Balancing Accumulation and Affordability: How Dutch Housing Politics Moved from Private-Rental Liberalization to Regulation

ABSTRACT This paper answers the question why the Dutch state has gone from vigorously stimulating private-rental growth and liberalization to actively restricting the tenure. Answering this question is important in understanding an emergent wave of more restrictive, or even “post-neoliberal” housing policies across countries. This paper presents an analysis of the changing private-rental politics in the period following the Great Financial Crisis, combined with a quantitative study of renters’ housing outcomes. The central argument is that policies promoting private-rental growth and liberalization and the subsequent turn restrictive policies are both outcomes of the state seeking to balance the property-led accumulation with middle-class residential demands. Supportive policies were the result of a presumed alignment of the interests of capital, the state and the middle classes, but ongoing liberalization has undermined middle-class housing affordability – revealing a key tension between capital and middle-class interests. This tension triggered new, more restrictive policies.


Introduction
In the years following the Great Financial Crisis (GFC), the Dutch government -both national and local -actively sought to promote growth and liberalization of the private-rental housing sector. This is perhaps best symbolized by then housing minister Stef Blok (in office as housing minister from 2012 to 2017) visiting domestic and international real estate fairs to entice (foreign) investors to buy up Dutch social housing. His arguments were that his government enabled stronger rent increases, competition from not-for-profit housing associations was restricted, and residential demand was high (Hochstenbach 2022). This would translate into appealing profit margins for investors. His narrative was accompanied by a string of policies indeed encouraging private landlordism -liberalizing rental contracts, weakening tenant protection, residualizing social rent, and removing other restrictive barriers. But, in recent years, state policies have started to change. Moving away from promoting the private rental market, the national government has implemented measures to actively restrict it. It has introduced an extra stamp duty for buy-to-let purchases limiting profitability (2021), it has introduced legislation allowing municipalities to completely ban buy-to-let purchases (2022), and it has announced tighter rent regulations that would affect most private-rental units (coming into effect in 2024) as well as scaling back short-term rent leases (decided mid-2023). The dominant political orientation of national government meanwhile has not changed. Consecutive cabinets led by Prime Minister Mark Rutte of the market-liberal VVD party have been in charge since 2010. So the puzzle this paper answers is as follows: why has the Dutch government, in a short time span, gone from vigorously promoting the private-rental sector to actively restricting it? This is not just an empirical puzzle. It requires us to think through the political economy of rental housing and particularly the role of the state. Political economy perspectives typically centre on how rental housing has become integrated in global capital flows (Beswick et al. 2016;Janoschka et al. 2020) as well as the property-led accumulation strategies of corporations, institutions and individuals (Fields 2018;Ronald and Kadi 2018;Aalbers et al. 2021) in which rental housing figures as "just another asset class" (Van Loon and Aalbers 2017). The state is typically seen to accommodate or ignite these processes (Gil García and Martínez López 2023), which may be considered emblematic of neoliberalization in which the state does not so much retrench but actively stimulates market forces (Brenner, Peck, and Theodore 2010;Oudenampsen and Mellink 2021). The state, however, does not only serve a property-led capital accumulation regime, it is also particularly responsive to the demands of middle-class populations (Schakel 2021;Boterman and Van Gent 2023). As housing affordability is under pressure, especially in cities and increasingly also for the middle classes, a new wave of post-neoliberal housing policies is emerging in different countries (see Kadi, Vollmer, and Stein 2021;Wijburg 2021;Byrne 2022).
The central argument of this paper, then, is that post-GFC policies promoting privaterental growth and liberalization, as well as the subsequent turn restrictive policies, are both the outcome of the state seeking to balance the property-led accumulation regime accommodating small-scale landlords and large investors alike, with the residential demands of the middle classes. This balancing act has inspired both neoliberal and, subsequently, post-neoliberal politics. This paper first draws on an analysis of policies to argue that the political promotion of private renting was the outcome of a presumed alignment of the interests of capital, the state and the middle classes. Capital turned to rental housing in search for sufficient and stable yields, the state turned to private investors for (rental) housing development, and the middle class turned to private renting as alternative tenures (social rent or owning) were closed off. Subsequently, it uses survey data to empirically demonstrate that private renting indeed came to play a more central role in housing dynamics and accommodating the middle class. While investments in new private-rental developments (build-to-rent) increased supply, the advance of buy-to-let further undermined the ability of the middle class to buy a house -still the ideologically preferred tenure -and rapidly increasing rents have rendered the private-rental sector unaffordable for many middle-class households as well. The paper therefore finally argues that the turn towards more restrictive policies is the result of a key tension in these housing politics. The accumulation regime needs to be balanced with the residential demands of the middle class. 1 Both are central to the private-rental project but increasingly at odds with each other.
The case of this paper is the Netherlands. Following Aalbers (2022), I view the Dutch housing system to be on a trajectory similar to many other countries, although acknowledging its unique starting point. Private-rental sectors across countries show similar development trajectories and are shaped by similar forces (elaborated upon below), albeit differentiated in magnitude and mediated by institutional arrangements. Importantly, in international comparison the Netherlands still boasts a large social-rental sector and relatively strong tenant protections. The findings help us understand the political economy of rental housing, the potentially contradictive role of the state, and the particular role of middle-class residential demands beyond the case of the Netherlands.

The Political Economy of Private Renting
For decades, the political economy of housing of many European countries was dominated by the ambition to expand homeownership to a majority of their populations (Doling and Ford 2007). Homeownership was the ideological project of the state (Ronald 2008) as the promise was that this would create independent, secure, proud and prosperous citizens. The primary fuel for homeownership expansion was mortgage debt. While credit is indeed necessary to smooth access, when debts continue to increase they do not so much broaden access but rather inflate house prices (Kohl 2021): rapidly increasing volumes of (mortgaged) capital are used to finance an incrementally growing housing stock. Ryan-Collins (2018) termed this the housing-finance feedback cycle, where increasing mortgages drive up house prices, which in turn again require larger mortgages to afford higher price levels. Levels of mortgage debt are therefore regularly used as a rough proxy for the depth of housing financialization, with the Netherlands having one of the highest mortgage-debt-to-GDP ratios in the world (Schwartz and Seabrooke 2008;Hick and Stephens 2023).
Housing insiders benefited, as house-price inflation for them equated to housingwealth accumulation (Arundel and Ronald 2021). This played a central dynamic in the reconcentration of wealth as returns on assets increasingly outstrip those on labour (Arundel 2017). In search of profitable reinvestment, these insiders increasingly ended up buying additional property to rent out, i.e. buy-to-let. These forces contribute to a rentierization of society (Christophers 2020), in which generating passive income through the possession of scarce assets becomes increasingly crucial in shaping economic opportunity and class position (e.g. on Australia, see Adkins, Cooper, and Konings 2021;Ryan-Collins and Murray 2021).
As multiple property ownership among the accumulating classes advances (Kadi, Hochstenbach, and Lennartz 2020), ownership rates among other groups have dwindled. This has led some to speak of a "post-homeownership" (Ronald and Kadi 2018) or "late homeownership" (Forrest and Hirayama 2018) society in which greater segments of the population have to resort to renting (Smith et al. 2022;Jacobs, Atkinson, and Warr 2022) as a coping strategy rather than a positive choice (Hulse, Morris, and Pawson 2019), for example because they are precariously employed (Arundel and Lennartz 2020). While in many Anglophone countries this led to decreasing homeownership rates, in the Netherlands they have remained stable at around 57% in the post-GFC decade. Yet, access is clearly polarizing with homeownership rates among young adults and lowincome households showing a steep drop (Hochstenbach and Arundel 2021). The longterm promotion of homeownership has, thus, via the re-concentration of wealth and property, set the scene for the contemporary revival in private renting (Aalbers et al. 2021).
Firms, corporations, institutions and trusts have also turned to rental housing in search for yield. Along with declining interest rates in the post-GFC years, returns on other asset classes have gone down, making rental investments comparatively more appealing (Waldron 2018;Fields 2018;Wijburg, Aalbers, and Heeg 2018), offering the prospect of direct yields (rental income) and indirect ones (property price inflation). These actors also respond to increasing demand for market-rental housing from populations unable to access other tenures, as well as from populations such as international knowledge workers looking for flexible housing (Huisman and Mulder 2022). Here, we essentially find another feedback cycle: as demand for private renting increases, more investors will flock to the market. They increase rental supply through new construction, but also by buying up existing property of other tenures. This further squeezes households, making them again more dependent on private renting which fuels demand.

Private-Rental Politics
In various countries, states have played a more or less proactive role in fuelling growth of the private-rental sector in recent years. They may do so not solely through individual measures, but also through coordinated "policy packages", i.e. coordinated regulatory efforts to attract and channel investment flows (Gil García and Martínez López 2023). Dutch housing politics are a case in point where the government has sought to attract a range of investors and landlords to develop new rental housing (build-to-rent) as well as buy up existing stock (buy-to-let). They have done so through coordinated efforts. This section outlines two key reasons why states may employ these strategies.
First, states typically depend on private capital to fund new housing construction, and often increasingly so as public models of housing provision have eroded. In particular, build-to-rent by large institutions, private equity, investment trusts and pension funds is increasingly prominent (Nethercote 2020;Brill and Durrant 2021;Aalbers et al. 2023). States have aimed to attract such rental investors as they are considered stable partners with long-term stakes and investment horizons, thus ostensibly providing "patient capital" (Deeg and Hardie 2016;Brill, Raco, and Ward 2022). Authorities may furthermore pursue entrepreneurial modes of urban development, often with considerable risks (Ferm and Raco 2020;Haughton, Allmendinger, and Oosterlynck 2013;Taşan-Kok and Özogul 2021). This has also been the case in the Netherlands where profit streams from build-torent are used to fund urban development, social-housing construction and other state activities (Van Loon, Oosterlynck, and Aalbers 2019; Savini 2017). To facilitate capital switches into the built environment (Harvey 1982), states may seek to remove regulatory barriers or introduce new measures to smooth the process (Bernt 2022;Nethercote 2022).
Efforts to attract private investments in rental housing are a central part of a financialization of the tenure (Fields 2018;Wijburg, Aalbers, and Heeg 2018). In the Netherlands this was, Aalbers et al. (2021) suggest, the result of the homeownership project running out of steam. Capital flows into owner-occupancy stagnated (but by no means disappeared), which led the state to open new investment opportunities in the rental market. This process can be linked to the concept of "circuits of finance", that is the sociotechnical systems that channel wealth and debt-fuelled investments (Halbert and Attuyer 2016). Norris and Lawson (2022, 4) argue these circuits shape "the forms and costs of housing provision and consumption." Yet, these circuits may be destabilized, for example as they are disrupted by economic crises or regulatory changes. This leads to "uneven flows of investment and disinvestment over space, time and asset category." In the Netherlands, as elsewhere (Nethercote 2020;Goulding, Leaver, and Silver 2023), these flows were redirected towards rental housing provision.
Second, states have turned to investors and landlords to provide housing for groups that are ill-served by other housing providers. This rationale pertains to both build-to-rent and buy-to-let. These are typically outsider groups lacking sufficient income, savings and employment security to buy, while social-rental alternatives have been dismantled (Kemeny 1995;Malpass 2004). The gradual residualization of social renting and concomitant dualization of the housing system have been key transformations in the Dutch housing system since the 1990s but have accelerated post-GFC. The still comparatively large social-rental sector is steadily declining in size and access has been restricted to those on a low income (Van Gent and Hochstenbach 2020). We thus see the growth of a squeezed middle -unable to buy and ineligible for social rent -which incentivizes governments to promote private-rental growth and entices investors.

Transformations in State Politics
State and market are not opposed, but instead fundamentally intertwined. Markets are socially constructed and embedded in social relations. States play a central role in defining, creating and maintaining markets, influencing their breadth and depth (Fligstein and Dauter 2007). This is certainly also the case for private-rental housing markets. As outlined in the introduction, the Dutch private-rental market has been subject to multiple rounds of policy transformations that alter state-market configurations. These include policies that support and fuel market actors as well as those that restrict them. It would be too simple, however, to simply assume that one set of policies is simply replaced by another one. Older policies may continue to exist or be reformed to reflect new goals and ambitions, while new policies may be layered on top of older formations (Streeck and Thelen 2005).
Ongoing market-oriented reforms may be combined with persistent regulation in key parts of housing systems -resulting in housing systems that resemble "monstrous hybrids" of regulation and marketization (Christophers 2013). Kettunen and Ruonavaara (2021) for example conclude that despite neoliberalization the private-rental sectors of most European countries remain regulated to various degrees (also Marsh, Gibb, and Soaita 2022).
The new policy layers themselves can also be contradictory. Neoliberalization and marketization may themselves not be de-regulatory at heart. Although some restrictive policies may be dismantled, new forms of regulation stimulating market forces may be introduced -essentially amounting to a form of "regulated deregulation" (Aalbers 2016).
A large literature underwrites the argument that neoliberalism does not assume an absent or retrenching state but rather a state that is very much active by creating the conditions for markets to flourish (Brenner, Peck, and Theodore 2010;Gerstle 2022). Furthermore, neoliberalization is an uneven, fragmented and contradictory process. Even the most market-liberal states are not purely the servants of capital, but take into account other interests as well -which may translate into restrictive regulation.
In the post-GFC period, some authors have identified a turn towards "post-neoliberal" politics (Davies and Gane 2021) as the neoliberal "political order" unravels (Gerstle 2022). This is not a sudden event but rather a gradual and uneven process in which neoliberal political ideas are increasingly criticized or discredited, e.g. in the wake of widening inequalities and cascading crises. Conversely, political ideas of de-commodification that until recently would have been considered unlikely have gained traction also in mainstream politics (Davies and Gane 2021).
In the field of housing, an incipient literature has noted renewed political efforts to dampen housing commodification and financialization through initiatives outside dominant market logics. Kadi, Vollmer, and Stein (2021) find a shift towards "post-neoliberal housing policy" in New York, Berlin and Vienna. Policies include a reinvestment in social housing, a broader regulation of rents and a strengthening of tenant rights (Byrne 2022;Marsh, Gibb, and Soaita 2022). Kholodilin and Kohl (2023, 2) speak of a "renaissance of rent control," accelerated by the COVID-19 pandemic. While not a complete overhaul, these policies amount to a shifting compromise between capital, labour and the state (Wijburg 2021).
Post-neoliberal housing policies typically find their origin, and continue to concentrate, in major cities. This has various reasons. For one, the adverse effects of the housing crisis have been most palpable in cities where residential demand and capital flows have concentrated (Wetzstein 2017). Those cities are furthermore frequently governed by relatively progressive authorities, more willing to push back against capital and national politics dictating market logics (Hochstenbach and Ronald 2020;Wijburg 2021;Kadi, Vollmer, and Stein 2021). In addition, cities have historically played a central role in the formation of social movements (Harvey 2012) and have in recent years seen the proliferation of new housing movements (Schipper 2015;Kadi, Vollmer, and Stein 2021;Card 2022).
The turn towards post-neoliberal housing politics remains partial and incomplete, though. Neoliberal logics may continue to dominate as evidenced by the influence of global capital, of conservative national governments, and of the influential real-estate lobby (Kadi, Vollmer, and Stein 2021;Wijburg 2021). New measures then struggle to fundamentally challenge existing market logics but have to operate within them to tackle market excesses. Writing about Amsterdam, Hochstenbach and Ronald (2020) discerned a politics of "regulated marketization" in which new regulations are embedded within logics of commodification while also seeking to correct for market failures. They furthermore go on to show that efforts of housing re-regulation are primarily concerned with the availability of affordable housing for middle-income households rather than for those worse-off (also Boterman and Van Gent 2023). Furthermore, while governments may impose new regulations that bracket commodification or financialization in certain domains or housing tenures, the result may be to simply shift, or refocus, capital flows and financial logics elsewhere, e.g. to different housing tenures or geographies (cf. Norris and Lawson 2022). The result, then, is a complex amalgam of contradictory, reinforcing or overlapping neoliberal and post-neoliberal housing policies.
The following sections will highlight a key tension in contemporary Dutch privaterental politics: how to balance the accumulation strategies of landlords and investors with the housing demands of middle-class residents? In achieving private-rental regrowth, the state depends on investor capital to ensure supply, which requires sufficient yields. This, however, goes at the expense of rental affordability for the middle classes. This central tension explains the emergence of increasingly complex policy arrangements pushing for liberalization (to enable yields) and re-regulation (to ensure middle-class affordability). From this, the paper posits that to understand the contemporary political economy of renting it is insufficient to focus only on capital and finance, but that the residential demands and concerns of particularly middle-class residents need to be incorporated as well. In particular, the promise to improve middle-class housing options is central to legitimizing housing politics.

Data and Methods
To understand the politics of liberalization and re-regulation, this paper draws on a macro-level analysis of housing politics and a micro-level analysis of residential outcomes in the private-rental sector, while drawing links between the two. This paper explores the logics of policy transformation by tracing how macro-level housing policies shape micro-level residential outcomes, which in turn trigger policy responses at the macro-level (cf . Coleman 1990). This implies an analysis of housing politics and housing outcomes, which is sensitive to their interaction over time and across scales. To do so, a (qualitative) policy analysis needs to be combined with (quantitative) residential analysis.
In analysing housing politics, the paper describes and traces the implementation of different key policy measures during the post-GFC period. To interpret the underlying motivations for policy implementation, it draws on exemplary quotes selected from policy documents, parliamentary notes and media statements. It particularly pays attention to how housing struggles are problematized, and how the private-rental market is constructed as part of either the solution or the problem (also Browne et al. 2019).
In analysing housing outcomes, this paper studies how the residential composition of the private-rental sector has changed and in which conditions these groups liveparticularly in terms of housing costs. To do so, this paper draws on descriptive analyses of the Netherlands' Housing Survey (Woon Onderzoek Nederland, or WoON). Statistics Netherlands conducts the cross-sectional survey every 3 years, gathering information about the living situation, preferences and characteristics of respondents.
This paper draws on the four most recent waves (2012, 2015, 2018 and 2021). These waves have been chosen for data consistency and because this time frame corresponds with the post-GFC period. It uses data on housing characteristics (tenure, rent levels, contract type, etc.) and household characteristics (age, income, household composition, etc.). In the case of individual-level characteristics such as age, the household main earner is taken as the reference person. A crude geographical distinction between respondents living in one of the four major cities and elsewhere is possible.
All rents reported in this paper are basic rents excluding utilities and corrected for inflation. A distinction is made between regulated and liberalized rents, which in 2021 concern units with monthly rents below and above the threshold of €752, respectively. This distinction is policy driven and determines rent regulation and household eligibility (further explained below). The threshold is subject to annual incremental changetypically close, though not identical, to inflation. A further distinction is made between rental units with rents below or above €1000 per month. Although this is not a legal distinction, it is frequently used by policymakers and politicians to distinguish between middle-income and expensive rental housing. These labels, it is important to emphasize, are normative. The term "middle-income" or "middle-segment" rental housing suggests normality and broad affordability, even when in reality rents may be above average and difficult to afford for those earning a middle income. As I am interested in the interaction between policies and household outcomes, this normative policy-driven categorization is particularly relevant.
To establish the socio-economic position, this paper draws on gross household income (inflation corrected). Households earning less than €500 per month are excluded from the dataset, as these are typically not living independently (for all waves, this concerns less than 1% of the cases). Subsequently, I constructed three equally large groups of low-, middle-and high-income households based on tertiles. In 2021, the middle-income group earned between €37,884 and €73,109 per year -roughly corresponding to the maximum income to be eligible for social rent (€40,024 in 2021) and twice the modal Dutch income, respectively. To assess rent burdens, this paper uses the pre-constructed variable of basic rent as a percentage of disposable household income. This measure excludes utilities as well as potential rent subsidies for low-income households (in the Dutch housing system, only low-income renters in the regulated sector are eligible for subsidies). While at the household level this is a crude measure of housing affordability due to substantial variations in other expenses, at aggregate levels it does indicate how affordability differs between groups and over time.
Extra attention is being paid to the housing outcomes for recent movers, as this is where changing regulations and other housing dynamics are most likely to have an impact. The questionnaire includes a question whether households have moved over the last 2 years. This was the case for around 15% of the respondents in 2012 and 17% in 2021.
After removing cases with missing tenure or income information, the datasets include 58,064 respondents between 2012 and 38,637 in 2021. Pre-made weights are used to correct for sample biases such as the descriptive overrepresentation of homeowners, the rich and the old. Appendix A gives an overview of the unweighted and weighted distributions of respondents over key variables.

Promoting Private-Rental Regrowth
For much of the post-WW2 period up until the early 2000s, the Dutch private-rental sector declined in size. Although there are various explanations for this decline, housing policies discouraging private renting played an important part. For one, most new developments focused on providing new owner-occupied or social-rental units. Rents became increasingly regulated, putting a strain on profit margins and triggering conversions to owneroccupancy. Other private-rental units were expropriated, often in the context of urban renewal, and placed under housing association management. The private-rental share declined from around 60% of all housing in 1947 to around 10% in the mid-2000s (Musterd 2014).
The post-GFC period subsequently saw a return to private-rental growth (Aalbers et al. 2021). As outlined above, private investment in rental housing was seen as crucial to boost overall housing supply and especially for middle-income households. Policy reforms therefore focused on making private-rental investments more attractive, e.g. by allowing for stronger rent increases (2015), especially in high-demand areas, and by introducing temporary contracts where permanent contracts used to be the norm (2016). Essentially, all these measures removed or minimized barriers for private accumulation through the rental property sector.
In 2019, minister of the interior Kajsa Ollongren (responsible for housing from 2017 to 2021) reflected on the crucial role of private-rental investors for boosting total supply, appealing to the image of good investors providing "patient capital", especially to invest in the middle segment: "Investors are necessary for a healthy housing market. They provide capital for the construction of new-build homes and have ensured significant development in the mid-rental segment in recent years. Investors, developers and market parties have billions available to invest in new construction. Building more is ultimately the structural solution [to housing issues]. We must therefore prevent sustainable investors from being chased away and ensure that excesses are tackled in a targeted manner." (Ollongren 2019, author translation).
These measures did, however, not only boost private-rental housing production but also contributed to a steep increase in buy-to-let tenure conversions, which do not increase total supply but shift housing from one tenure (owner occupation) to another (private rent). In 2017, Ollongren explained that she did not consider the advance of buy-to-let problematic: "Buy-to-let can, generally speaking, contribute to a desirable expansion of the supply of rental housing, especially when this expansion concentrates in the still relatively small middle segment" (cited in Ronald 2020, 1635). This quote illustrates how private-rental expansion, even through conversions rather than actual increases in supply, was justified as it was seen to improve the housing opportunities of the middle classes.
Beyond housing liberalization, private-rental growth was further smoothed by the ever-growing dominance of market logics in planning. While the (local) state can shape what gets build -determining the tenure, type and pricing of houses on offer, it increasingly depends on private investment as public funds have been slashed. Area developments have thus become more dependent on generating revenue streams from profitable market housing, which are subsequently used to fund infrastructure, public space as well as unprofitable social housing (Savini 2017). This encourages local states to prioritize more expensive build-to-rent schemes.
Supportive policies contributed to clear private-rental growth in the post-GFC decade. Between 2012 and 2021 the private-rental share in the total housing stock increased from around 12% to 14% (Figure 1). While this may seem like a modest increase, it is underpinned by a much more central role of private rent in contemporary housing dynamics. As the total Dutch housing stock grew by 646,000 units over the 2012-2021 period, net growth of the private-rental stock was 254,000 units. In other words, private-rental growth represented 39% of the total housing growth. In most major cities and university cities, private-rental dynamics were even stronger and dominated total housing growth. In Amsterdam, for example, the private-rental share increased from 25% to 30% (Figure 1) as 36,000 private-rental units were added to the city's stock (Figure 2). These dynamics are fed by new developments (build-to-rent) and conversions (buy-tolet). While exact figures to decompose growth are unavailable, some rough estimates are possible. Data from Statistics Netherlands show permits for the construction of around 120,000 rental units by market developers -typically larger institutional investors -were granted between 2012 and 2021, some 22% of the total permits granted. 2 In the four largest cities, rental units by market developers even represented 41% (37,000 units) of all building permits granted over this period.
The Dutch land registry estimates that buy-to-let was responsible for a further net growth of the private-rental sector by some 75,000 units between 2009 and 2019 (Kadaster 2021). A wide range of landlords are involved in these purchases, ranging from small-scale private individuals and small companies to large multinational firms such as Patrizia and Blackstone (Hochstenbach 2022;Fernandez, Hartlief, and Hudig 2022). Another 62,000 units were converted from owner-occupation into private rental through "keep-to-let" -in which former owner-occupiers move but hold on to their property to subsequently rent out (Kadaster 2021).
From these trends, we can derive that even though private rent remains the smallest tenure, it has come to play a central role in (1)

Restructuring the Private-Rental Sector
Private-rental growth was only one part of the political project. Another key part was the liberalization of the private-rental sector. Historically, the vast majority of private-rental units (as well as social-rental units) were rent regulated, with maximum rents based on an elaborate "point system" in which dwellings score points based on features, quality and size. The point system is a form of second-generation rent control, in which the number of points determines the maximum starting rent with annual increases within a tenancy tied to inflation levels. All dwellings scoring up to 141 points cost up to €752 per month (2021 level, subject to annual incremental change). Any unit that scores more points passes the "liberalization threshold" above which rent controls are absent.
Crucially, the regulated sector is not only cheaper with strict rent controls in place but access is also limited to lower-income households. Liberalization was therefore necessary to accommodate middle-income groups, one of the specific policy aims. Here, it is worth pointing the strict income criteria to be eligible for regulated housing are not natural givens but political choices. They were only formally implemented in 2011. This came after institutional investors filed a complaint at the European Commission for unfair competition between market providers and not-for-profit associations. The EC agreed and compelled the Dutch state to introduce income criteria, but it was the latter that decided upon a particularly stringent limit (Elsinga, Haffner, and Van der Heijden 2008; Van Gent and Hochstenbach 2020). The political choice to exclude middle-income households from rent-regulated housing in turn necessitating the creation of a new middle segment. This decision, effectively, eliminated competition from social-housing providers and gave private investors free rein in accommodating middle-income groups in rental units with higher rents. One of the key ways in which national government smoothed rent liberalization, was through the inclusion of assessed real-estate values in the aforementioned point system determining maximum rents (in 2015). Before that, liberalization was only possible through physical improvements to the dwellings. Especially in high demand areas the result of including these values was that most rental units were suddenly eligible for liberalization without any improvements. With real-estate values increasing, this was the case for more and more rental units. As liberalization is only possible in the case of residential turnover -with sitting tenants being protected -the large-scale introduction of temporary contracts (in 2016) further increased the potential for liberalization.
Data show a clear increase in the share of private-rental dwellings that are rent liberalized, with the trend accelerating after 2018. At the national level, the share of private tenants with a liberalized rent increased from 32% in 2012 to 53% in 2021 (Figure 3). These are all households that spent a minimum of 752 euros in rent per month (2021 level, excluding utilities). These shares still include a large number of incumbent tenants whose rents have therefore remained rather stable. Looking at recent movers -operationalized as those who moved in the last 2 years -therefore reveals a stronger trend towards rent liberalization: from 39% in 2012 to 65% in 2021.
In the four largest cities -Amsterdam, Rotterdam, Den Haag and Utrecht -increases were even steeper. Here, the share of liberalized private-rental units increased from 30% in 2012 to 59%, while these shares for recent movers increased from 39% to 76%. There are various factors that explain why liberalization increased at a relatively fast rate in cities: house price increases were relatively steep, residential turnover consistently higher and overall private-rental growth (whether through build-to-rent or buy-to-let) strong. These interlocking dynamics all enhanced the scope for liberalization.
The above has shown both a clear growth ( Figure 1) and liberalization (Figure 3) of the private-rental sector. Coherent policy efforts refocused investment flows to the tenure. Increasing numbers of households consequently find a place to live in the tenure, but these increases are unevenly distributed (Figure 4). In particular, young adults, singleperson households and urbanites are overrepresented in the private-rental sector. And this is, again, more evident when looking at recent moves. Among recent movers, the share of private renters among middle-income households increased from around 22% to 33%. In line with policy ambitions, the private-rental sector increasingly came to accommodate squeezed middle-income groups unable to access other tenures.

Private-Rental Unaffordability
As middle-income households became more dependent on private rent, affordability of the tenure has steadily worsened. Liberalization combined with high levels of demand led to increasing rent levels. Median monthly rents among recent movers increased from €677 in 2012 to €850 in 2021 (after inflation correction). The upward pressure is especially apparent in the four major cities where rents went up from €658 to €1000. 3 In political discussions, it is typically emphasized that middle-income households should not spend more than €1000 per month on rent (excluding utilities). While it is certainly possible to debate this level (for being too high, too low or insensitive to contextual variations), Figure 5 shows that in recent years particularly the more expensive private-rental sector has grown. This works as follows: when dwellings pass the liberalization threshold of 141 points (with a corresponding maximum monthly rent of €752), landlords are free to determine the rent. Especially in high demand areas, rents can quickly jump to far higher levels, resulting in a "missing middle." In 2015, around 15% of the recent movers into the tenure paid at least €1000 in rent per month, and in 2021 this share had doubled to over 30%. Among middle-income households, this increase was even more evident. Between 2015 and 2021, their share of moving into expensive private-rental units almost quadrupled from around 10% to around 38%. At the same time, the share moving into middle-segment housing remained roughly stable around 38%, while those moving into more affordable regulated rental units went down from 47% to 20%. 4 Also among other incomes, a steep decrease in affordable private rental units can be discerned with particularly the most expensive segment increasing ( Figure 5). The political ambition to stimulate the private-rental sector was motivated by the promise to create middle-income housing, but these figures underscore that in reality it was particularly the expensive sector that expanded.
The total increase in rents still downplays the actual rate of change. As rents have increased, private tenants have in fact been getting less in return. The share of private tenants moving into one-room studios doubled from around 7% in 2012 to 14% in 2021. Overall, the share of private tenants moving into relatively small housing situations (defined as the number of rooms being equal to, or less than the number of occupants) increased from around 18% to 29%. In both social-rental and owner-occupied tenures, similar trends are not discernable. While in some cases this may signal a conscious decision by private tenants to live relatively small as well as an increase in dwelling quality beyond size, these patterns at least suggest living relatively smaller may be an increasingly common coping strategy to overcome higher rents and a lack of alternatives. Furthermore, following the roll-out of temporary contracts, residential insecurity has become more commonplace among private tenants. Although national-level longitudinal data are not yet available, data from 2021 show that 13% of all private tenants had a temporary contract. Among recent movers this was 29% -with the share being slightly higher (32%) among middle-income households in private rent. We thus see a triple trend where private tenants pay more and get less housing in return and experience more residential instability. These trends are particularly pronounced among recent movers and in larger cities.
Subsequently, the paper analysed to what extent private tenants spend a relatively high share of their income on housing. Figure 6 focuses on private tenants who spend 30% or more of their disposable income on rent (excluding utilities). 5 These households are potentially rent burdened. Among all private tenants, these shares went up from 34% in 2012 to 46%. While still a rough proxy, these figures indicate that a growing share of Figure 6. Share of private tenants spending 30% or more of their disposable income on rent, 2012 and 2021 (with 95% confidence intervals). Source: own adaptation of the WoON survey. private renters may struggle to afford the monthly rent -with almost half of all private tenants being potentially rent burdened.
These shares are particularly high among low-income households, with around 60% of them spending 30% or more of their income on basic rent (compared to around 50% in 2012). Rent burdens have, however, also clearly expanded to middle-income households, with around 43% of them experiencing relatively high rent burdens (compared to around 28% in 2012), while remaining relatively stable among high-income households. The same analyses were performed for tenants spending 40% or more of their income on rent (see Appendix B). The trends over time are highly similar -with clear increases for most categories -although overall shares are of course substantially lower (increasing from 16% of all private tenants in 2012 to 22% in 2021).

Re-Regulating the Private-Rental Sector
The above has shown how the state has promoted private-rental regrowth and liberalization, particularly to accommodate middle-income households excluded from other tenures. However, while this indeed opened up the private-rental sector for middleincome households, affordability of tenure has steadily worsened over recent years and is now increasingly unaffordable not only to low-income households but also to middleincome ones. This represents a key tension.
The key tension is the result of two interests that are increasingly conflicting. On the one hand, there are the accumulation strategies of the providers of private-rental housing: investors and landlords. Rent liberalization has unlocked their potential for accumulation, which in the context of high demand, leads to steep rent increases. On the other hand, there are the housing needs of middle-income households: unable to access other tenures and in need of relatively cheap housing. In the years directly following the GFC, these interests appeared aligned, but as liberalization continued they are increasingly in conflict with each other. Further rent increases boost yields but reduce middle-class odds to find suitable housing.
The housing concerns of the middle classes were central in the proliferation of a "housing crisis" in Dutch political and public debates (see Hochstenbach 2023). While these debates began emerging in 2019, they culminated in multiple bottom-up housing protests in late 2021 across larger cities. The first two rallies in Amsterdam and Rotterdam were the largest housing protests since the early 1980s and focused on the political roots of the crisis situation. Importantly, an analysis of political and public debates revealed that disproportionate attention was directed towards the housing woes of the middle classes (presented in detail in Hochstenbach 2023). Not in the least, rampant housing costs and precarity in the private-rental sector were problematized, feeding societal and political calls to intervene in the rental market and tackle market excess through re-regulation.
Middle-class concerns -increasingly vocalized in debates -led national government to pursue re-regulation of the private-rental sector, after having actively stimulated liberalization for years. Since 2020, various measures have been introduced or announced to tighten the grip over the private-rental sector. Two types of measures can be discerned: policies limiting buy-to-let purchases and policies restricting rent levels.
First, following calls from city governments, the national government has implemented two measures to restrict buy-to-let purchases (by both landlord individuals and companies). In 2021, an extra stamp duty for landlords came into effect to make such investments financially less viable. While the "regular" stamp duty stood at 2% of purchase price, for buy-to-let investments this was increased to 8% and, in 2023 even to 10.4%. In the case of a €300,000 buy-to-let purchase, this means that the stamp duty increased from €6,000 (2%) to, eventually, €31,200 (10.4%). Conversely, the stamp duty was scrapped for first-time buyers as of 2021. In 2022, the national government furthermore allowed municipal authorities to ban buy-to-let. For juridical reasons of proportionality, the ban can only target parts of the stock, such as dwellings below a certain price level or located in specific neighbourhoods. All four major cities and many other municipalities have since implemented or announced such a ban.
Both measures were the result of concerns that landlords were crowding out regular home-buyers, particularly starter households. Political support for these restrictive measures was broad: left-wing parties wanted to curb landlord speculation, and right-wing parties wanted to uphold the ideal of widespread homeownership. Whereas only a few years earlier minister Ollongren remarked that buy-to-let could contribute to a desirable expansion of middle-income rental housing, the dominant frame has quickly shifted towards the idea that buy-to-let contributes to a problematic decrease in middle-income homeownership opportunities and a problematic deepening of housing affordability stress.
Second, the national government has announced measures to more broadly regulate rents again. The effects of the extra stamp duty and municipal buy-to-let ban are limited to new acquisitions, but leave the existing private-rental stock and new constructions untouched (build-to-rent), while political and public concerns also revolved around excessive rents in these segments. A general rent regulation would tackle this issue. Mid-2022 housing minister Hugo de Jonge therefore announced his intentions to expand rent regulation further using the existing point system. Under previous regulation only units scoring up to 141 points are regulated, but from 2024 onwards this includes all units scoring up to 187 points (corresponding to a maximum rent of around €1100 per month). Many "expensive" dwellings will be subjected to rent controls again with rents falling back to the middle-segment range (€752-1100). 6 Elaborating on his plans for re-regulation in a letter to parliament, De Jonge stated: "Affordable rental housing for people with a middle income has become less and less obvious in recent years. Scarcity has led to excesses in which the rent is no longer in proportion to the quality of the home" (Ministy of the Interior 2022, 1, author translation).
He furthermore notes that these measures are particularly crucial in cities where demand pressures are highest, where elsewhere homeownership might still be in reach for middleincome groups. However, while the proposed regulations will improve affordability for these "middle-segment" units, they do little to enhance low-income housing affordability. Indeed the Affordable Housing Programme of the ministry (Ministry of the Interior 2023, 16) envisions a steep increase in middle-income housing supply up until 2030, while the share of affordable housing units is projected to further decrease.
Landlord lobby organizations, investors and developers have subsequently warned that the proposed regulations undermine the financial viability of new projects -especially in combination with increasing building costs, high inflation and economic uncertainty. Desirée Uitzetter, chair of developers' association Neprom, elaborates in a press release: "With the proposed regulation, the permitted rent will soon be hundreds of euros per month lower than what developers, investors and municipalities counted on in their plans. At first glance, that sounds like good news for tenants, but those homes will ultimately not be built because they have become unfeasible for investors" (Neprom 2022, author translation).
Although such statements and press releases, of course, reflect vested interests -in this case of developers and investors -they do highlight the heightened tensions between middle-income housing affordability and investor accumulation. International scholarly evidence suggests rent controls may restrict new constructions, although precise effects are mixed and context dependent, and can be offset by other measures (Kholodilin and Kohl 2023).
This points to a key tension as the state, on the one hand, seeks to improve middleclass housing affordability, but, on the other hand, depends on investors to provide new homes. This is already reflected in the Rutte IV coalition agreement presented in December 2021, which states: "Middle-segment rental housing will get some form of rent control, in such a way that housing for middle-income households becomes affordable and it remains profitable for institutional investors to invest in these homes" (VVD, CDA & CU 2021, 17, author translation).
In regulating rents, revenues for buy-to-let and build-to-rent investors alike will be capped. Alongside rising interest rates, high inflation and geopolitical uncertainty following the war in Ukraine, investor returns are further squeezed. This may lead institutional investors to pull out of development projects, undermining new supply. In response, the national government has decided on implementing several exemptions to maintain investor yields, among other things by granting new dwellings a general 5% top-up in rents.
These measures do not only signal a shift from rental liberalization to regulation but also represent the state seeking to strike a new balance between investors and middleclass residents. The analyses in this paper suggest that this may prove difficult as these groups' interests have proven to be difficult to commensurate.

Conclusion
This paper has studied the regrowth and liberalization of the Dutch private-rental sector in the post-GFC period. Specifically, we address the question why and how state policies shifted from vigorously promoting the private-rental market to actively restricting it. While the findings pertain to the Dutch case, some conclusions with broader relevance can be distilled from them. In particular, these conclusions are relevant to other countries on a similar private-rental trajectory (Aalbers 2022).
This paper has shown how private-rental regrowth, liberalization and subsequently reregulation are rooted in the interaction between state policies, capital flows and residential demands (also Wijburg 2021). In the years directly following the GFC, the state ostensibly facilitated both landlord accumulation and middle-class residential demands through their politics of rent liberalization. Middle-income households' diminishing access to owner occupancy and social rent were compensated through re-investment in the private-rental market. Yet, this period was a temporary one as there is a key tension between these two ambitions. As liberalization progressed, landlord accumulation undermined middle-class housing opportunities, residential stability and housing affordability. This tension represents a conflict, or antagonism, between housing as part of financial accumulation strategies and housing as part of social reproduction practices (Byrne 2019). Or, in more economic terms, it represents a conflict between housing's production function (for landlords) and consumption function (for tenants). The new regulations are not simply a return to pre-GFC conditions though, as the middle classes remain excluded from social rent and are accommodated in a newly regulated middle segment instead.
These findings deepen our understanding of the variegated and uneven paths of housing liberalization and subsequent re-regulation (Brenner, Peck, and Theodore 2010). The politicization of the housing crisis and steadily worsening housing opportunities have urged governments across various urban and national contexts to pursue progressive measures at the expense of landlord interests (Wetzstein 2017;Byrne 2020). Recent years have seen incipient debates around an emerging "post-neoliberal" housing politics. This paper speaks to these debates. For one, it similarly highlights a notable turn in political ambitions and efforts -away from rental liberalization and towards regulationbut also underscores that fundamental alternatives so far remain eschewed (see also Kadi, Vollmer, and Stein 2021;Byrne 2022;Wijburg 2021). That is, new measures do put a brake on commodification, but do not challenge dominant market logics, as is for example evidenced by a continuing dependency on investors in ensuring sufficient new supply. The paper forwards debates around post-neoliberal housing politics by arguing that the political focus on the "middle-class housing crisis" have been central in understanding the turn towards re-regulation (also Boterman and Van Gent 2023). It has urged the state to change course. Yet, the prioritization of middle-class concerns and interests over those of more disadvantaged populations may also help explain why more fundamental housing alternatives -such as broadly accessible social-rental housing -remain largely unexplored.
More broadly, while we see a trend towards the re-regulation of rental housing markets, this does not necessarily add up to overall de-commodification or definancialization. New regulations may certainly disrupt financial flows into private-rental housing, but these flows do not simply disappear. Instead, in an attempt to again realize the normative ideal of homeownership among the middle classes, the focus of financial circuits may again shift towards owner-occupancy through debt and wealth investments (see Norris and Lawson 2022;Hochstenbach and Aalbers 2023). Indeed, this possibility echoes the conclusion by Norris and Lawson (2022, 13) that policymakers are concerned "by the fact that middle class home buyers are no longer among the winners from financialization" and are therefore now "debating how to tame financialization". Again, we see here that regulatory measures are undertaken to serve middle-class housing demands.
Previous studies on post-neoliberal housing policies have noted how they first took hold in major cities. This is perhaps no surprise, as cities are typically the places where the housing crisis is most severe (Wetzstein 2017), where progressive local governments are (more) willing to intervene (Wijburg 2021;Hochstenbach and Ronald 2020) and where social movements have the strongest voice (Card 2022). The analyses presented in this paper have focused primarily on the national level, but also underscores important urban dimensions. The growth, liberalization and subsequent calls for re-regulation of the private-rental sector have been most prominent in the largest cities -Amsterdam in particular. Indeed, a previous study (Hochstenbach and Ronald 2020) noted that both the liberalization and re-regulation of the private-rental sector initially manifested itself at the urban scale before "going" national. It suggests the importance of "the urban" as an incubator of new policies and policy ideas, before they can be mobilized and transported to other contexts and scales . The findings of the present paper show how they have rolled out at a national scale. This paper has charted the interaction between policy transformations and residential outcomes, highlighting a crucial interaction between the two in shaping the political economy of renting. Methodologically, this requires integrating critically reflexive policy analyses with more quantitative, "positivist" analyses of residential outcomes. These methodological traditions, however, rarely join forces. This paper is a modest answer to pleas to do so (see also Wyly 2009;Sheppard 2014;Uitermark, Hochstenbach, and Groot 2023).
Finally, this paper has problematized private-rental growth and liberalization. One of the central outcomes of this paper has been the expansion of housing unaffordability, instability and insecurity to broader segments of the population, and the deepening of those problems for the worst off (also Dewilde 2018; McKee, Soaita, and Hoolachan 2020; Waldron 2023). Nevertheless, this should not be reason to uncritically embrace new (postneoliberal) housing politics containing elements of re-regulation and decommodification. It should rather be an invitation to critically unravel these new policies, exposing new tensions and contradictions that may emerge from them. Notes 1. While not the focus of this paper, indirectly, middle-class households may also have an interest in yields through the investment of their future pensions in housing. 2. Own calculations based on CBS (2023) data. Note that the year a permit was granted does not align with the construction year but with the start of a project. 3. Covid initially suppressed rents due to economic uncertainty and reduced demand from international migrants (especially students and expats). This, however, appeared to be a temporary dip. 4. Middle-income households are typically excluded from the regulated sector. The fact that still 20% of middle-income households moved into regulated rent may indicate they experienced an income gain after moving. Furthermore, some exceptions to these general rules exist. 5. This analysis does not zoom in on recent movers, due to relatively low numbers of observations and large confidence intervals for some subgroups. 6. In order to respect property rights and existing rent contracts, new regulations only apply to new tenancies.

Disclosure statement
No potential conflict of interest was reported by the author.

Funding
The