Housing financialization as a self-sustaining process. Political obstacles to the de-financialization of the Dutch housing market

Abstract

The 2007–2009 financial crisis exposed the risks of housing financialization. Yet the political dynamics shaping post-crisis efforts to de-financialize housing have received surprisingly little analysis. The financialization literature posits that de-financialization policies have been hampered by a policy consensus on the desirability of the pre-crisis status quo. I examine this claim through a detailed analysis of Dutch macroprudential policy reforms, which aimed to mitigate housing-related systemic risks. It finds a fragmented rather than coherent policy community, with the central bank and financial conduct authority pushing for ambitious policies. While they influenced reforms during the housing bust (2008–2013), the government ensured that these financial supervisors would remain peripheral to the future determination of these policies. As the subsequent housing market recovery reduced the urgency to reform, supervisors were unable to impose further de-financialization policies. Housing financialization thus appears self-sustaining, by making mortgage-related policies politically too important for the government to consider a significant empowerment of the actors that might challenge it.

Keywords:

• Housing policy
• de-financialization
• post-crisis reforms
• macroprudential policy
• financial stability

Introduction

Housing scholars have documented how many countries’ housing sectors became increasingly dominated by financial actors, markets, practices, measurements and narratives in the decades preceding the 2007–2009 global financial crisis (Aalbers, 2016; Engelen et al., 2009). Far from a ‘natural’ economic development, this financialization of housing was to an important extent shaped by the state (Aalbers, 2008; Belotti & Arbaci, 2021; Çelik, 2021; Christophers, 2017; Gotham, 2009; Walks & Clifford, 2015). The subsequent financial, economic and housing crises exposed the risks thereof (Jordà et al., 2016), with observers calling on governments to implement thorough reforms of financial regulations and housing policies to mitigate systemic risks linked to real estate (Turner, 2017). The housing literature, however, has generally described post-crisis developments as a reconfiguration rather than a reversal of housing financialization (Aalbers, 2017; Waldron, 2018; Wijburg & Aalbers, 2017). Crucially, scholars show how many states have taken actions that (to varying degrees) have buttressed financialization, including by purchasing and reselling banks’ troubled mortgage portfolios (Byrne, 2016; García-Lamarca, 2021; Waldron, 2018), by rebooting securitization markets (Engelen, 2015; Walks & Clifford, 2015), and by facilitating the rise of institutional corporate landlords in the private rental sector (Beswick et al., 2016, p. 321; Fields & Uffer, 2016; Nethercote 2020; Wijburg et al., 2018).

The literature has been remarkably silent, however, on post-crisis attempts to de-financialize housing, which Wijburg (2021, p. 1277) identifies as a ‘surprisingly under-studied research topic’. Given that the state is a central actor to financialization, a particularly interesting issue is whether public authorities have also pushed for de-financialization policies, and what obstacles they have encountered in this process. But although housing scholars acknowledge that the state consists of separate institutional bodies with potentially conflicting preferences (Aalbers, 2016; Christophers, 2017), many studies treat it as a unified actor that has generally – albeit to varying degrees and in uneven ways – supported financialization, before and after the crisis.

This article evaluates this treatment through an in-depth analysis of macroprudential reform (Baker, 2013). As discussed in the following section, macroprudential policy can be considered a de-financialization strategy as it aims to weaken the destabilizing interaction between mortgage market developments and financial dynamics (Wijburg 2021, p. 1279). It does so by enabling public authorities to impose temporary or permanent system-wide restrictions on lenders (such as tougher requirements for their mortgage exposures) and borrowers (such as loan-to-value and loan-to-income limits) in order to fight systemic risks (IMF et al., 2016), particularly those linked to real estate (Hartmann, 2015). Macroprudential policy thus constitutes a crucial component of one of the three lines of inquiry into de-financialization as identified by Wijburg (2021, p. 1279), namely financial market reforms aimed at dismantling finance-led housing accumulation.

The paper focuses on macroprudential reform in the Netherlands in the period 2008–2020. Scholars have labelled the Dutch housing market as thoroughly financialized (Engelen et al., 2009), with rapidly rising housing prices and mortgage debt levels in the 1990s and 2000s leading to a substantial bust between 2008 and 2013. This makes it a suitable case to study to what extent policymakers have tried to weaken the housing-finance link. Building on an extensive analysis of policy documents and 14 expert interviews, I discuss policy negotiations between key public actors – in particular: the Autoriteit Financiële Markten (AFM; the Financial Conduct Authority), De Nederlandsche Bank (DNB; the central bank and banking supervisor), the government, and the relevant ministries (which execute government policies but also have an autonomous role) – on macroprudential reform.

Far from a unified actor, the analysis shows how the state’s institutional bodies have actually clashed on the desirable direction of reform. In the crisis’ aftermath (2008–2013), DNB and AFM pushed for strict macroprudential rules and asked for the institutional power to determine the rules’ future calibration. Yet while the government and the Ministry of Finance were open to the idea of stricter rules, they vehemently opposed a strong institutional position for these financial supervisors. In institutional terms, these supervisors have thus been side-lined. This, in turn, has shaped macroprudential action during the subsequent housing market ‘recovery’ from 2014 onwards. Although this was economically the right time to implement more ambitious macroprudential constraints, politicians were hesitant to hamper housing access. While calling for stricter measures, DNB and AFM have ultimately failed to influence reforms, reflecting their weak institutional position.

The state’s reluctance to implement de-financialization strategies thus does not necessarily reflect broad internal support for housing financialization, but rather the institutional weakness of those agencies that may push through reforms. This appears to be directly related to the political sensitivity of a financialized mortgage market: with so many voters being directly affected by these policies − 60 percent of all Dutch households have a mortgage (Eurostat, 2022) – the government considered it politically unfeasible to shift policy responsibilities to the more conservative AFM and DNB. Through its impact on macroprudential institutional reforms, housing financialization thus seems to have a self-sustaining character: the mortgage market’s political relevance incentivises the government to suppress de-financialization forces.

The article’s contribution to the housing literature is threefold. First, it answers the call to investigate post-crisis de-financialization attempts (Wijburg, 2021). While it does not study the breadth of potential de-financialization strategies – for example, it does not address rental markets (cf. Fields & Uffer, 2016) – the mortgage market’s centrality to housing financialization warrants thorough analysis. Second, it contributes to the study of ‘financialization through the state’, identified as ‘one of the research frontiers to be pushed in the coming years’ (Aalbers, 2019), by showing the conflicting perspectives within the state on de-financialization strategies. Third, it stresses the inherent political dimension of these policies (cf. Bengtsson, 2015; Jacobs & Pawson, 2015), showing how the salience of financialized housing markets makes de-financialization strategies politically difficult.

This paper proceeds as follows. The next section discusses the housing (de-)financialization literature and its treatment of the state, and it argues that macroprudential reforms may be considered a de-financialization strategy. The subsequent two sections document Dutch institutional and regulatory macroprudential reforms during the housing bust and the subsequent recovery. The final section contains a brief conclusion, highlighting the article’s broader relevance and indicating potential avenues for future research.

The state and housing de-financialization

The ‘housing financialization’ concept has been used to describe a plethora of developments that have profoundly changed housing systems in the past decades (see below). While many studies focus on changes in the Global North (including this article), a thriving literature now also addresses developments in the Global South and peripheries of the Global North (e.g. articles in Aalbers et al., 2020; Çelik, 2021). Scholars have generally been critical of the consequences, unearthing increased economic instability (Waldron, 2018) and social inequalities (Rolnik, 2013).

As ‘financialization’ refers to various developments and has multiple definitions – which partly explains why the financialization literature has recently received critical scrutiny (cf. Christophers, 2015) – it is key to give some conceptual clarity. In this article ‘housing (de-)financialization’ refers to the increasing (decreasing) role of ‘financial actors, markets, practices, measurements and narratives’ in housing sectors (cf. Aalbers, 2017, p. 544). ‘Financialization’ therefore does not solely refer to situations where ‘finance’ is introduced in a domain where it previously played no role whatsoever; after all, mortgage finance has a long history, while the strong rise of mortgaged homeownership is still key to housing financialization (Aalbers, 2017). Similarly, de-financialization is not solely about ‘the negation of any form of finance’, but may also refer to ‘the re-embedding of finance in a more regulated form’ (Wijburg, 2021, p. 1289).

The article also relies on the concept of ‘housing (de-)financialization policies’, which are those public policy measures that aim to increase (decrease) the role of finance in housing. So while Aalbers (2017, p. 551) and Loomans & Kaika (2021, p. 18) have dismissed tighter mortgage lending norms as irrelevant for de-financialization, it would still be warranted to call them examples of de-financialization policies, even if their effectiveness may be limited.

The state’s role in the financialization of housing

The concept of housing (de-)financialization policies is useful given the housing literature’s emphasis of the state’s crucial role. Scholars have emphasized the different strategies and tools with which states have contributed to housing financialization. Its regulatory role obviously features prominently. For example, Fields & Uffer (2016) demonstrate how the deregulation of rent-stabilised housing contributed to the financialization of rental housing in New York, while Gotham (2009) shows that state policies and legal actions facilitated the securitization of mortgage loans (see below). The state also directly operates on the housing market. Belotti (2021) suggests that the state operated as a shareholder to boost and steer equity investments towards affordable housing, both rental and owner-occupied housing (see also Belotti & Arbaci, 2021). Byrne (2016) and García-Lamarca (2021) reveal that states created ad-hoc vehicles to absorb non-performing mortgage loans. Yeşilbağ (2020) and Çelik (2021) describe housing financialization in Turkey as a strategy orchestrated entirely by the state.

The literature also tells us that countries differ in important ways in how housing financialization manifests itself and how governments steer this process (Fernandez & Aalbers, 2016; Schwartz & Seabrooke, 2009). Schwartz & Seabrooke (2009) coin the concept ‘varieties of residential capitalism’ to highlight that there are significant differences in countries’ housing systems. They argue that these variations shape the politics of housing, thereby giving rise to different reform trajectories. So while in some countries, including the USA, the UK, the Netherlands, Spain and Ireland, states strongly pushed mortgage market financialization, in countries such as Germany and France this was less the case. Fernandez & Aalbers (2016, p. 72) concur that housing financialization is ‘uneven in nature, resulting from the interaction of a global pool of capital and national systems of housing and housing finance’. Yet they argue to ‘focus on systemic transformations rather than on the fundamental diversity of national models’, emphasizing that countries have experienced common trajectories towards an increasingly liberalized financial environment.

As mentioned above, housing financialization refers to different developments, and the state’s role obviously depends on the issue. Consider one of the most crucial manifestations of housing financialization: the strong rise of mortgaged homeownership in the decades preceding the global financial crisis. Underpinned by a neoliberal ideology stressing the benefits of private property, many governments introduced policies aimed at increasing homeownership rates, even if this was only possible through ballooning household indebtedness (Fernandez & Aalbers, 2016). Through these asset-based welfare programmes (Doling & Ronald, 2010), households were encouraged to perceive their house not so much as place to live in but as a financial investment, so with its ‘exchange value’ gaining precedence over its ‘use value’ (Aalbers & Christophers, 2014).

Many states also supported another (related) pre-crisis manifestation of housing financialization: securitization, involving lenders selling parts or their mortgage portfolios to other financial actors. Housing scholars have documented how the state not only provided the necessary institutional frameworks, but also how semi-public institutions (particularly in the USA) dominated these markets (Aalbers, 2008; Gotham, 2009; Wainwright, 2009; Walks & Clifford, 2015). But while policymakers celebrated the creation of these mortgage-backed securities as a way to distribute credit risks, increasingly lenders created credit risks to distribute mortgage-backed securities. The rise of mortgaged homeownership and securitization mutually reinforced one another, leading to a ‘credit cycle on steroids’ (Turner, 2017, p. 104), historically unprecedented private debt levels, and rapid house price inflation (Jordà et al., 2016).

While the 2007–2009 crisis exposed the downsides of these policies, housing scholars have documented how governments have responded in ways that were mostly conducive to housing financialization, rather than fundamentally challenge this process; although again states did so to varying degrees and in uneven ways (cf. Aalbers, 2017). One line of research shows how state-backed vehicles attempted to resolve the real-estate crash by buying and reselling banks’ distressed property assets (Byrne, 2016; García-Lamarca, 2021), thereby ‘rebooting the mechanisms that led to the crisis conditions in the first place’ (Waldron, 2018, p. 208). Another line analyses states’ attempts to keep securitization markets afloat. For example, Engelen (2015) shows how public actors helped Dutch banks resist tougher global regulatory requirements, while Walks & Clifford (2015) document how state support underpinned Canadian securitization markets, which paradoxically encouraged risky mortgage lending.

The crisis also boosted rental market financialization (Nethercote, 2020), although this process predated the crisis (cf. Fields & Uffer, 2016). Its most visible manifestation has been the rise of institutional investors – private equity firms, hedge funds, real estate investment trusts (REITs), and publicly listed real estate firms – on ‘private’ rental markets (cf. Beswick et al., 2016). Post-crisis manifestations include REITs buying (rock-bottom priced) property in the crisis aftermath (Waldron, 2018; García-Lamarca, 2021). States have also steered these developments, for example by introducing (tax) legislation conducive to a bigger role for these investors.

Housing scholars generally acknowledge that the state is not a unitary actor but rather ‘a system of interacting institutions’ (Christophers, 2017, p. 63). While studies generally focus on the national level, some also take the supranational (e.g. Engelen, 2015; García-Lamarca, 2021) and local levels (Belotti & Arbaci, 2021) into account. Accounts of the national state’s role in mortgage market financialization – this article’s topic – often pay attention not only to the government and the relevant ministries, but also to regulatory agencies such as banking supervisors, central banks, or consumer protection agencies (cf. Engelen, 2015; Walks & Clifford, 2015). These institutions need not necessarily pursue the same goals: ‘Some state agents actively [.] create the conditions for the financialization of housing [.] while other state agents may try to limit financialization pressures’ (Aalbers, 2017, p. 550).

And yet, particularly with respect to mortgage market financialization, the literature often treats the state as a unified actor of which the different components more or less pursue the same goals. Regulatory agencies such as central banks are generally portrayed as actors that support national governments in their attempts to reconfigure (rather than fundamentally challenge) housing financialization. Explanations include their being part of a (trans)national financial elite that benefits from pre-crisis practices (Engelen, 2015) or their dependence on financialized markets for successful policy implementation (Walter & Wansleben, 2020). But while particular post-crisis actions of financial regulators have certainly boosted rather than undermined financialization – for example central banks’ Quantitative Easing programmes (Fernandez & Aalbers, 2016) – in the domain of mortgage policies they have also pushed for stricter rules (Hartmann, 2015). The state may therefore be more divided than how it is often portrayed in the literature: not all agencies are necessarily on the ‘housing financialization bandwagon’.

Macroprudential regulation as a de-financialization policy

This brings us to the article’s main issue, namely the extent to which state institutions have actively pushed for de-financialization policies, focusing on macroprudential regulation. Hailed by policymakers as the solution to financial markets’ boom-bust patterns, macroprudential policy can be defined as ‘the use of primarily prudential tools to limit systemic risk’ (IMF et al., 2016, p. 4). This was to address a key flaw of pre-crisis policies, namely their focus on financial risks for individual actors (a microprudential approach). During a credit boom these risks appear low, therefore warranting lenient policies, while from a macroprudential perspective this only reinforces the boom and the build-up of systemic risk (Borio, 2018).

As collapsing mortgage markets in the USA and several EU countries were central to the 2007-9 crisis, housing would obviously be a central concern for macroprudential policy (Hartmann, 2015). As research confirmed, financial stability risks ‘have been increasingly linked to real estate lending booms, which are typically followed by deeper recessions and slower recoveries’ (Jordà et al., 2016, p. 107). These systemic risks can be mitigated through temporary or permanent system-wide restrictions on borrowers and lenders (see Table 1). Borrower-based instruments include loan-to-value limits, loan-to-income limits, and amortization requirements. These instruments reduce payment and ‘negative equity’ risks for borrowers, thereby protecting mortgage providers from losses, and limit systemic risks by constraining mortgage credit creation. Lender-based instruments include the imposition of tougher requirements for banks’ mortgage exposures. Such system-wide constraints protect the banking sector against mortgage-related losses, and may also disincentivize mortgage credit creation. Such constraints may apply at all times, but a crucial innovation of macroprudential policy is that policymakers may apply them in a countercyclical manner, so in response to rising systemic risks (Borio, 2018).

Table 1. Lender-based and borrower-based macroprudential instruments.

Type	Macroprudential tools	Explanation
Lender-based instruments	Countercyclical capital buffer	An additional buffer on top of normal requirements activated when credit grows too rapidly
	Higher capital requirements for lenders’ mortgage exposures	Extra stringent capital requirements for lenders’ mortgage exposures, for example higher ‘risk weights’
Borrower-based instruments	Loan-to-value limit	Limits the amount people can borrow relative to the value of the house
	Loan-to-income limit	Limits the amount people can borrow relative to their income
	Debt-service-to-income limit	Limits the amount people may spend on debt service (interest and principal repayment) relative to their income
	Amortization requirements	A requirement for borrowers to pay-off the mortgage over time, thereby incrementally decreasing their leverage

In the EU, lender-based macroprudential instruments have been embedded in the Capital Requirements Directive (CRD) IV and the Capital Requirements Regulation (CRR). These include a countercyclical capital buffer and additional capital requirements for banks’ mortgage exposures. The European Central Bank (ECB) has limited macroprudential powers, being allowed to ‘top up’ national measures. It also hosts the European Systemic Risk Board, an agency tasked with monitoring systemic risks and issuing warnings and recommendations (on a comply-or-explain basis) to national authorities. The ECB has made clear, however, that it does not take financial stability considerations into account in its monetary policies (including Quantitative Easing). Despite this EU dimension, macroprudential competences – particularly those linked to real-estate – are mostly located at the national level, reflecting their political sensitivity (Stellinga, 2021).

Macroprudential policy may be considered a de-financialization strategy because it aims for ‘the re-embedding of finance in a more regulated form’ (Wijburg, 2021, p. 1289). Yet like de-financialization policies more generally (Wijburg, 2021), macroprudential policy is an under-studied research topic in the housing literature (but cf. Belfrage & Kallifatides, 2018). This contrasts with (what is often called) the International Political Economy (IPE)-literature on financial regulation, which has studied the role of ideas, interests, and institutions in shaping post-crisis macroprudential reforms (Baker, 2013; Stellinga, 2020; Thiemann, 2019).

A crucial contribution of IPE-studies is their emphasis on the political conflicts that have characterized macroprudential reforms. Notwithstanding a broad consensus on the desirability of implementing macroprudential frameworks (Baker, 2013), key actors often vehemently disagreed on the nature of reforms, both at the international and national level (Thiemann, 2019). Crucially, this literature portrays financial regulatory agencies – particularly central banks – as having been at the forefront of macroprudential reforms, which have often resulted in a strengthening of their powers (Baker, 2013). This buttresses the case to critically assess the housing literature’s apparent treatment of the state as a unitary actor.

This literature, however, has mostly studied political dynamics shaping macroprudential reforms in general, rather than related to housing. While IPE-scholars have emphasized these measures’ political sensitivity due to their impact on mortgage credit conditions (Fuller, 2019; Kohl, 2021; Stellinga, 2021), we know little about how this has shaped the institutional design of countries’ macroprudential frameworks, let alone how this has affected their actual functioning (Thiemann and Stellinga, 2022). Overview studies show that countries differ in the distribution of macroprudential responsibilities between ministries and financial supervisors, particularly with respect to housing issues (Edge & Liang, 2019, pp. 16–17), suggesting that countries have responded differently to these political sensitivities. Similarly, policy evaluations show that some countries more successfully mitigate systemic risks pertaining to post-crisis housing booms than others (e.g. European Systemic Risk Board, 2022). The IPE-literature, however, is mostly silent as to why this may be the case.

Put differently, we lack an answer as to what determines a country’s success in implementing housing de-financialization policies. As the Dutch case presented below suggests, the degree of pre-crisis housing financialization may play a key role here: housing financialization increases the political sensitivity of mortgage-related policies, making it politically difficult to empower those public actors willing to implement de-financialization policies. While such policies have certainly been implemented between 2008 and 2013, notably through tougher loan-to-value and loan-to-income limits, financial supervisors lacked the powers to implement even tougher measures when the housing market ‘recovered’ from 2014 onwards. While the central bank initiated some macroprudential action in 2019, it lamented the lack of tools to substantially cool down the boom. In the 2007–2009 crisis aftermath, politicians had been reluctant to give these supervisors a strong institutional position in the determination of these rules, reflecting these measures’ high political salience. Despite subsequent calls to strengthen supervisors’ position, no meaningful institutional change was implemented after 2013.

The bust (2008–2013)

In the crisis aftermath, average Dutch housing prices dropped by 20%, resulting in over one-million households having an ‘underwater’ mortgage (Commissie Huizenprijzen, 2013). After years of debate, instigated by the Dutch Financial Conduct Authority (AFM), the government adopted tougher borrower-based instruments. Crucially, however, it refused to structurally empower the AFM and De Nederlandsche Bank (DNB) on these issues, reflecting the political sensitivity of constraining mortgage credit in a financialized housing market (see Table 2).

Table 2. Main conflicts during bust period [2008–2013].

	Borrower-based limits (regulation)	Borrower-based limits (institutional)	Power of Financial Stability Committee (institutional)
The Dutch central bank (DNB)	Prefers ambitious borrower-based limits, but slow implementation to limit procyclical effects of tougher rules	Prefers significant formal influence over borrower-based limits	Prefers a strong Financial Stability Committee, with the Ministry of Finance on board to ensure political commitment
The Financial Conduct Authority (AFM)	Prefers ambitious borrower-based limits, and quick implementation during political window of opportunity	Prefers full responsibility for borrower-based limits or significant influence	Prefers strong institutional position for Financial Stability Committee
Ministry of Finance	Prefers significant borrower-based limits, but slow implementation to limit procyclical effects and to ensure political consensus	Prefers limited influence of supervisors over borrower-based limits	Prefers a weak Financial Stability Committee and an observer status for the Ministry of Finance
Political parties (in government coalition)	Rutte I (2010-2012) coalition parties prefer modest borrower-based limits. In 2012 several parties with more reform ambition implement an ad hoc ‘Spring agreement’ with Rutte I coalition parties	Political parties generally oppose a strong role for financial supervisors in determining borrower-based limits, either preferring self-regulation (on the right) or government control (on the left). Rutte I coalition maintains self-regulation; Spring agreement implements political control	Political parties generally oppose a strong institutional position for the DNB in the Financial Stability Committee, considering it has failed during the financial crisis

The post-crisis reforms of borrower-based instruments

Dutch pre-crisis policies had mostly stimulated mortgaged homeownership, including through generous tax benefits (Priemus, 2010). A self-regulation code specified loan-to-income limits, yet in practice many banks deviated from these norms (DNB & AFM, 2009). Moreover, the code did not contain a loan-to-value limit. In 2008, an average new mortgage had a loan-to-value ratio of 120%, while tax benefits on mortgage payments discouraged amortization. Mortgaged homeownership therefore expanded rapidly: in 1990 the total level of mortgage debt as a percentage of GDP was less than 40%; in 2008, it was 110%. In the same period the average price of a house more than doubled (Commissie Huizenprijzen, 2013). A significant share of the mortgages (about 33 percent) was securitized, although two thirds of these mortgage-backed securities remained on banks’ balance sheets (DNB, 2015). This suggests that they were mostly developed as a funding strategy and to reduce regulatory requirements. DNB and the Ministry of Finance had actively supported these securitization practices (cf. Engelen et al. 2009).

The AFM – from 2006 onwards tasked with supervising mortgage lending – played a leading role in post-crisis reforms. In April 2009, AFM director Hans Hoogervorst dropped the bomb by dismissing the self-regulation code. He argued for stricter loan-to-income norms and a loan-to-value limit of 100%, indicating that the AFM would present new guidelines within a couple of months (Ministry of Finance, 2009). While the AFM reasoned from a consumer protection-perspective, it also considered pre-crisis norms undesirable from a financial stability (macroprudential) perspective (Interview 7).

The AFM had hoped to receive public backing by DNB, yet, according to a former AFM official, DNB was hesitant, as it worried that ‘this would influence the housing market too much, and therefore also the economy, because it was really a stressful time’ (Interview 7).¹ The Ministry of Finance was also unhappy about the AFM’s timing, according to a former Ministry official: ‘this was at a time that a huge housing market crisis started, so there was a debate on the balance between reducing risks and protecting the economy from being harmed by more stringent rules’ (Interview 5). Moreover, the AFM’s proposal was politically controversial: a parliamentary motion calling upon the government not to embrace the 100% loan-to-value limit received support from 145 of the 150 parliamentarians (Kamerstukken II, 2009). The AFM, however, favored policy reform even if it would have negative macroeconomic effects: ‘Hoogervorst considered it to be the ideal time to reform, because you have to strike the hammer when the iron is hot’ (Interview 7).

While the AFM appeared to be alone in its quest for stringent borrower-based measures, there was actually broad support within the Ministry for a loan-to-value limit and a stricter enforcement of loan-to-income norms. According to a former official, ‘we thought that the AFM was right: there were loopholes in the code of conduct, it did not work properly. So we broadly agreed that things needed to be done’ (Interview 5). Another former official concurs: ‘before the crisis, mortgage lending practices were getting out of control with very high loan-to-income ratios, and nobody even paid attention to loan-to-value ratios because people believed that prices could only go up’ (Interview 3). To generate consensus within government, the Ministry circulated policy notes showing that existing norms were undesirable from a macro-economic and a consumer protection perspective. It also aimed to convince parliamentarians by emphasizing the systemic and personal financial risks of Dutch mortgage lending practices (Interview 5).

But because of their political and economic sensitivity, the implementation of reforms went much slower than the AFM had wanted. In 2010 the Rutte I-government backed away from tough public rules by giving the banking sector a chance to modify its code of conduct, much to the chagrin of the AFM (Interview 7). According to a banking sector representative, however, it was ‘self-regulation with a gun to our head’ (Interview 6). The new code introduced a loan-to-value limit of 110%, a requirement that at least 50% of the mortgage would be subject to amortization, and less room for banks to deviate from the loan-to-income guidelines (Interview 5).

A political crisis in Spring 2012 – the Rutte I-government collapsed over austerity measures – paved the way for a political coalition that implemented more significant changes. The coalition set the loan-to-value limit at 106%, which would be reduced by 1% each year until reaching 100% in 2018. It also ended self-regulation: the loan-to-income and loan-to-value measures were incorporated in a public regulation. Additionally, it made tax benefits for new mortgages conditional upon loan amortization. This significantly reduced the demand for interest-only mortgages, thereby ensuring that borrowers’ leverage would decrease over time.

These reforms were the result of a careful process of political consensus-building by the Ministry of Finance, itself being pushed in this direction by the AFM and (more in the background) DNB. The housing market crisis had introduced a political window of opportunity. A former Ministry official argues that ‘in times of crisis there is a higher sense of urgency for these kinds of reforms, while in good times nobody sees the necessity to do so’ (Interview 5). Former AFM board member Kockelkoren concurs: ‘when the sun shines [.] we don’t see the necessity to repair the roof. And so only when it pours people feel and understand that we need to do so’ (cited in: Mulder, 2014, p. 9). Still, the housing market’s macroeconomic relevance made policymakers wary of bold steps, implying a preference for incremental reforms. A slow reduction of the loan-to-value limit was meant to limit the impact on the housing market: ‘we decided to do it very slowly and not go to 100% at once, because that would increase the risk of further negative effects on the housing market. So we won’t do that, we will do a 1% reduction each year’ (Interview 5).

The institutional side-lining of financial supervisors

While these reforms embedded financial stability considerations in housing policy, the government simultaneously ensured that AFM and DNB did not obtain a strong institutional position to influence the future calibration of these borrower-based measures. During the debates between 2008–2012, the AFM had appeared willing to determine these measures, a move pre-empted by the Minister of Finance De Jager:

where exactly is the border between independent supervision and policymaking? In this domain I have taken the position that these types of rules should be determined by us, by me as the Minister and therefore also by parliament. If the AFM were to do this, there would be no possibility for parliamentary influence (quoted in: Tweede Kamer, 2010, p. 14).

A former AFM official remembers that the government was hesitant to delegate substantial responsibilities to the AFM: ‘when we said that the self-regulation code was not stringent enough, politicians were eager to shift to public regulation, to prevent that a very independent public agency such as the AFM would become leading on such a politically salient topic’ (Interview 7).

The same applied to DNB. A Ministry official recalls that politicians and the relevant ministries were quite hesitant to give DNB a strong role: ‘Parliament considered the DNB to have failed during the crisis [.]. And we considered DNB to have an overly narrow perspective, looking only at banks’ balance sheets’ (Interview 10). Yet DNB itself was also hesitant to obtain a stronger role:

we asked ourselves: “given that these issues are so politically controversial, do we even want responsibility for these tools?” When the Governing Board discussed this at that time, I believe there was disagreement on whether or not DNB should actually aim to obtain these instruments (Interview 1).

Still, several things made DNB worry that the institutional set-up would leave too little room for macroprudential considerations in the future determination of borrower-based measures. The norms were formally part of a section in the law on consumer protection, and within the Ministry of Finance the ‘consumer protection’ department rather than the ‘financial stability’ department was responsible for them (Interview 5). Additionally, the task to do the underlying calculations for what constitutes ‘prudent’ loan-to-income norms was delegated to an agency (the National Institute for Family Finance Information) that, in the words of a representative, ‘only looks at the affordability for the consumer and doesn’t do anything with macro-economic issues’ (Interview 11). Moreover, from 2013 onwards the Ministry of the Interior – traditionally very much focused on ensuring housing access (Interview 9) – obtained a stronger role on housing market issues, implying that the government would be increasingly hesitant to introduce more stringent borrower-based tools.

To ensure that the macroprudential perspective would be taken into account in the (future) determination of these tools, in June 2013 DNB asked for a formal right of recommendation in the decision-making process (DNB, 2013, p. 10). A DNB official recalls:

The position of DNB was that [loan-to-income and loan-to-value limits] are important macroprudential instruments, and the way in which decision-making is organized doesn’t do justice to that. [.] So for the loan-to-income limit they only look at households’ incomes and expenses, but it does not address the macro-financial situation, or the build-up of debt, or direct effects on the housing market. And for the loan-to-value limit the same applies. And so we argued that the macroprudential side of policymaking should have more traction and more influence over these ratio’s (Interview 1).

The government responded negatively, however, indicating that DNB had ample room to advice on these measures, and that a formal right of recommendation was unnecessary (Ministry of Finance, 2014, p. 1). Instead, the DNB and AFM would be included in the yearly consultation of the National Institute for Family Finance Information on new loan-to-income calculations (a consultation procedure that includes several other stakeholders). For the loan-to-value limit, the AFM and DNB could issue advice on their own initiative.

Financial supervisors’ weak position was also reflected in the new Financial Stability Committee, set up in November 2012 and consisting of representatives from DNB, AFM and the Ministry of Finance. The Committee’s task was to signal financial stability risks and issue recommendations, yet two facets seriously weakened these recommendations’ potential power. First, they had a non-binding nature. Second, while the Ministry of Finance was formally a committee member, it would not participate in the decision-making process.

A Ministry official recalls that DNB had preferred a different set-up:

DNB of course wanted the Committee to have strength, so it insisted on having the Ministry of Finance on board […]. It wanted to ensure that the Ministry would feel bound by the outcome of the deliberations. But we said that this is not compatible with the Minister’s accountability to Parliament. You cannot have public servants in such a committee taking very intrusive measures (Interview 10).

The Ministry of Finance (2012a, p. 5) therefore made crystal clear that ‘the Minister of Finance must always be free to make his own decisions on policy matters’. By abstaining from the decision-making process, the government ensured that it would not be implicated by the Committee’s recommendations, which obviously could also be directed at the politically sensitive borrower-based measures (Interview 10). The Committee’s non-binding recommendations would therefore de facto come from DNB and AFM.

While the loan-to-value and loan-to-income limits were thus out of reach, DNB did gain control over lender-based macroprudential measures contained in the EU’s banking rules. Targeting banks, these instruments’ impact on borrowers is more indirect, making them politically less controversial. Still, the government deemed it necessary that DNB would coordinate its actions with the Minister of Finance before activating tools, especially for those included in the EU Capital Requirements Regulation. This was not only because these tools involved a deviation from the Single Rulebook, which in theory could be blocked by the EU Council (Interview 1), but also because they affected the national real estate market (Ministry of Finance, 2013, p. 2). This once again shows the government’s determination to maintain control over macroprudential measures influencing the housing market.

Interim conclusion (I): political obstacles to empower financial supervisors

Returning to the issue of the state’s position on de-financialization policies, the analysis suggests that its different institutional bodies are much more divided than the housing financialization literature appears to suggests. Crucially, not only financial supervisors but also the Ministry of Finance pushed for tougher borrower-based limits, although it took considerable time and effort to generate consensus within government and parliament. Still, the political salience and macroeconomic relevance of a financialized housing market implied that only incremental steps were acceptable.

More fundamental for the future implementation of de-financialization policies, however, was the fact that the government institutionally side-lined financial supervisors, specifically on the tools which directly constrain households’ borrowing-capacity. Such tools not only affect first-time buyers, but also (through their impact on demand and prices) homeowners’ wealth. This makes these tools politically very sensitive, and particularly so in the Netherlands, given that around 60 percent of all households have a mortgage, considerably higher than the EU average of 27 percent (Eurostat, 2022; cf. Fuller, 2019; Kohl, 2021). While financial supervisors’ weak institutional position did not obstruct post-crisis reforms, it did hamper steps during the boom period.

The boom (2014–2020)

Post-crisis reforms were ambiguous. Financial supervisors’ macroprudential arguments played a meaningful role in the reform of the loan-to-value and loan-to-income limits, but these actors were simultaneously kept at a distance from the future determination of these limits. As the bust turned into a boom post-2014, the political momentum for further changes waned, even though from a macroprudential perspective this was the ideal time to implement more fundamental reforms. Lacking the institutional power to counter political inertia, however, financial supervisors failed to push through further de-financialization policies (see Table 3).

Table 3. Main conflicts during recovery and boom period [2014–2020].

	Loan-to-value limit (regulation)	Loan-to-income limit (regulation)	Lender-based measure (regulation)	Power of Financial Stability Committee (institutional)
The Dutch central bank (DNB)	Prefers further reduction of loan-to-value limit to 90%, in incremental steps to ensure political consensus	Prefers a different calibration method to limit procyclical effects. Supports AFM	Activation of a lender-based macroprudential instrument as a last resort, but prefers the modification of borrower-based instruments	Considers comply-or-explain advisory powers desirable but not absolutely necessary. Prefers more formal legal position for the Committee
The Financial Conduct Authority (AFM)	Supports DNB in its quest for a lower loan-to-value limit	Prefers a different calibration method for the loan-to-income limit to limit financial risks for borrowers	No explicit position on this issue	No explicit position on this issue
Ministry of Finance / Ministry of the Interior	Consider further reductions of the loan-to-value limit undesirable, and unnecessary given previous steps	Consider the calibration of the loan-to-income limit to strike an adequate balance between housing access and consumer protection	Prefer the activation of a lender-based instrument over the modification of borrower-based instruments	Strongly oppose comply-or-explain advisory powers. A more formal legal position for the Committee is unnecessary but acceptable
Political parties (in government coalition)	Rutte II (2012–2017) coalition parties disagree on lowering the loan-to-value limit. Rutte III (2017–2021) coalition parties agree on no further reduction	Coalition parties in Rutte II and III consider the calibration of the loan-to-income limit to strike an adequate balance between housing access and consumer protection	No explicit position on this issue	Coalition parties in Rutte II and III oppose comply-or-explain advisory powers

Regulatory inaction on borrower-based limits

As banks and the housing market recovered from the bust, DNB emerged as the leading actor in calling for further reforms. To be sure, it did not challenge all pre-crisis policies that had contributed to the housing boom. For example, it remained generally positive about mortgage securitization, consistently arguing that ‘if done in a responsible manner, securitization has a positive effect on banks’ funding, capital position and risk profile’ (DNB, 2010; cf. Engelen, 2015 for a critical take). Yet especially on borrower-based limits, DNB called for tougher rules. It now publicly argued for a stricter loan-to-value limit, indicating that it should continue its decline after reaching 100% in 2018. It considered Dutch households’ high leverage a major threat to both financial and economic stability: ‘as 1.3 million households had a mortgage that was under water, it was obvious that this was a relevant fact from a macroeconomic point of view’ (Interview 1).

Domestically, DNB received much support from the AFM. A former Ministry of the Interior official notes that DNB and AFM teamed-up to increase their influence, ‘with the AFM supporting DNB’s call for a lower loan-to-value limit and DNB in return supporting the AFM calls for a stricter loan-to-income limit’ (Interview 4). International pressure came among others from the International Monetary Fund (IMF, 2014, p. 19), which argued that beyond 2018 ‘the authorities should aim for a faster pace of LTV [loan-to-value] reduction, eventually reaching 80 percent’.

In 2013 the Ministry of Finance (2013, p. 11) seemed open to the idea of a further reduction, indicating that ‘if the housing market recovers robustly, further proposals will be made about the ultimate loan-to-value ratio and the steps towards this level after 2018’. Yet this idea was politically controversial. The Rutte II-government (2012-2017) consisted of the left-leaning PvdA-party and the right-wing VVD-party, and the latter was skeptical about a further future reduction of the loan-to-value (Interview 3). VVD Member of Parliament De Vries (2014) gave a clear signal when she called on the government to refrain from committing to a reduction of the limit beyond 2018.

Policy officials within the Ministry of Finance were also not convinced. They considered that the incentive to amortize mortgage loans (a precondition for tax benefits on mortgage payments) reduced the need for a lower loan-to-value limit, while the welfare costs of a lower limit were high (Interview 9). The Ministry of the Interior focused on accessibility, and a lower limit would hamper first-time buyers with little savings, pushing them to the expensive private rental market, which would further reduce their ability to save money (Interview 3).

To challenge the DNB’s perspective, the Ministry of the Interior requested the Bureau for Economic Policy Analysis – an observer in the Financial Stability Committee that had been critical of loan-to-value limit reductions (Interview 1) – to investigate the macro-economic consequences of a further reduction (Interview 4). Unsurprisingly, its study (CPB, 2015) argued that a further reduction would have negative economic consequences, while financial stability benefits would be negligible. Instead, the Bureau suggested ‘other measures, particularly eliminating tax benefits’ (Interview 12). According to a Ministry of Finance official, a further loan-to-value reduction would have been unlikely even without this study, but it was still an important blow to DNB’s narrative (Interview 9).

Expecting limited political support, the Financial Stability Committee’s (2015) recommendation to continue the annual 1% reduction until the loan-to-value limit reached 90% in 2028 – de facto coming from the AFM and DNB, as the Ministry of Finance abstains from such recommendations – was explicitly directed at the next government. The Rutte II-government thus deferred this issue, stating that existing measures were sufficient, using the Bureau for Economic Policy Analysis’ study to flag that even macroeconomic experts disagreed (Ministry of Finance, 2015). The subsequent Rutte III-government (2017–2021) did not do anything on the loan-to-value limit, keeping it at 100% from 2018 onwards, ‘to ensure that the housing market’s accessibility for first-time buyers is not unnecessarily hampered’ (Rutte III Coalition Agreement, 2017).

Financial supervisors have also had moderate success in shaping the loan-to-income limits. The AFM was concerned about changes in the calculation method that could expand households’ ability to borrow (Interview 11). DNB mainly worried about the method’s procyclical effects: as the economy booms and interest rates fall, people can borrow more, which is problematic from a macroprudential point of view. The AFM and DNB repeatedly flagged these problems, but did not gain much support inside the relevant ministries. A former Ministry of Finance official indicates that the ministries barely consider the loan-to-income limit to be a macroprudential instrument, and mainly look at whether consumers can afford the mortgage (Interview 9). While the method for calculating the loan-to-income limits was somewhat modified in 2016 to alleviate DNB’s concerns, there were no modifications to limit the impact of falling interest rates on consumers’ borrowing capacity. As such, over the years the lower interest rates – mainly caused by the European Central Bank’s expansive monetary policies – resulted in much more lenient loan-to-income limits (Ministry of the Interior, 2020).

The activation of a lender-based instrument

From 2018 onwards, DNB got increasingly worried about systemic risks pertaining to rapidly rising housing prices and borrowers’ increased risk taking. With no action expected from the government on the borrower-based limits, DNB turned to its own macroprudential toolkit, containing only lender-based instruments. The countercyclical capital buffer had been available from 2016 onwards, but it was deemed inadequate for this particular problem, as it targets all credit exposures, including corporate credit. Moreover, the credit-gap-to-gdp, the most prominent indicator of excessive credit growth, had been significantly below zero ever since the financial crisis. ‘The cycle wasn’t really strong, so it was difficult to use the countercyclical buffer’, according to a former DNB official (Interview 14). As such, DNB consistently kept the buffer at 0%.

DNB therefore focused on banks’ risk-weights for mortgages exposures, which DNB considered to be problematic from a macroprudential perspective, as they decrease when house prices rise (Interview 1). It was encouraged in taking action by the European Systemic Risk Board’s (2019) recommendation to activate capital-based measures to strengthen banks’ resilience. After a formal check with the Ministry of Finance and a public consultation, DNB announced in October 2019 that it would activate Article 458 of the EU Capital Requirements Regulation to induce an increase in banks’ mortgage risk-weights. While postponed due to the COVID 19-crisis, the measure entered into force in January 2022.

Yet a former DNB official argues that this measure aims to increase banks’ resilience, not to stop the housing market boom: ‘Capital-based measures have a more indirect effect on housing markets. So it is more about increasing the resilience of the banking system than affecting the market. So their effect is different, and in a way, they are less effective’ (Interview 14). DNB considers borrowing-based limits much more effective:

if you want to cool it [the housing market] down it is much more effective to do that through loan-to-value or loan-to-income ratios. And changing risk weights doesn’t help much in this regard, because it has a very small effect on the pricing. [.] But the debate on borrowing limits was over, there was limited willingness in The Hague to do something about this (Interview 1).

The Ministry agreed to the increase of banks’ capital requirements. As a Ministry official explains, ‘this measure was also less politically sensitive, because it less directly affects the housing market and is more focused on banks’ (Interview 8). Yet it responded negatively to the European Systemic Risk Board’s recommendation to tighten borrower-based measures: ‘invoking Article 458 to change risk-weights is worlds apart from tightening loan-to-value and loan-to-income measures, politically speaking [.]. Loan-to-value and loan-to-income measures are still very sensitive within the government’ (Interview 8).

Institutional inertia

Financial supervisors’ limited impact on the borrower-based tools after 2013 is a reflection of their weak institutional position. While according to respondents there have been informal deliberations on giving the DNB and AFM a stronger say and depoliticizing the determination of these limits, this was not considered a viable way forward (Interview 1, 5, 7). A former Ministry of Finance official argues that ‘on the one hand the benefit is to make this less political, but on the other hand… the issues are inherently political, so this would create a lot of pressure on the supervisor and therefore also on the minister of Finance, who is ultimately responsible for the supervisors’ conduct’ (Interview 5). A DNB official maintains that Dutch politicians would never agree on policy delegation: ‘if you would suggest to the Minister of Finance to let DNB determine the loan-to-income and loan-to-value limits, or better yet if you suggest that in parliament, well everybody would say “no way, they are way too important”. While in many countries it is fully accepted that the supervisor or central bank determines these limits’ (Interview 1).

Similarly, debates on strengthening the Financial Stability Committee’s position have had limited effects. Over the years international organizations have repeatedly advised to increase the its powers. The Financial Stability Board suggested to give the Committee the power to make formal recommendations to the government on the use of the loan-to-value and loan-to-income limits and even argued that in ‘the longer term, the authorities should consider reallocating the powers for setting these limits to the FSC’ (FSB, 2014, p. 9). Unsurprisingly, these suggestions have been ignored by the government (Interviews 1; 8).

Yet even the more modest suggestions by the International Monetary Fund (IMF, 2017), the Financial Stability Board (FSB, 2014) and the European Systemic Risk Board (ESRB, 2019) to give the Committee’s recommendations a more binding character through an ‘act or explain’-mechanism have been dismissed. As the Committee’s recommendations de facto come from the AFM and DNB, the government is reluctant to give these more traction. A Ministry official indicates that it takes these international recommendations with a grain of salt: ‘to put it bluntly, when the FSB and the IMF recommend something, de facto DNB is speaking’ (Interview 10). Yet the opposition to an ‘act or explain’-mechanism is also a matter of principle, officials emphasize, referring to the democratic principle that a Minister only has an accountability relationship with parliament (Interview 10; Interview 13).

There was only political willingness for a more limited step, namely embedding the Committee (originally set up through a ministerial decree) in primary law. DNB had pushed this as it could open up the possibility for a stronger future role (Interview 1), given that a strong legal position is a precondition for allocating instruments to the Committee (Interview 10). A Ministry official stresses, however, that a different legal position will not change the Committee’s actual functioning (Interview 13).

Interim conclusion (II): political obstacles to countercyclical action

As in the post-crisis period, the Dutch state was also internally divided on macroprudential (de-financialization) policies during the boom period. But whereas the housing market crisis created the political momentum for reforms, the subsequent recovery made it politically unattractive to constrain mortgage credit. Reducing households’ ability to borrow when prices rise is a risky political strategy, while price rises reduced the number of households with ‘negative equity’. The public actors that propagated a macroprudential perspective had been institutionally side-lined, implying that they had no power to push through further de-financialization policies. As such, the pre-crisis financialization of Dutch mortgage markets seems to have had a self-sustaining aspect: the high political salience of the owner-occupied segment has pre-empted the delegation of policy responsibilities to the actors that would have been willing to introduce more ambitious de-financialization measures.

Conclusion

This article has described how the Dutch state is internally divided on de-financialization policies, with financial supervisors having pushed for reforms that often proved politically unfeasible. The government’s resistance to empower financial supervisors on borrower-based measures is linked to the political sensitivity of these tools, which seems especially high in the Netherlands given its thoroughly financialized housing market. While the crisis certainly created the momentum to implement de-financialization policies, the subsequent recovery erased the political will to go further. Pre-crisis financialization thus seems to have created the very political obstacles to de-financialization policies.

The argument that housing financialization has a self-sustaining quality resonates with broader debates in the housing literature. This includes political science approaches stressing the path dependence in housing systems. Scholars have demonstrated how housing policies decided at one point in time constrain policy choices at a later point in time, thereby introducing a self-reinforcing mechanism (cf. Bengtsson et al., 2017). The argument also chimes with approaches emphasizing how mortgage market financialization is part of a broader transformation towards an asset-based welfare state, in which ‘individuals accept greater responsibility for their own welfare needs by investing in financial products and property assets which augment in value over time’ (Doling & Ronald, 2010, p. 165). This literature shows that not only do public agencies have deep stakes in this state of affairs, but homeowners as well (Malpass, 2008), again creating political obstacles to reform. The present study contributes to these debates by highlighting the struggles within the state, showing how ‘de-financialization forces’ attempt (but in this case fail) to obtain an institutional position which would allow them to counter the strong interdependence between finance and housing. It suggests that scholars not only need to pay attention to how financialization is ‘contested from within society and the economy’ (Wijburg, 2021, p. 1276), but also from within the state.

Evaluations of macroprudential policy show that many countries have been quite active since the global crisis. Yet the Bank for International Settlements finds that these actions ‘have not necessarily prevented the emergence of familiar signs of financial imbalances – for instance, in the form of outsize credit growth’ (Borio, 2018, p. 5). Edge and Liang (2019, p. 13) find that while ‘the use of macroprudential tools has been growing and now is substantial, the actual frequency of change for most tools is very limited, suggesting most are not used in a time-varying way to address cyclical vulnerabilities’. Especially borrower-based tools – which have ‘proved to have a larger and more discernible effect than, say, [.] the countercyclical capital buffer’ (Borio, 2018, p. 5) – appear to be difficult to tighten, as the Dutch case also shows. Policy evaluations generally link this to political economy reasons: ‘really “biting” instruments are “socially charged”’ (Hartmann, 2015, p. 78).

Although the article has focused on de-financialization policies for the owner-occupied segment, clearly a promising line of research would be to investigate public agencies’ conflicts over rental market financialization. Moreover, as the housing financialization literature – like the present study – focuses mainly on changes in the Global North (Aalbers et al., 2020), it would be interesting to assess public actors’ attempts to push housing de-financialization policies in emerging economies. Although housing policies may be politically controversial everywhere, it would be fruitful to investigate whether we see more willingness to constrain excessive mortgage credit in those countries where mortgaged homeownership is less prominent. Studies suggest a greater willingness in emerging economies to empower central banks on loan-to-value limits (Villar, 2017), yet it is an open question whether this also implies limited political control and meaningful action.

What does this mean for the fate of strategies aiming to de-financialize the Dutch housing market? In response to the COVID 19-crisis, the government introduced several policies aimed at supporting the housing market, including the removal of the 2% stamp duty on housing transactions for first-time buyers, thereby contributing to a further rise of house prices. The Rutte IV-government, installed in January 2022, considers housing construction to be the main solution to the boom. While several steps also appear to counter excessive overbidding – for example the stepwise elimination of tax exemptions for family donations used for house purchases – borrower-based limits have once again remained untouched, with the government explicitly stating that the loan-to-value ratio will not be modified while it is in office (Rutte IV Coalition Agreement, p. 14). The political reform logic is clear: only when it rains there is ambition to repair the roof (Mulder, 2014). Whether it will start to rain again any time soon, remains to be seen.

Politicians generally consider the strong price rises as caused solely by low interest rates and insufficient supply, and have a hard time seeing lenient borrower-based norms and subsidies for buyers as part of the problem. If anything, many politicians identify more lenient norms and additional subsidies as the solution to accessibility problems, even if on aggregate this is counterproductive: first-time buyers ultimately do not benefit from this if it only leads to ever increasing housing prices. DNB’s story emphasizing systemic risks appears to lack traction in parliament. One way forward may be to reframe the argument for strict policies also being about house price stability: emphasizing the benefits of stable housing prices could create more political support for measures that hurt in the short run, but by delinking financial dynamics and housing dynamics would over the long term better safeguard the various public interests that are at stake. Yet the analysis suggest that an integral approach, which also encompasses rental markets, is essential for housing de-financialization. Post-crisis reforms in this domain have mainly facilitated the financialization of ‘private’ rental markets, thereby solving one problem by creating another. Investigating how social housing provision may play a role in de-financialization, therefore, also seems essential (Van Gent & Hochstenbach, 2020).

List of interviews

Table

Interviews	Organization	Date
1	DNB 1	24-03-2021
2	Dutch Home Owners Association (VEH)	25-03-2021
3	Ministry of Finance 1	30-03-2021
4	Ministry of the Interior	30-03-2021
5	Ministry of Finance 2	01-04-2021
6	The Dutch Banking Association (NVB)	02-04-2021
7	AFM	07-04-2021
8	Ministry of Finance 3	14-04-2021
9	Ministry of Finance 4	14-04-2021
10	Ministry of Finance 5	30-04-2021
11	National Institute for Family Finance Information	08-05-2021
12	CPB	20-05-2021
13	Ministry of Finance 6	01-06-2021
14	DNB 2	07-06-2021

Acknowledgements

I would like to thank Matthias Thiemann (Sciences Po Centre d’études européennes) for inviting me to join his research project on post-crisis macroprudential reforms, which allowed me to do the empirical research for this article. I am really grateful to Gertjan Wijburg for his comments and suggestions on an earlier draft. I also want to thank all my respondents for their willingness to help me with this research. Finally, I would like to thank the editors and five reviewers for their suggestions during the review process.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Bart Stellinga

Bart Stellinga is a senior research fellow at the WRR. He is also a lecturer at the Institute for Interdisciplinary Studies of the University of Amsterdam (UvA). He studied political science (MSc) and philosophy (MA) at the UvA. He obtained his doctorate at the UvA in 2018 for his thesis ‘The financial valuation crisis’, addressing the politics of post-crisis financial reform in the European Union. He has published in leading political science journals, including Journal of Common Market Studies, Review of International Political Economy, Regulation & Governance, and Business & Politics. At the WRR, dr. Stellinga has contributed to publications on the privatization of public sector activities, food policy, financial supervision, and monetary policy.

Notes

1 All quotes from the interviews and from Dutch sources are the author’s translation.