was reflected in increasing consumption. In the housing market, too, the effects of increased private consumption were observed, which mainly resulted in greater demand than supply of houses, resulting in higher house prices (Salverda, 2005, p. 42, p. 57).
Significant institutional shifts in regulation have taken place since the 1980s. Deregulation has gradually become the norm in thinking about the role of the government in ‘private’ economic life. The Geelhoed parliamentary committee (that was concerned with the deregulation of the overall legislative framework) also concluded that government intervention should be proportionally and procedurally austere. In addition, it should not entail too much regulatory burden for both citizens and businesses (Deregulering, wetgevingskwaliteit en geconditioneerde zelfregulering, p. 157). The question that arises from this is how the Dutch deregulation doctrine has penetrated in financial regulation.
The Credit Supervision Act (own translation of ‘Wet toezicht kredietwezen’) of 1992 can provide some insight into the trend of deregulation. In the explanatory memorandum, minister Kok (1992, par. 3.2) states that this law mainly concerns accelerating the credit practice
of financial institutions. It is also emphasized that this law is designed according to the principles of deregulation thinking (Kok, 1992, par. 6.2). Illustrative is the fact that the government explicitly emphasizes that there is no intention to pass on the supervisory costs of the law to the financial sector, which is in line with the characteristic of the Geelhoed Committee (Kok, 1992, par. 7.2).
For the lending institutions, this meant a reduction in transaction costs, which translates into higher profit margins. Within traditional institutional economics, transaction costs are seen as a stressor for a perfectly functioning market (Blom, 2010, p. 23). However, if the negative external effects are considered, as explained by Nas (2016), these transaction costs could be seen as the value of the negative external effect. In fact, if a bank becomes insolvent, it is highly likely that it will be kept alive by the state, as banks play an essential role in preserving financial stability (Hüpkes, 2003, p. 459; Brooks & Lombardi, 2016, p. 61).
A remarkable point in the explanatory memorandum is that Kok (1992, par. 3.2, quoted) emphasizes that the ‘value of the guilder is partly influenced by the actions of the credit institutions’ and that the relevant regulations must embrace this goal. From an economic intuition, this position is understandable, since the value of the guilder is important for a favorable exchange rate and associated export volumes (Blanchard, Amighini & Giavazzi, 2016). It therefore seems plausible that the cabinet considers generous lending growth politically and economically desirable. However, financial institutions themselves are responsible for their financial soundness (Kok, 1992, par. 3.6). Bail-out is explicitly dismissed a spossible policy option when a credit institution threatens to get into serious financial trouble. This should then serve as a proper incentive from the government to lending institutions. Risky loans should then not be provided to consumers, based on basic economic reasoning (since the potential marginal costs are higher than the marginal returns of a risky loan, seen at an aggregated level).
In addition, this illustration is also ideologically interesting. Minister of Finance Kok is a member of the Social Democratic PvdA, which at the time formed a cabinet with the right-wing CDA. The earlier sketch of the Dutch regulatory landscape in the early 1990s shows that the principles of deregulation are ideologically well suited to the CDA, but the PvdA is nevertheless more focused on the practical implementation of policy, in particular on its distributive elements (Deregulering, wetgevingskwaliteit en geconditioneerde zelfregulering, p. 159). Therefore, the semantics in the explanatory memorandum of Kok fit better with the PvdA's ideological signature.
4.2 Deregulation enters the financial markets
As briefly mentioned, the deregulation trend has also had an impact on financial regulation. From the early 1990s, three general principles can be observed on how the government approached the financial markets. First, financial markets are comparable to non-financial markets (Stellinga, 2015, p. 34). Second, market forces ensure economic growth, because capital can be better allocated (ibid., p. 34). Third, facilitating market forces is an objective in itself for the government (ibid., p. 34). These points can be observed separately in the above analysis of the Credit Supervision Act. Facilitating market forces practically meant less regulation. In addition, Underhill (in Stellinga, 2015, p. 50) indicates that self-regulation within the financial sector was seen as preferable to government regulation. Government regulation was considered to be slow and inefficient, while self-regulation was effective and innovative. In addition to the deregulation trend, a second movement is taking place in the early 1990s that concerns the business strategy of banks.
Blom (2010, p. 29) indicates that before the 1990s, the continental European banks were mainly engaged in activities within their own territorial borders, whereby deposits and mortgage lending were core activities. However, in the early 1990s, there was a shift to an Anglo-Saxon business model, with complex financial products and risky investments normalizing within banks' business activity (Blom, 2010, p. 29). This shift is in line with the general development of (neo) liberalization in the 1980s. Lending and deposits are less profitable than, for example, derivatives on the financial markets. Van Poeck and Van Gompel (2013, p. 334) also argue that the trends of deregulation and liberalization have allowed banks to develop into investment banks.
The above has shown that a general deregulation and liberalization trend can be observed in the relationship between the Dutch government and the financial sector in the early 1990s. The starting point of this process-tracing study has been mapped. The question now is to what extent there will be path dependence or path deviation. In the next section, we will focus on the period from the 1990s to approximately 2013. Reflection will be given on the theoretical mechanisms discussed, in order to provide a sufficient interpretation of the empirical findings.
4.3 Laissez faire in credit supply
The Netherlands experienced an increase in short-term credit of up to 50 percent in the 1990s (Fase, 1995, p. 100). This fact is supported by the fact that in the same period the M2 (secondary money supply) experienced an increase of no less than 500 percent (Fase, 1995, p. 100). Money
creation has therefore been an important cause of this development, while the public sector has had hardly any share in it. Loans were therefore found to become more normal for consumers, to which lenders responded favorably. As mentioned in section 4.1, financial regulation has created few obstacles to provide credit.
Bank money creation arises primarily from issuing loans (in all forms). Mortgages, due to the size of the loan, have a significant share in this. Aalbers (2008) emphasizes that mortgage levels were increasingly stretched during the 1990s. The so-called loan-to-value (LTV) ratio increased to ensure that consumers could continue to buy houses, as real wages did not grow at the same rate as house prices (Aalbers, 2008, p. 158). The increase in mortgages (and their size) has led to rising house prices and rising household debt. Where the total household debt was 95 percent of GDP in 1990, it had already increased to 188 percent of GDP in 2000 (DNB in Becker, 2005, p. 1090). DNB also notes (in Becker, 2005, p. 1093) that the annual growth in household debt was 12 percent.
The increased loan-to-value ratio has resulted in an increased demand for mortgages, as consumers were able to borrow more to buy a house. However, the supply-side situation is more relevant to this study. Mertens (2017, p. 14) indicates that the increased loan-to-value ratio is an implicit consequence of the abolished interest rate restrictions, which, according to Mertens (2017, p. 14), are again the result of financial reforms, initiated in political institutions. In terms of dismantling interest rate restrictions, the Netherlands was ahead of the rest of Europe, making it possible for financial institutions to operate in a wider segment of the market and subsequently increase their profit margins (Stellinga, 2015, p. 39).
Aalbers (2008) focuses on the financial institutions themselves. The mortgage acceptance policy would have been relaxed to such an extent that it had become considerably easier for a consumer to obtain a mortgage, while the mortgage level was not necessarily in balance with the financial resilience of that consumer. Aalbers (2008, p. 157) illustrates this with the fact that the credit limit has been stretched by more than 86 percent for a two-income group (with a total income of EUR 41,000). This illustration is consistent with the mechanisms outlined in the theoretical framework, such as by Roberts (2019). Namely, the lender continues to provide credit to the point where the marginal benefits equal the marginal costs. This seems to be the case here, due to the combination of the dismantled interest rate restrictions and the increased loan-to-value ratio.
In addition, financial institutions introduced a new mortgage product in the mid-1990s; the so-called investment mortgage. The investment mortgage does not entail a monthly repayment, as is the case with annuity mortgages, which makes this mortgage riskier for
consumers when changes in disposable income arise. However, it became more attractive for financial institutions to sell this type of mortgage, because higher profit premiums applied when selling an investment mortgage (Aalbers, 2008, p. 158). This incentive to sell risky loans was driven by a shortage of regulation and preferences of the cabinet primarily aimed at promoting home ownership.
4.4 Political motivation for credit
In the Dutch political debate, facilitating credit lending and increasing home ownership are policy goals for the cabinet, despite the possible risks for the financial stability. GroenLinks MP Vendrik (in Tweede Kamer, 2000, no. 1421) asked questions about the overvaluation on the Dutch housing market. When asked whether the Dutch development can be compared to the American stock market, Minister Zalm (in Tweede Kamer, 2000, no. 1421) replied in the negative and states that the overvaluation has had a positive spending effect. To a follow-up question on how Zalm assesses the increased mortgage debt, Zalm replies (in Tweede Kamer, 2000, no. 1421) that this has to do with, among other things, low interest rates and the changing acceptance policy of lenders. Zalm (in Tweede Kamer, 2000, no. 1421) then indicates that home ownership is still below the European average, but home ownership is growing, which is accompanied by an increase in mortgage debt. To similar questions from SGP MPs Van der Staaij and Van der Vlies (in Tweede Kamer, 2000, no. 807) about the explosive growth of mortgage lending, Zalm emphasizes that the increased mortgage debt may bring some sensitivity to the housing market, but that the financial stability of the Netherlands is far from being at stake. In addition, Zalm (in Tweede Kamer, 2000, no. 807) points to the importance of good information provision for consumers regarding mortgage products, but that there is by no means a place for the government in the interaction between consumer and the lender.
Deregulation thinking clearly emerges in the political discourse, as discussed above; in other words, the state should not penetrate the free market with new laws. As stated by Stellinga (2015, p. 34), the basic principle that the government must facilitate market forces is widespread. In this respect, market forces were expressed in new mortgage products, such as the investment mortgage, which are risky for the consumer. This is also an elaboration of the pattern identified by Blom (2010, p. 29) that banks are increasingly focused on providing risky products. In the theoretical framework, some scholars within the critical political economy were already introduced. Roberts and Soederberg (in Metz, 2018, p. 79) stated that debts are in principle asymmetrical. If this asymmetrical character were to be applied to the above explanation, it could be argued that the smooth acceptance policy of banks around mortgages
