Just cities in the neoliberal era? : examining the variegating effects of neoliberalism on the urban housing systems of Amsterdam and Stockholm

J

UST

C

ITIES IN THE

N

EOLIBERAL

E

RA

?

Examining the variegating effects of neoliberalism on the urban housing systems

of Amsterdam and Stockholm

University of Amsterdam

Adriaan de Jong

1

By Adriaan de Jong

10492984 First reader Prof. Dr Richard Ronald Second Reader Prof. Dr Maria Kaika

Date 11th _{of June 2018}

Master thesis Urban and Regional Planning

Graduate of School Social Sciences University of Amsterdam

2

Acknowledgments

I would like to thank everyone who helped me in the process of this thesis. The field of housing systems and welfare regimes is very challenging, voluminous and ever growing. I am grateful to Richard Ronald who provided me with the necessary help when I got lost in the literature. His calmness, directness and guidance during the process of this thesis were of great importance. Furthermore, I would like to thank everyone who helped me with the interviews. Finally, I want to thank Maria Kaika for being the second reader.

3

TABLE OF CONTENTS

2.3. WELFARE REGIME AND HOUSING 9

2.4. NEO-LIBERAL ERA 10

2.5. CONCLUSION 12

3. METHODOLOGY 14

3.1. RESEARCH DESIGN 14

3.2. CONCEPTUAL FRAMEWORK AND OPERATIONALIZATION 15

3.3. SUB QUESTIONS 16

3.4. GENERAL CRITERIA 18

3.5. LIMITATIONS TO THE RESEARCH 19

4. MACRO-DYNAMICS UNDER NEOLIBERALISM 21

4.1. THE NETHERLANDS 21

4.2. SWEDEN 26

4.3. CONCLUSION 28

5. HOUSING OUTCOMES IN URBAN HOUSING SYSTEMS 29

5.1. AMSTERDAM 29

5.2. STOCKHOLM 35

5.3. CONCLUSION 38

6. COMPARING URBAN HOUSING SYSTEMS AND WELFARE REGIMES 39

6.1. AMSTERDAM 39

6.2. STOCKHOLM 40

6.3. CONCLUSION 40

7. CONCLUSION 44

JUST CITIES IN THE NEOLIBERAL ERA? 44

5

1. I

NTRODUCTION

In his book “Worlds of Welfare Capitalism” Esping-Andersen (1990) provided a foundation for international welfare state comparison. In this book he described the three ideal types of welfare states: the social-democratic, corporatist and liberal welfare state. This framework has been the basis of many international studies on welfare, including housing. Nevertheless, Esping-Andersen did not incorporate housing in his book, because of the strong link with the market due to its commodified character (Malpass, 2008).

On the basis of the welfare regime framework by Esping-Andersen (1990), academics have shown how western European housing systems have become more liberal (Kadi Musterd, 2015; Christophers, 2013; Wacquant, 2009; Hedin, Clark, Lundholm, & Malmberg, 2012; Lind & Lundström, 2007). The process of neoliberalisation in housing is observable in the sale of social housing, privatization of social housing companies, promotion of homeownership and concluding marketisation of housing. Neoliberalisation is increasing pressures on western European urban housing systems. This is partially due to decentralization; causing local governments to get more responsibilities in the provision and management of housing. Also, due to the marketisation of housing, real-estate prices are skyrocketing, which has a direct excluding effect on lower incomes in the inner-city. Many western European cities are coping with housing problems like inaccessibility and segregation (Andersson & Magnussen, 2014). In September 2017 UBS (2017) published their global real estate bubble index. This index demonstrates that house prices in Amsterdam, Munich, Toronto, Sydney and Stockholm rose by more than ten percent in 2016. Susan Fainstein (2005) describes Amsterdam as a prime example of the ‘Just City’. A city which is relatively successful in the provision of housing for all incomes (Kadi & Ronald, 2014). Therefore, Amsterdam is characterised as a social-democratic city regarding the regulation of housing. The Netherlands is characterised as a hybrid form between a social-democratic and corporatist welfare state (e.g. Hoekstra, 2003). Sweden is commonly known as the main example of a social democratic welfare state. However, neoliberal reforms have heavily impacted these regimes and their housing system (e.g. Grander, 2017; Christophers, 2013). Cities are logically also influenced by this trend. Uitermark (2009: 248) argues that the neoliberal reforms will have such a thorough effect on the housing system of Amsterdam that ’fututure Amsterdammers will not enjoy a just city Brenner et al. (2010) debate how different geo-institutional contexts will cause different local effects of neoliberal processes. Meaning that cities like Amsterdam and Stockholm will show different forms of the implementation of neoliberal influences in policy. However, these housing systems show similar high-pressure characteristics, like inaccessibility and high real estate prices, due to the marketisation of housing and the increased demand for housing in western European cities (UBS, 2017). This research investigates whether the urban housing systems and welfare regimes of Amsterdam and Stockholm demonstrate variegating effects of neoliberalisation as Brenner et al (2010) argue. Prior researches have shown that the production of new houses is very helpful in explaining developments in housing systems and welfare regimes (Hoekstra, 2003; Zhou & Ronald, 2017;).This issue is researched by illustrating the effects of neoliberal reforms on different scales.

The organization of this paper is as follows. First the influence of neoliberal macro dynamics on Sweden and the Netherlands are demonstrated on their housing systems. Next, we take a closer l look at the production of new housing in the urban housing systems of Amsterdam and Stockholm to show variegating effects. Finally, we will be able to tell whether the urban housing systems and welfare regimes of Amsterdam and Stockholm show variegating impacts by using the framework by Zhou & Ronald (2017).

7

2. T

HEORETICAL

F

RAMEWORK

This chapter discusses the relevant literature for this research. In order to analyse the nature of welfare regimes and housing systems in the neoliberal era, it is essential to study the significant theory on these concepts. The first part aims at explaining welfare regimes, in order to empirically research this concept in the further chapters. After which the relevant theory on housing systems theory is discussed. This is followed by the literature on the integration of welfare regime theory in housing and finalized by an explanation of the concept of neoliberalism and the corresponding theories.

2.1. W

R

After the second world war, when economies and demographics were booming, there was an increasing position of western governments to foster their inhabitants. This resulted in a large role for the governments providing basic needs for the people. This governmental act is called the provision of welfare services. However, the organization of the provision of welfare services is diverse among Western European states. The way in which this provision is organized can show wat ideals are pursued in a particular society.

Arguably the most influential conceptualization of welfare state regimes was made by Gøsta Esping-Andersen (1990) in the book ‘The three words of welfare capitalism’. His analysis defined three ideal types of welfare regimes. These classifications were based on the thought that a regime is a complex system of neatly intertwined arrangements between state, market and the family.

Esping-Andersen (1990) argued that there is not one form of the welfare state. He defined three welfare state typologies on the basis of his analysis of 18 OECD countries: the ‘Nordic social democratic regime’, the corporatist regime on the European continent and the Anglo-Saxon liberal model. The three regimes were separated on the basis of three main pillars. Firstly, de-commodification, which stands for the level to which a family is able to live contentedly without partaking on the labour market. Secondly, stratification, which stands for the way in which the government is counteracting segregation. Thirdly, the way in which employment is facilitated is crucial in analysing welfare regimes and the arrangements between the State family and the market (Hoekstra, 2003). This relation shows what body is directing the regime. The regimes however are not precise classes of their welfare state type. The regimes are meant as a framework to analyse differences between welfare states.

The social democratic welfare state type is recognised by a large role for the state (Esping-Andersen, 1990). The main idea of this welfare type is to lift the social economic position of the lower and middle-class incomes. This is achieved through direct and indirect subsidies, with higher subsidies for the lower incomes. Through this policy, governments try to decrease income inequality, and set a proper standard of welfare for all households. In this way the state tries to minimize the dependence on the market for lower income households (Esping-Andersen, 1990). Social-democratic welfare state types are therefore characterised by a high level of de-commodification. The government is responsible for the provision of many welfare services (Hoekstra, 2003) The Nordic states like Sweden and Norway are examples of traditional social democratic welfare state types.

The corporatist welfare state type is recognised by a less active role of the government in the provision of welfare services (Esping-Andersen, 1990). The government is supported by social companies and other institutions in the provision of welfare services. The traditional family is favoured in these systems and the welfare is distributed to preserve traditional roles. Countries from the European continent mostly have a form of a corporatist welfare state type, like Germany and Belgium.

8

In liberal welfare state types there is a dominating role of the market over the state (Esping-Andersen, 1990). The provision of welfare services is controlled by the market and state intervention is only meant for the very poorest. However, these welfare state types are characterised by relatively large poverty problems. Poverty mostly strikes vulnerable households which are not capable of maintaining themselves through the market (Lennartz, 2011). These households need help of the state which is mostly lacking. These welfare state types are mostly visible in Anglo-Saxon countries like the United Kingdom and Ireland.

Nonetheless, Epsing-Andersen (1990) did not incorporate housing in his research on welfare services. However, housing is seen as a universal right for everyone (Grander, 2017). Malpass (2008) argues that Esping-Andersen (1990) did not incorporate housing because of its very context specific and vague role in the welfare state, due to its relation to the market as a commodity.

2.2. H

As shown in the latter, the role of the welfare state is defined primarily on the basis on the provision of welfare services. These welfare services are: social security, healthcare, education and housing. These welfare services are traditionally characterised by their high level of de-commodification (Hoekstra, 2003). However, Kemeny (2001) and Harloe (1995) debate that housing is different from the other welfare services. Harloe argues that housing is the least de-commodified element of the welfare state. This is because housing has increasingly become a tradable commodity and an integrated part of the world economy. The financial crisis which started in 2007 displayed the influence of the housing market on the world economy (e.g. Schwartz, 2009; Harvey, 2012). Torgensen (1987) characterised housing as the ‘Wobbly pillar under the welfare state’, due to influence of the market and the dependence on the market in the provision of housing.

Harloe’s (1995) book ‘the people’s home’ was a very influential work on housing systems. Harloe was influenced by the convergence theory, which determined that the national housing systems would converge because of international housing policy and economic and demographic developments. Harloe used a neo-marxist approach to analyse the development of housing systems. He argued that the interventions in housing depended on the profitability of private investment in housing. Which means that in times of crises there would be a relatively large level of de-commodification, or state-built housing. In times of economic prosperity, intervention would be little and private investment would flourish, which means a high level of commodification. Harloe described two models of social housing systems, the mass model and the residual model. The mass model is characterized by high level of de-commodification and state intervention in housing. The residual model is a system whereby housing is only provided for the very poor. The Residual model was the ‘normal’ model according to Harloe. The convergence approach today is still relevant. Recent studies argue that the pressure of globalization and international competition will lead to a roll back of the welfare state, and lead to international liberal convergence (Genschel, 2009).

However, most comparative researches and frameworks on housing and welfare use a divergence approach (Hoekstra, 2010). This means that these theories acknowledge the importance of different social contexts in housing systems. Kemeny’s (1992, 1995) approach in comparative housing research has been very influential. Kemeny distinguishes two different types of housing systems on the level of corporatism in political structures. He does this by comparing rental systems of western European countries. These systems are either unitary or dualist, unitary meaning that there is a corporatist political structure. These systems are also known for large rental sectors and collectivist ideology. The dualist system is non-corporatist, has a small rental sector and a privatist ideology.

Theories that use the divergence approach show how different societal contexts influence welfare systems and create different realities. This approach also shows how the paths of welfare development differ per country. Path dependence analysis is increasingly used in social sciences over the last decades (Bengtsson & Ruonavaara, 2010).

9

‘Path dependence is often seen as the basic causal mechanism influencing various social and political processes in historical versions of institutional theory’ (e.g. David, 1985; Hall & Taylor, 1996; Mahoney, 2000; North, 1990;

Pierson, 2004; Putnam, 1993; Sewell, 1996; Thelen, 1999). The main idea is that a historical development or event has led to the choice of a certain path, because of this it is very difficult to take on another approach. Several academics have applied this method in analysing housing practices and policy (Bengtsson & Ruonavaara, 2010; Malpass, 2011; Boelhouwer & van der Heijden, 1992; Kleinman,1996:15). They state that the power of path dependence is that it can help to understand continuity. However, convergence theory claims that trends like neoliberalisation, economic growth and globalization cause systems to head in the same direction (Genschel, 2009).

2.3. W

Kemeny (1995) argues that there are some difficulties in the theoretical framework by Esping-Andersen (1990), Which makes it difficult. to be applicable to housing research. Kemeny claims that the use of corporatism by Andersen troubles analysis of housing systems. Kemeny uses a political science definition for corporatism, he emphasizes the direct relation between social arrangements in a society and the structure of a housing syst em (Hoekstra, 2012). Kemeny (2006) argues that the main difference in housing systems is that they are either more collective or private. Kemeny modified Esping-Andersen’s framework with two changes: labour led corporatism, which was profound in Scandinavian countries, due to the strong position of labour movements. Second, continental European countries were mostly characterized as capital led corporatist.

The typology of welfare states by Esping-Andersen (1990) can be very helpful in the analysis of housing systems. It was translated to use in the analysis of housing systems by multiple academics (eg. Hoekstra, 2003; Zhou & Ronald, 2017). One framework of the welfare state ideal types in housing is proposed by Hoekstra (2003, 2010, 2012), he has extracted the three main pillars of the welfare state and translated this to housing criteria, Hoekstra (2003: 61) argues that “the relation between the State, the family and the market determine which welfare services

are provided, how they are distributed and for which groups they are destined.”

Hoekstra (2003; 2013) states that the influence of welfare state regimes is determined by the relation between the State, markets, and the family, in housing. This is most evident in the organization of the production of new houses. This relation between the State, markets, and the family in housing is decisive in the de-commodification and stratification of housing. De-commodification in housing is the extent to which households are capable of providing their own housing, apart from their earnings on the labour market. De-commodification is represented by subsidisation and price regulation of housing. The subsidies can be subject subsidies, like income support, or object subsidies, which are specific for housing. Stratification, concisely said, is the way in which social housing is allocated. As an example, Liberal regimes are assumed to provide little price regulation and subsidies, furthermore, housing allocation would be regulated primarily by the market. Therefore, liberal regimes show vast signs of segregation and polarisation in the housing market, because social housing is only provided for the very poorest (e.g. Forrest & Murie, 1990; Hamnett, 2003). In social democratic regimes, there are many state subsidies, a highly regulated market, and allocation of social housing on the basis of need.

Ronald and Zhou (2017) have compared the housing systems of Chongqing and Beijing in China. This study has shown how different urban housing systems can be diverse in the same country, due to different contextual backgrounds. It is remarkable that they focused on cities, because most comparative research on housing is between countries. The framework they have used to analyse this can be found on the next page in figure 1, this framework is based on the framework by Hoekstra (2003). The framework by Zhou & Ronald (2017) is very clear and useful in analysing housing systems and welfare regime. The framework can help to display certain traits of a

10

system and it can help in the comparison of multiple cases. This framework will be used to analyse the differences between the urban housing systems and welfare regimes of Amsterdam and Stockholm.

2.4. N

-

Neoliberalism started to increase affecting housing policy in Western Europe around the beginning of the 1990s (e.g. Musterd, 2014; Hedin et al, 2012). It was characterized by privatization of social housing and decreasing subsidies and increasing home-ownership, which was encouraged by several governments, like in Sweden and the

TABLE 2: THE DIFFERENCES BETWEEN HOUSING SYSTEMS AND THE FOUR WELFARE REGIMES (ZHOU & RONALD, 2017)

11

Netherlands. Academics identify this trend as the rollback of the welfare state in European countries that used to have strong social policy (Dewilde & Ronald, 2017). Furthermore, due to globalization, international migration and changing demographics, the contexts of the welfare regimes have changed (Stephens, 2016). These developments have led to new patterns of stratification and asks for other interventions.

The concept of neoliberalism is widely used and widely debated. Brenner et al. (2010) explain it as market-based responses to regulatory problems, it tries to tap into new forms of commodification and always strives to make profit. Furthermore, Brenner et al argue that processes of neoliberalisation do not lead to convergence among different policy systems. This is different from, for example Genschel (2009), who claims that globalization will lead to increasing worldwide convergence. Nonetheless, Brenner et al describe how institutional contexts have led to growing spatial legislative differences among countries.

Earlier it was argued that housing is described as the most commodified welfare service (Malpas, 2008). Meaning that decentralization, deregulation and marketisation due to neoliberalisation will hit this welfare service particularly hard. An example of a neoliberal trend in housing is the promotion of homeownership as described by Ronald (2008). Due to increased homeownership, mortgages debts have increased in western European countries which has led to increased financialization of housing (Aalbers, 2016; García-Lamarca & Kaika, 2016). This financialization caused increasing real-estate prices which makes several housing markets increasingly inaccessible for young people and lower incomes. Therefore, the financialization has also led to new patterns of stratification (Stephens, 2016).

Neoliberal ideology has also contributed to the decentralization of housing policy. The central government confines itself in creating a framework for the local policy makers, this is the case in modern corporatist welfare states like the Netherlands (Boelhouwer & Hoekstra, 2012). Boelhouwer and Hoekstra (2012) ask the question whether more corporatist welfare regimes will develop into modern corporatist regimes, with a smaller role for the government and a bigger role for the market. However, recent dynamics show that despite decentralization, higher governments are still very influential in housing policy. One example is the new legislation for social housing companies in the Netherlands and public housing companies in Sweden, because of a conflict with the European competition law (Grander, 2017; Ronald & van Duijne, 2018).

Dewilde (2017:1) argues in her article that housing outcomes have become less associated with housing regimes, but rather the welfare state type and the country’s economic affluence. I focus on cities because they directly govern the production of new houses. Therefore, the influence of the welfare regime should be noticeable in cities. The recent financial crisis had shown how neoliberal characteristics like the marketisation of housing had caused evident imbalances in the world economy. However, Springer (2015) dismisses the claims and states that this would lead to an ending of neoliberalism, as neoliberalism is not a single entity which could come to an end. This supports the idea by Brenner et al (2010) that neoliberalism is not a uniform regime, that neoliberalism is not evenly implemented over countries and regions. “Cities in their analysis play a prominent role, as scales of

experimentation and variegation of neoliberalism.” (Gonzalez et al, 2017: 4). Brenner et al (2010: 184) state that

neoliberalism should be analyzed as a ‘systemic production of geo-institutional differentiation’ of market-oriented restructuring. This is the basis of variegated neoliberalism, meaning that institutional differences will show different patterns of neoliberal reforms in different places. Gonzalez et al (2017) did a research on the difference of governments reaction to the financial crisis and have shown how different national institutional contexts matter to this reaction. Nonetheless, they conclude that all cities they researched still grab back to pre-crisis strategies which have a highly market-oriented character. This research will demonstrate whether the urban housing systems and welfare regimes of Amsterdam and Stockholm are affected in a distinctively variegating manner.

12

2.5.

This chapter provided the theoretical framework for this research. The chapter was introduced by explaining the relation between relevant literature on welfare regimes and housing systems. It was explained how social democratic regimes are characterised by collectivist ideology and housing system. Liberal regimes are characterised by privatist ideology and a focus on the market. Hoekstra (2003) has described using his framework to link housing to welfare regimes that the ideal types are not meant to exactly illustrate a certain system, yet the ideal types have to be used to show difference in the traits of certain systems which can be linked to welfare regime theory.

The concept of convergence was demonstrated and debated by the idea of variegated neoliberalism (Brenner et al, 2010). However, the literature has shown that the global regimes and housing systems show strong resemblances. An example is the influence of the EU on housing policy. Nonetheless, this liberal trend does not mean that all regulatory systems are heading in the same way. This research will test the hypothesis whether the policy impact of neoliberalism is variegated because of institutional contextual differences. This will be done using the cases of the urban housing systems of Amsterdam and Stockholm. The contemporary urban housing systems of Amsterdam and Stockholm are compared using the framework proposed by Zhou & Ronald (2017). Using this framework, it will be possible to tell whether the variegation of neoliberal impact is showing in urban housing systems and their welfare regimes.

14

3. M

ETHODOLOGY

This chapter explains what methods are used to explore how urban housing systems and their welfare regimes have developed in the neoliberal era. This chapter starts off by explaining the comparative design and the main question of this research. After this, the conceptual framework and operationalization of the main concepts are demonstrated This is followed by an explanation of the methods used, and the data which was gathered to answer the sub questions. Subsequently, the criteria considering the validity and replicability of the design of this research are clarified. Lastly, the limitations to this research are presented.

3.1. R

This research will analyse the housing systems of Amsterdam and Stockholm. These cities are investigated because of their seemingly relatively comparable contexts. First, they are coping with severe pressure on their housing system. Second, cities largely control the organization of the production of new houses. Therefore, they can actually make useful policy to overcome housing systems problems. Third, these cities are both based in countries with social democratic and corporatist traits. Due to these similarities and seemingly comparable contexts, it is interesting to see whether these housing systems show signs of divergence.

This study is based on a deductive principle as it tests a hypothesis, namely that the policy impact of neoliberalism is variegated because of geo-institutional differences. The hypothesis is tested by examining the characteristics of the urban housing systems and welfare regimes of Amsterdam and Stockholm.

Ronald & van Duijnne (2018) have shown how the rental system in Amsterdam is moving from a unitary to a more dualist system. Van Duijne (2014) has shown that Stockholm shows a similar trend in the move towards a more dualist system. However, he debates that both systems also variegating patterns in the way in which reforms are implemented. Brenner et al. (2010) argue that geo-institutional contexts create different forms of neoliberalist implementation. This research displays if the variegating nature of the implementation of neoliberal reforms will appear by investigating the production of newly built houses in Amsterdam and Stockholm. The main question which is explored is:

To what extent are the urban housing systems and welfare regimes of Amsterdam and Stockholm diverging in the neoliberal era?

Comparative research is a widely use method among welfare regimes and housing systems (e.g., Kadi, 2014; Ronald & Elsinga, 2012; Lennartz, 2016; Ruonavaara, 2008; Ronald & Zhou, 2017) However, most of these researches focus on national systems. An example of urban housing system comparison is Van Duijne (2014), who focused on the comparison of Amsterdam and Stockholm, in their development from a unitary towards dualist rental systems. Zhou & Ronald (2017) compared the urban housing systems and welfare regimes of Chongqing and Beijing and displayed the differences between these systems. This research will draw on the method of comparing housing systems and welfare regimes from Zhou & Ronald (2017), by using their proposed framework, from table 1 (page 11).

Kadi (2013) stated that the research on cities is highly dependent on the just implementation of other scales, like the national and European scale. The influence of the national and the supranational scale are imbedded in this research by examining the macrodynamics that have shaped the housing system and welfare regime in Sweden

15

and the Netherlands over the last 30 years. These findings are the framework for the second sub question which examines the organization of the production of new houses in Amsterdam and Stockholm.

Finally, the outcomes of the first and second sub question will help to use the framework by Zhou & Ronald (2017). This framework will show how the housing systems and welfare regimes of both contemporary urban housing systems are constructed and how they relate.

The sub questions will guide this research towards answering the main question. As shown in the latter, the sub questions have a clear hierarchy in the way they contribute to answering the main question. The sub questions for this research are:

• How have macro-dynamics shaped the welfare regimes and housing systems in the neoliberal era in the Netherlands and Sweden?

• How has the organization of the production of new houses changed in the urban housing systems of Amsterdam and Stockholm in the neo-liberal era?

• To what extent are the urban housing systems and welfare contexts of Amsterdam and Stockholm diverging?

3.2. C

This section shows the relation between the different concepts which are explored. On the next page the conceptual framework of this research is illustrated. Figure 1 shows the relation between the independent variable and the dependent variables, and its attributes. The neo-liberal era is the independent variable, it is the defined from the literature. The urban housing system and welfare regime outcomes are based on their attributes. These attributes are derived from the framework for the analysis of housing systems and welfare regimes to compare cities, by Zhou & Ronald (2017).

FIGURE 1: CONCEPTUAL FRAMEWORK, OWN PRODUCTION

Neoliberal era Independent variable De-commodification Attribute 1 Stratification Atrribute 2 State-Market Family relationship Attribute 3 General housing policy objectives Attribute 4 Urban housing systems and welfare

regime outcomes

16

The definition of the independent variable ‘neo-liberal era’ is derived from the literature. In order to be able to use this variable, first a clear definition of neoliberalism is needed. Van Duijne (2014) has clearly defined neoliberalism as follows: ‘It prioritizes market-based or market-disciplinary responses to regulatory problems; it

strives to intensify commodification in all realms of social life; and it often mobilizes speculative financial instruments to open up new arenas for capitalist profit-making (van Duijne, 2014: 19; based on Brenner et al.,

2010; Harvey, 2005). The neoliberal-era in this research stands for the evident rise of the impact of neoliberal reforms on housing policy up until the present. It is widely agreed upon in housing research that neoliberal impact on housing policy vastly started growing around 1990 and has affected housing policy ever since (e.g. Christophers, 2013; Grander, 2017). The ‘neoliberal era’ is operationalized using the welfare regime framework by Esping-Andersen (1990), and other academics who have made it applicable to housing research (e.g. Hoekstra, 2003; Zhou & Ronald, 2017). By looking at the impact of the macro developments like national policy changes, signs of a neoliberal trend in the macro-dynamics considering housing is illustrated. The data which was used was qualitative data derived from policy documents, literature and expert interviews with academics.

Urban housing system and welfare regime outcomes are actually two concepts, namely housing systems and welfare regime. Nonetheless, welfare regime theory has helped to investigate the development of housing systems. Different forms of state support and relations between the state the market and the family can be found in housing systems, this state support, or lack of support says something about characteristics of the welfare regime and its housing system (Hoekstra, 2003). Several researches have used this link between welfare regimes and housing systems in international housing research (e.g. Stamsø, 2009; Venter et al. 2015; Ronald & doling, 2010; Ronald & Zhou, 2017). Hoekstra (2003) argues that the housing system is an essential part of the welfare state, therefore he states that the theoretical framework by Esping-Andersen can be applied to explain developments within housing systems. This research builds on this idea, and implements it to urban housing systems. The definition of ‘housing system’ in this research is based on the framework by Hoekstra (2003). As stated earlier, the organization of housing can be seen as a key service and characteristic of a welfare state. Hoekstra argues that the relation of between the State-market and the family is most important in the organization of welfare services. ‘The

relationships between State, market and family determine which welfare services are provided, how they are distributed, and for which groups they are destined. In other words, the mix between State, market, and family is decisive for the de-commodification and stratification in a country.’ (Hoekstra, 2010; 36). Therefore, Housing

Systems in this research is: the relationship between State, market and family in the organization of housing. The dependent variable, urban housing system and welfare regime outcomes consists of four attributes. These attributes are characteristics of housing systems and welfare regimes. These attributes are based on the framework by Zhou & Ronald (2017), see table 1 page 11. These attributes are measured using a mix of quantitative and qualitative methods. They are measured using statistics on housing, literature studies, expert interviews and extensive policy research.

3.3. S

This section describes the data and methods which were used to answer the subquestions of this research. Every sub question is first shortly introduced and summarised, after this the data and methods are explained.

M

-

N

S

The first sub question illustrates the macro dynamics which have influences the housing systems in Sweden and the Netherlands. This sub question is divided in different levels of influence, namely national influence and supranational influence. This division helps to show that there is a hierarchy of authorities which influence housing. The goal of this sub question is to set a framework of the main dynamics which affects the concerning

17

housing systems. Furthermore, this sub question will also help to show how neoliberalism has a variegating affect in on the national scale due to historical geo-institutional contexts (Brenner et al. 2010). Moreover, this sub question is important in integrating the national and supra-national in the analysis, which is important to be able to properly examining the urban perspective (Kadi, 2013).

To answer the first subquestion a mix of qualitative data was used. The first part is a historical analysis of the housing systems and welfare states of the Netherlands and Sweden in the time before neoliberal influences, before the 1990s. However, some academics debate that neoliberal reforms have started far before this time, mainly in Sweden (Christopher, 2013). This data was gathered through qualitative literature and policy analysis. Furthermore, eight interviews were carried out with different experts concerning housing in both systems. A list of the interviews is included in the appendix. The respondents were mixed in their expertise, for both the Netherlands and Sweden (Amsterdam and Stockholm). For both cases interviews were conducted with housing market experts from the municipality, CEO and head of asset management from the Public and social housing companies, and academics who specialize in both housing systems. The variety of the expertise of the respondents has contributed to the integrality of data.

H

A

S

The second sub question focusses on the frame of the main influences which originates from the first sub question. The influence on the urban housing systems by the macro system dynamics is measured in this sub question. This will show how the different scale of governmental bodies (national and supranational) have an impact on local urban housing contexts. This is done by analyzing changes in the characteristics in the production of newly built dwellings. As Hoekstra (2003) debates, the influence of a welfare regime is most evident in the production of newly built dwellings. This is especially the case when looking at municipalities, which in both cases are responsible for the zoning plans. The characteristics will tell how the macro influences have affected the housing systems. Therefore, it will be possible to tell whether the welfare regimes in different contexts are diverging or converging.

For this sub question a secondary quantitative data analysis was executed. The data which was used to analyze the characteristics of newly built dwellings is derived from multiple sources. For Sweden this data was gathered from the SCB, this is the national platform for statistics. For the Netherlands, the quantitative data is derived from the CBS (national statistics agency), The AFWC (local data on housing in Amsterdam) and from the O+S statistics department from the municipality of Amsterdam. The quantitative data was processed using different excel models, these are available upon request. There were three main topics in the quantitative data which is relevant when looking at the production of newly built housing, which were derived from the macro dynamics on the national scale.

• Tenure type: the development of the ratio of tenure shows a clear trend towards a preferred tenure type (e.g. Christopher, 2013). The preference of a certain tenure can be linked to welfare regimes. Therefore, this data was helpful in the analysis of the housing system.

• Sale of social (public) housing: the sale of social or public housing is considered to be ideology driven (Grander, 2017). Therefore, these statistics tell what welfare regime is present.

• Distribution of permits for new development: this distribution of permits shows what parties are active in the production of new homes. The nature of this party, either public or private, will display strong characteristics of the relation between the State the market and the family in the distribution of welfare services. And therefore, illustrate the welfare regime at place.

18

Furthermore, the quantitative data is supplemented with results from the interviews which are described in the first section, and policy analysis of both municipalities.

C

OF

A

S

The last sub question is a comparison between the contemporary urban housing systems and welfare regimes of Amsterdam and Stockholm. This last sub question builds on results from the first two sub questions. The last sub question will tell what differences and similarities there are in the implementation of neoliberal ideology in the local urban policy of Amsterdam and Stockholm. This will be done using the framework by Zhou & Ronald (2017). The data used to ‘fill in’ the framework originates from the first two sub questions. The completion of the table, which is also shown in table 2 on page 43 will display the similarities and differences in the characteristics of the contemporary housing systems and welfare regimes of Amsterdam and Stockholm.

After the completion of the sub questions, I was able to answer the main question and defend the hypothesis.

3.4. G

There are multiple considerations which have to be thought off, in order to secure the validity of a research. This section elaborates on the general criteria considering the validity of this research.

First, the internal validity is referred to when a research is able to demonstrate that the there is a link to a certain outcome and the object which is studied (Bryman, 2012). The research hypothesized that neoliberalism has different effects on local housing systems due to geo-institutional differences (Brenner et al, 2010). This outcome is studied through the triangulation of data. With the use of quantitative housing statistics, expert interviews and policy/literature studies to examine the hypothesis. The triangulation of the data enhances the robustness of the research (Yin, 2012). The research design which was chosen was carefully constructed, to be able to generate a proper internal validity.

The external validity refers to the applicability of the findings in other contexts (Bryman, 2012). This research focused on the specific contexts of Amsterdam and Stockholm, therefore the outcomes of the specific contexts do not apply to other contexts. However, the outcomes of this research say something about the impacts on western European urban housing systems. If you look at is from that perspective, one could argue that the variegating effect of neoliberalism on housing will also show in other western European urban housing systems, due to the differences among local contexts.

Finally, the replicability is arguably the most important criteria in doing research. The replicability refers to whether a research is conducted in an unbiased manner (Bryman, 2012). This chapter has carefully described the methods which were used to collect the data, and the way in which this data was processed. The transcriptions of the interviews and spreadsheets of the quantitative analysis are all available upon request. Furthermore, this research has used the existing framework by Ronald & Zhou (2017) to process analyze the data, the use of this framework contributes the replicability. As it transparently illustrates the choices and interpretations which were made in the analyses.

19

3.5. L

This research is conducted in a relatively short period of time. Therefore, this research has some limitations which has an undesirable influence on the reliability of this research. Firstly, eight experts with different specialization were interviewed for this research. This is a relatively small group, which means that more interviews with other respondents could have led to other outcomes. Nonetheless, the triangulation of data secures that the research does not solely rely on the results from the interviews. The results from the interviews were mostly used as a guidance in the research and were very helpful in illustrating the different contexts.

Furthermore, the statistical analyses was conducted using secondary data. This means that this cannot validated. However, this data was gathered by the SCB, CBS, AFWC and O+S Amsterdam, these organizations are known for their extensive contribution in the provision of data, therefore, the validity and completeness of this data can most probably be relied upon.

21

4. M

ACRO

-

DYNAMICS UNDER NEOLIBERALISM

This chapter is an analysis of the macro developments that have shaped the Dutch and the Swedish Housing systems and their welfare states. These processes are prone to powers from political economic regulation, demographic developments and ideological transformations. This chapter will examine how macro developments have affected welfare contexts and housing systems under neo-liberalism. This will be done chronologically and by looking at two levels of governmental bodies; the national government and the EU. Finally, the macro changes of both countries will be compared. By doing so, it will become clear how both systems have developed differently under the neoliberal trend in the world over the last 30 years. This chapter will answer the question: How have

welfare contexts and housing systems developed under neoliberalism in the Netherlands and Sweden? The answer

to this sub question will create a framework for the second sub question which looks at the changes of the characteristics of the development of new housing in the capitals of Sweden and the Netherlands. In conclusion, this chapter is going to examine how neoliberalism has affected the housing systems and welfare regimes of Sweden and the Netherlands, in the following chapter the impact of neoliberal reforms on the local urban housing systems and welfare regimes will be examined.

4.1. T

N

T

D

Since the end of the second world war, the Netherlands has been very active in the provision of housing for its inhabitants (Kadi & Musterd, 2014). The Dutch housing system was highly regulated and was recognized for its high level of de-commodification. Up until the 1980s housing production in the Netherlands was dominated by the non-profit social housing corporations. Moreover, the government heavily subsidised the supply side of housing. Social housing was not only destined for the lowest income groups, but for a much broader set. Social housing was therefore not stigmatised or only destined for the very poor. Unlike the US, where social housing was only meant for the very lowest incomes. The Dutch government provided support by giving object subsidies for the production of various types of new homes (Hoekstra, 2003). Moreover, the government provided subject subsidies on such a large scale that most people were eligible for these subsidies, not only the poor. It is thus argued that, the abundance of the rental subsidies showed the strong social democratic character of the Dutch welfare state (Doling, 1999).

However, the Dutch housing system has also been recognised for its corporatist traits. This was displayed in the strong interference of the government in owner occupied housing (Hoekstra, 2003). The Dutch government provided premium house purchase regulation (Premie-koopregeling) which were available for certain households. The providing of such premiums in order to make the owner-occupied housing market accessible for lower-middle incomes is a typical characteristic of a corporatist regime. The subsidies were provided and distributed by the central government. The regulation also stimulated the production of new houses that were in the range of households with lower-middle incomes, who could then buy a house due to the premium house purchase regulation, this is a characteristic of a corporatist regime (Hoekstra, 2003; Zhou & Ronald, 2017). Moreover, the municipality also had a large role in the allocation of houses. Households could get a preferred status in the allocation in a neighbourhood if they had favoured social economic relations (passenheidscriteria) (Hoekstra, 2003). This regulation was maintained in order to control the social economic characteristics of neighbourhoods and cities.

The characteristics displayed in the latter show the hybrid nature of the Dutch welfare system in housing. A social democratic system with corporatist traits (Esping-Andersen, 1990). The large government influence shows that the government saw housing as a welfare good, basic need and their responsibility. People in the Netherlands in

22

general could all get proper housing without being dependent on the market. The consequence was a large de-commodified share of housing. However, housing was a big expenditure on the countries’ economic balance. By the end of the 1980s and the beginning of the 1990s the government considered reformations which would lead to less government responsibility and expenditure.

N

In the 1990s economically, liberal influences matured in affecting the Dutch housing system and welfare regime. Many of the traditional social democratic traits vanished due to neoliberal influences. Corporatist traits however were getting more influential. In 1989 the Dutch government introduced the policy document housing in the 1990s (volkshuisvesting in de Jaren 90). The policy document was the start of decentralisation and privatization of many state responsibilities in housing (van Duijne, 2014). Due to the changing role of the central government, the new function was to create a framework for local governments and private companies to act in. This was a transformation from direct to indirect state interference in housing (Hoekstra, 2003).

The large-scale object subsidies were cut in the beginning of the 1990s, as well as the premium house purchase regulation. In 1995 the government cleared out all the subsidies and loans by buying all the outstanding obligations and object subsidies (Hoekstra, 2003). This process was called the bruteringsoperatie. Unlike the object subsidies which were abolished, the subject subsidies remained (huursubsidie). This subsidy was mainly aimed at lower incomes, nonetheless with 800.000 eligible households, this still accounted for a relatively large share of all households. Nonetheless, the focus on subsidising lower incomes shows a move away from a social democratic regime, the abolishment of object subsidies is a character of a liberal regime (Zhou & Ronald, 2017).

In the 1990s corporatist traits of the Dutch welfare regime grew vastly. However, as stated above the object subsidy of the premium house purchase regulation was abolished. This corporatist policy was replaced by budget subsidies for local authorities. The idea of this policy was that through the decentralisation of subsidies, local authorities could stimulate local initiatives. Local governments had to work within the framework regulated by the central government, but they could decide for themselves what they wanted to subsidise. Hoekstra (2003) states that this is a typical corporatist trait. State control over rental prices also transformed in a more indirect way of controlling. Furthermore, the neoliberal wave also influenced the social housing sector. Social housing associations were decentralized and privatised in order to increase efficiency (de Jong, 2013). However, they maintained their state security in case of financial problems. Due to the privatization, the social housing companies gained a strong financial position. At the end of the 1990s the social housing companies were financially privatised which broadened their capabilities and strengthened their financial position (Koster, 2015). The social housing companies increasingly engaged in commercial activities to finance the social housing. Consequently, these companies became highly influential players on the real-estate market. There was a shift from providing to social housing towards the maximization of profits, the social housing companies therefore acted like businesses (Blessing, 2012) The allocation of housing in the Netherlands remained its corporatist character, allocation rules for people from different classes and statuses. In 1993, the Housing law (huisvestingswet) was introduced, this policy ensured the allocation on the basis of certain conditions, a typical corporatist trait (Hoekstra, 2003; Zhou & Ronald, 2017). The policy can be seen as a correction to certain market circumstances, like inaccessibility of housing for certain incomes in a neighbourhood, in order to maintain certain social economic relations. The rental market was partially regulated by the market. Dwellings with a rent below 565,- gilders had maximum rental increase which were controlled by the government. The liberated expensive rental market however did not have any controlled maximum price rise. This displays the liberal transformation, because prices on the expensive rental market were decided on the basis of supply and demand.

23

The 1990s were a decade of transition for the Dutch housing system. The developments show a move away from social democratic traits, like the abolishment of the object subsidies. Furthermore, corporatist traits were amplified through the decentralization of state responsibilities to local governments. However, one can argue that the decentralization of state responsibilities also characterises a liberal regime. The way in which housing associations started acting is a clear sign of a liberal turn, as they were privatised and social housing was therefore ‘marketized’. In 2001 the Dutch government introduced the ‘what people want, where people live’ policy document. Kadi & Musterd (2015) argue that this policy, as well as the ‘volkshuisvesting in de Jaren 90’ gave way to a stronger market orientation of the Dutch housing system. One of the main pillars supported by the Dutch government was the promotion of home-ownership. This was ‘subsidised’ by the central government by providing the deductibility of mortgage interest. In this way homeownership was favoured and stimulated. “There is basically an embracing

of the growth of homeownership, with subsidies and policies that support the sector and promote the increase of house prices” (Kadi & Musterd, 2015; 251). The dynamics show a transition of subject subsidies for lower

incomes towards subsidizing home-ownership, and therefore the relatively higher incomes. The total size of the deductible mortgage rents in the Netherlands was €14 billion in 2006, against €3 billion in total for rental subsidies (Conijn, 2008). In 2001 the promotion of homeownership was secured in the Homeownership act (BEW). In this act the goal was set to increase homeownership in the Netherlands from 53% to 60%. This was primarily promoted in the big cities, where there was still a very large share of social housing. This mainly had to be done through the demolition of cheap social housing and the conversion to owner occupied housing (Van Kempen & Priemus, 2002).

One of the latest national regulations is the ‘landlord levy’. This is a policy whereby social and private landlords are renter stimulated to increase rents in price-controlled dwellings, with increased revenues being taxed to offset lost government revenue derived from reductions in transaction taxes for homebuyers (Kadi & Ronald, 2014: 279). The reforms in the Netherlands have contributed to the marketisation of housing. This has led to a decline of inexpensive units in the housing stock and thereby enhanced the inaccessibility of the housing market for lower incomes.

I

E

As stated earlier housing associations increasingly engaged in commercial activities, due to the privatization (Blessing, 2012). However, the fact that the state supported market actors, which the social housing companies were, led to complaints from other market actors. Complaints were raised that this support was in conflict with European Competition law (Elsinga & Lind, 2013). This resulted in tighter regulation for the housing associations. They had to split their commercial activities from their social housing activities and had to focus on providing and maintaining social housing. Moreover, the social housing corporations were important actors in the changing tenure preference of the Dutch government. These social housing companies were very influential actors in the housing market as they controlled and possessed a very large share of the Dutch housing stock. By selling part of this stock, the social housing corporations could help the States goal of promoting homeownership (e.g. Aalbers, 2004; Aalbers & Holm, 2008; Kadi &Musterd, 2014)

The beginning of the 2000s was characterized by a strong neoliberal trend in the Dutch Housing system. This was clearly visible through the promotion of homeownership and the sale of social housing. Thereby, the provision of social housing through the privatized social housing companies. The still relatively large share of social housing in the Netherlands, secured the social democratic character of the country. However, all macro changes show a turn towards a regime with more liberal and corporatist traits.

In 2011 new legislation showed a particular turn in the Dutch social housing sector (Kamerbrief Woonvisie 2011). New regulation measures stipulated that 90% of all social housing should to be allocated towards lower income

24

households, with incomes below €34.085 (Ronald & van Duijne, 2018). This shift shows a turn from corporatist character towards a more liberal housing regime, because this means that the social housing is now only destined for incomes who have difficulties or are unable to find housing on the market (Zhou & Ronald, 2017). This legislation was introduced because the government felt that the social housing sector was too big. The government increasingly saw social housing as a safetynet for the poor, which shows the gap between social housing and market housing (Forrest and Murie, 1988). This also caused that social housing has become increasingly stigmatized (Elsinga & Lind, 2013). This turn was also driven by the continuation of the promotion of homeownership. By selling of social housing, home-ownership increased and social housing was increasingly meant for the lower target group. Van Duijne (2014: 62) states that this turn in the allocation of housing, and the promotion of homeownership shows that: ‘housing as a social structure is increasingly becoming more private

than collective’. This illustrates both a liberal turn in the Dutch housing system, as well as a turn towards a more

dualist rental model (Kemeny, 1995).

As a result of the interventions by the European commission about the housing associations and their interference with competition law, the Dutch government-imposed restrictions on the housing associations (Elsinga & Lind, 2013). These restrictions were reinforced after the right winged liberal Cabinet (Rutte I) came to power in 2010. This liberal coalition imposed new laws which had several goals. Not to only to comply with the requirements of the European commission, but also to limit the actions of the associations. The reason for these measures were the problems caused by the housing associations. As explained earlier, housing associations had to act in a more businesslike manner in order to be able to finance the production of social housing and the regeneration of urban areas (Ronald & van Duijnne, 2018). However, regardless of the deregulation and privatization of the housing associations, the associations were still responsible for the social housing objectives in the Netherlands.

In 2012, the former biggest social housing association in the Netherlands Vestia, was on the verge of collapse (Winter, 2016). Apart from commercial real estate, some social housing companies invested in other commodities. Vestia had engaged in investing in high risk financial products. This speculating in highly insecure financial derivatives led to a total loss of 2.7 billion euros (Köning, 2016). Vestia was saved by the national government and the social housing corporations, who paid a combined 670 million euros to save Vestia. Other problems were several examples of self-enrichment by high ranked employees of the associations and questionable investments. Köning (2016) argued that these harms were caused by a lack of regulation, due to the privatization and deregulation imposed by the government.

The most recent housing law was introduced in 2015 (‘De woningwet 2015), the main goal of this new law was to minimize financial risks and to introduce new playing rules for the Housing Associations (Rijksoverheid, 2015). With this new law, a new supervising body called the authority for housing associations was introduced (autoriteit

woningcorporaties). This new body supervises the governance in the Housing associations and annually reports

on the development of the sector. The Woningwet 2015 has four main principles (Rijksoverheid, 2015):

• First, the housing corporations have to focus on their main task, providing social housing. The corporations are however allowed to develop new projects under very strict rules. First of all, in case a corporation wants to develop new property, there must be no private parties who are interested in the development, because they are not allowed to compete with the private market. This is analysed through a ‘market-test’. Moreover, the social housing corporations have to pay regular land prices, and the development has to pass a prior efficiency test.

• Second, the social housing companies have to restrict their focus area on municipality or small region. The Social housing companies contribute to the housing strategy of municipalities. They set agreements with these municipalities over their housing goals. This shows the decentralization of policy, a character of a corporatist regime (Zhou & Ronald, 2017).

25

• Third, social housing corporations had to separate their commercial portfolio from their social housing portfolio. Hereby, the state secured the fact that social subsidies are not used for commercial investment. • Fourth, governance is of great importance in the new law due to the prior problems with the management of the corporations. New board members have to pass a certain appropriateness test, and they are not Allowed to act in certain side activities, which might cause a conflict of interest.

Head of Asset management at the biggest social housing company in Amsterdam (Ymere), claims that: “The

intervention of the European Commission came in handy for the national government. They had to step in to regulate. We understand that, and agree on the fact that regulation was needed. However, the capacity in which they imposed new regulation is increasingly showing to be contradictory. We have little to no space to invest, and the tests we have to go through and the increased operational costs we have are unbelievable.”

As shown in the latter, the respondent argues that it has become almost impossible for social housing associations to develop new property due to the new legislation, primarily in places where market demand is high. Furthermore, a recent report of Aedes has shown that the new rules have tripled the administration costs for the Housing associations (Aedes, 2017).

In October 2015 the Dutch government revised the point system for the appreciation of rental dwellings. Social housing dwellings are priced based on points, a dwelling used to only get points for its size and the services in the dwelling. Since October 2015, the market value (waarde) has become part of the point system. The

woz-waarde is the estimated market value of a dwelling, based on transactions of comparable dwellings in the

neighbourhood (Rijksoverheid, z.j.). Moreover, when social dwellings reach above a certain amount of points they are liberalized and are not social dwellings anymore. These have to be sold, because social housing associations are allowed to engage in commercial activity under strict regulation. Due to the implementing of woz-waarde in the point system, more dwellings are liberalized. Also, these dwellings are mainly located in the better areas, because woz-waarde is higher in these areas. This is contributing to the allocation of housing based on income, therefore this is a clear liberal turn in Dutch housing system and its welfare regime.

The Dutch government changed the point system to increase the liberate units for the so-called mid-income problem. In the Netherlands there is a shortage in housing for middle incomes. Middle incomes are incomes above €34.085, which is the border for incomes eligible for social housing (Ronald & van Duijne, 2018). This shortage is created because of the fact that middle incomes, are not eligible for social housing, and market prices are too high for them. Therefore, the Dutch government thought It was a smart to liberate the more expensive social housing, in order to facilitate middle incomes. This addition of woz-waarde in the point system is reinforcing social stratification because it diminishes social housing from the more expensive areas. A housing market researcher for the municipality of Amsterdam, argues that this policy is driven by the ideology of the ruling coalition, she claims: ‘Blok just says, the left cities have to stop complaining about problems with social housing,

because those have long been solved. The people can live on the outskirts of the city just fine if there is no place in the cities for social housing anymore.’ This argument strengthens the claim that the state’s welfare regime is

getting more liberal. However, the claim was made by a representative of the municipality of Amsterdam. Therefore, this context is important for the influence of the claim. Nonetheless, academics like van Ronald & van Duijne (2018) have shown how social housing is pushed towards the outskirts of the city, and is increasingly being segregated.

The recent macro changes in Dutch housing show a clear liberal trend in the Dutch housing system. First, Housing is increasingly linked to income. Furthermore, Due to the new regulation, housing associations are almost unable to develop new property. Lastly, social housing is increasingly stigmatized due to the maximized income, and the fact that it has become a safety net for the poor. Ronald & van Duijne (2018) show that the Dutch rental system in recent years has gradually been sliding towards a more dualist system, due to the unbalancing of the social rental

26

sector and the private market. The new legislation, for example on the woz-waarde, demonstrates a clear trend towards a more liberal housing system and a liberal welfare regime in the Dutch central government.

4.2. S

T

S

“A country where everyone lives in decent housing which costs less than 25 per cent of their income? A country where everyone has the choice between rental housing, cooperative housing, and single- family housing? A country where lack of individual financial resources does not relegate one to living in central city slums? A dream world? No. Affordable, quality housing for all exists in Sweden. “ (Kenn, 1996:63).

The quote above illustrates how the Swedish housing model symbolizes the social democratic welfare regime. Housing in Sweden until the end of the 1980s had been a core part of their well-known social democratic Swedish welfare state (Turner &Whitehead, 2002). However, many academics have shown how changes towards a more liberal housing system have been going on in Sweden since the 1960’s (eg, Chritophers, 2013; Lindbom, 2001; Lind, 2007).

The Swedish housing system has some unique features. Starting off with the bostadsrätt, which is neither a form of the rental nor owner-occupied sector (Christophers, 2013). The person who lives in such a dwelling is part of a cooperative called the bostadsrättförening. This cooperative owns the block of dwellings, which a single

bostadsrätt is part of. The person who lives in the dwelling pays a fee and has the right to the dwelling for an

unlimited period and can transfer this right to a new person. This system originates from the pre-war era in Sweden, however it was not institutionalized until after the first world war. The Tenant-Ownership Act of 1930 and the Tenant-Ownership Control Act of 1942 were key for this, as it introduced price restrictions. The owner could only sell the right to the apartment based on share of the cooperation’s assets. This price was based on the taxation value of the property, which in turn is based on the rental income of the cooperation. Therefore, the owner could not get market prices for the dwelling. Therefore, this form of housing was a de-commodified form of housing which symbolized the Swedish model (Ruonavaara, 2005: 221). This form of housing on the basis of collectivism and de-commodification shows the social democratic character of the system.

However, in 1968 the Tenant-Ownership Control Act of 1942 rules ended. The shareholders of the

bostadsrättförening could sell their right to the dwelling on the basis of market prices (Ruonavaara 2005: 223).

Before the liberalisation of the bostadsrätt transfer, the dwellings accounted for 9% of the housing in Sweden. By 1990, this share had risen to 15% of the total housing stock. This process was also accelerated by the deregulation of the credit markets in Sweden (Christophers, 2013). Prices in Sweden increased in average by 8% per year from 1980 to 1990.

The Swedish social housing sector emerged like in many other countries between the two world wars (Christophers, 2013). However, social housing in Sweden is named public housing, because it is not restricted to certain income groups, it is however provided by the government. This lack of restriction to income exemplary for the inclusive character of Swedish housing and welfare (Grander, 2017). Social housing in Sweden is managed by municipal housing companies, these companies are owned by the municipalities (Bengtsson, 1994). Between 1965 and 1974, the public housing sector accounted for 40% of the one million homes built in that period. The public housing share stabilized at approximately 25% of the total stock.