Who helps the students : welfare state or free market? : comparative research on welfare state regimes and student housing outcomes

(1)

June 26

Second reader: Dhr. Dr.

David Evers

(2)

Acknowledgements

Working on this thesis has been a challenging and demanding but in the end very satisfying experience. It would not have been possible without a number of people who helped me along the way. First and foremost, I would like to express my gratitude to my supervisor Assoc. Prof. Dr. Tuna Tasan-Kok for the great amount of time and effort she invested into my research ideas and me. I have been very fortunate to have such a motivating, encouraging and knowledgeable supporter, who always provided helpful and constructive feedback. Tuna introduced me to comparative case study research and taught me a lot about academic writing. Tuna, thanks for all the attention and guidance, and for the trust you put into my research ideas, which motivated me and helped to overcome my doubts. I want to express my gratitude to my internship supervisors Manu Moritz and Wouter Onclin, from The Class of 2020, who, first of all, inspired me to start research on student housing and gave me a chance to make my research findings practically applicable. They helped me not to doubt the usefulness of what I have been doing. Manu from the very beginning was deeply engaged with my work; he helped me to develop a great deal of networks and provided contacts of my interviewees. I’m very thankful for his valuable guidance and support. I am also grateful to my fellow classmates from the same group project Amal, Steven, Mike and Therri for their constructive comments on my topic idea, for always being around to share some moments of frustration, and not least, for the good time.

(4)

1. Introduction

Throughout Europe there has been an increasing imbalance between the supply and demand of student housing due to the growing student population, tightening urban real-estate markets, dismantling of welfare state functions and increasing involvement of private sector. Demand for international higher education continues to grow. Recent years have shown a rapid rise in the number of higher education students globally. UNESCO reports a 49% increase from 2005 to 2014, with the total number of students enrolled in tertiary education worldwide now reaching 207.5 million (UNESCO, 2017). Unlike the rest of the world, Europe is experiencing a lower growth of domestic students due to lower birth rates, but it is offset by increased international student enrolment. Europe already attracts 35% of all international students worldwide (The Class of 2020, 2017).

Apart from the quality of education, students also consider other practical matters when choosing the study destination, notably living costs, accommodation costs and accessibility of affordable places to live. Many European cities, especially big university hubs, have a severe shortage of student accommodation, which is affordable and of good quality. This leads to a situation where students either drop out from their studies, choose not to enroll in programs or find themselves renting small rooms in old, uncomfortable, and possibly unsafe buildings. Apart from that many domestic students continue to live with their parents despite preferring to live independently (The Class of 2020, 2017).

In the past, many countries relied on the state to provide resources as the primary means to ensure student housing supply and students’ financial support. However, since 1970 a shift towards neoliberalism has occurred in political economy of welfare states and brought deregulation, privatization and withdrawal of the state from many areas of social provision, including student housing and education (Harvey, 2005: 3). Nowadays the traditional government role has shifted from that of housing providers to housing enablers and regulators. Consequently, welfare state policies that guaranteed favorable housing subsidy systems have been gradually fading, so the solution needs to be found elsewhere. Generally this means that the private sector and PPPs will be the ones to take over the state’s functions in order to meet the demands of growing student population.

Tremendous demand for student accommodation has raised investors’ and developers’ interest in the student real estate market. Savills reported a record level of investment in student housing across Europe in October 2016 (Savills, 2016). International cross-border investment in the student housing sector accounted for 40% of all deals in the last three years. The majority of international investments have been made in the UK, which has the most unregulated European housing market. Purely market-driven student housing provision and liberal policies create a situation, where more and more private actors are entering the niche. It results in more beds available for students, but since private actors are seeking for profit, these options are less affordable. Contrastingly, countries with a strong state’s role, extensive housing regulations and generous welfare benefits succeed in providing social student housing of good quality and for affordable prices. However, they are facing challenges to attract international investors because tightened market rules do not allow private developer to make their business models work. Consequently, such countries are facing housing shortage, because social housing corporations are not able to provide sufficient numbers of accommodation and the private sector is not active.

The discussion so far suggests an evident tension. On the side of an unregulated housing market, the dominant role of the private sector with market-driven housing provision ensures the sufficient supply of student housing. But a high level of privatization together with a lack of welfare benefits produce high rent prices, and hence a low affordability level. On the other hand,

(5)

housing markets with extensive construction and rent regulations scare away private investors, who could fill up the housing supply and deal with shortage issues. But the existence of a social housing sector together with government benefits towards students’ rent and incomes increases the affordability of student housing.

Using the welfare regime theory of Esping-Andersen adjusted by Hoekstra I make an attempt to understand why certain countries provide relatively higher levels of affordability and accessibility than others. This research addresses the following question: How does a welfare state regime influence student housing outcomes: affordability and accessibility? To answer the main exploratory question two sub-questions are pursued through the literature review and cross-country case studies:

• What is the role of the welfare state in student housing provision?

• What are the affordability and accessibility levels on student housing markets?

• How do different welfare state regimes address the problem of affordable and accessible student housing? What tools do they provide?

The goal of this research is to produce knowledge about the role of the welfare regime in configuration of student housing outcomes in the European context and compare the differences in affordable and accessible student housing provision between different welfare state regimes.

To grasp the different approaches to affordable and accessible student housing provision I use a multiple-case study research design. This type of design is common within housing studies, particularly those investigating the role of a welfare state model in urban housing markets (e.g., Kadi, 2014; Kadi & Musterd, 2014; Ronald & Elsinga, 2012; Lennartz, 2016; Ruonavaara, 2008). Welfare regime theory provides a ready comparative framework that justifies a selection of cases, in which each new country represents a type of welfare regime. In order to determine the differences in student housing provision in different welfare regimes, I choose representative countries from two extremes: social democratic and liberal. I will compare the liberal welfare state regime, with the most privatized student housing provision and unregulated student housing market, and the social democratic, with the biggest share of public provision of student housing and extensive regulations. As a third case study, I analyze the transitional regime, located between the liberal and social democratic and incorporating the characteristics of both. As the data collection method, I use a qualitative method of analysis and conduct semi-structured interviews with experts from the sector. I analyze the data using thematic analysis.

In this thesis I support the vision on housing as an important and integral part of the welfare state and assume that housing is embedded as an element of the welfare state. This position represents a divergence approach. The divergence approach implies that social reality is context dependent, which is important for cross-national comparison (Hoekstra, 2013: 6). Applying this approach to the welfare regime theory means that all welfare states follow different paths that determine their future development and bring a certain degree of path-dependency. Of course, general societal trends have moved welfare states towards a common direction: economic growth, neoliberalization, internationalization, etc. Nevertheless, every country reacts to these trends differently, which is why despite seeing some convergence trends, I follow the divergent approach in the welfare state regimes and housing systems.

My findings contribute to several broader debates. First, I continue the discussion about the neoliberal transformations of welfare states, by providing concrete evidence on the transformations in the student housing sector. Secondly, I take a novel look at housing outcomes, expanding the widely used concept of affordability with accessibility measurements. The term affordability has been one of the topical policy indicators to the central concern of housing policy and a key theme of contemporary housing studies. In my opinion, affordability as

(6)

a measure for housing outcomes is not sufficient on its own as it does not capture the dimension of housing access. In this thesis I apply a more comprehensive approach to housing outcomes by combining two interdependent but substantive concepts. Last but not least, derived from the interviews with the key experts in student housing sector from three different countries my findings reveal the crucial trends for improving accessibility and affordability of student housing in Europe. The findings of this thesis will be used for the blog articles and an annual

(7)

2. Theory and hypotheses

In order to give an insight into the role of welfare regime in student housing outcomes it is first necessary to explore the key theoretical concepts. This chapter begins with integrating existing knowledge of the welfare state regime theory and the position of housing in it. Next, the characteristics of student housing markets are given, followed by an overview of student housing outcomes: affordability and accessibility. At the end of the chapter five hypotheses are developed for empirical testing.

2.1. Welfare state regimes. First classification by Esping-Andersen

Models of society have been developed to generalize about the diverse values and practices that shape the relationship between the state, market, and citizens in certain nations. The ‘ideal’ types generalized from these social models are used to compare and explain the more complex reality in particular states and regions. A number of typologies of social models have been distinguished in different spheres: policy, planning, market, welfare, housing systems, etc. The book of Gøsta Esping-Andersen ‘The three words of welfare capitalism’ (1990) has sparked interest in welfare state regime classification. His comparative analysis of welfare states has led to clustering the systems into ideal types – welfare regimes. According to him, a welfare state regime is a complex system of neatly intertwined arrangements between state, market and the family (Esping-Andersen, 1990). Gøsta Esping-Andersen distinguished three ideal welfare regimes: the Anglo-Saxon liberal, the Nordic social-democratic and the corporatist regime on the European continent. Based on his analysis of 18 OECD countries, Esping-Andersen clustered them around these three types, which fundamentally differ from each other. Esping-Andersen distinguishes three keys to welfare identity: de-commodification (“the degree to which individuals or families can uphold a socially acceptable standard of living independently of market participation”), social stratification (the role of welfare states in maintaining or breaking down social stratification) and employment (Esping-Andersen, 1990). In short, he states that different interactions between state, market and the family create distinctive systems that possess attributes relating to de-commodification, social stratification and employment (Stephens, 2016).

The liberal welfare state type is characterized by a dominant role of the market in welfare provision and a weak position of the state in structuring citizens’ social and economic life. The core idea is that direct state interventions are minimized to help only those in need (the poorest of the poor); the welfare state is left to the “free” market. That is why the supply of social security is characterized by a strong private involvement (Esping-Andersen, 1990). Liberal welfare states see the most effective way to deal with social exclusion in targeted in-work benefits rather than social insurance. As revealed by high poverty rates, liberal welfare models perform poorly in dealing with equality issues (Lennartz, 2011). Poverty is especially evident among vulnerable households, such as lone mothers, young families with children, and others (Esping-Andersen, 1947: 15). Typical representatives of this model in Europe are the UK and Ireland. The social-democratic type, on the contrary, leaves a dominant role to the state. The main idea is to uplift the poor to a middle-class standard with comprehensive state intervention. This is achieved by granting welfare transfers on a middle-class standard universally, so that the poor will benefit more from these transfers leading to a relatively low degree of income inequality. Welfare in this type of regime is the most de-commodified, meaning that the level of household’s welfare does not strongly depend on household’s labor market position (Esping-Andersen, 1990). Minimizing the dependency on the market though is followed up by introducing a policy

(8)

of maximizing citizens’ employability and productivity (Esping-Andersen, 1947: 12). Typical examples of social democratic regime are the Nordic countries: Sweden, Finland, and Norway.

According to Esping-Andersen’s typology the corporatist type of welfare regime is a mix of state and market in the welfare provision and stands in between liberal and social democratic systems. The main peculiarity of this type is the strong role of family. Welfare is given to people according to their specific status to preserve the traditional status differentials; rights were attached to class and status. Therefore, this welfare system is very fragmented and does not pursue a universal social security system (Esping-Andersen, 1990). Typical representatives of this type in the Esping-Andersen’s classification are Germany, Belgium, Austria and Italy.

The study ‘Three worlds of welfare capitalism’ has provoked an ongoing debate in the comparative social policy literature and sparked further research on typologies of welfare states. This process has led to the development of alternative typologies, many of which extend the range of countries included in the analysis of Esping-Andersen, by using different criteria for typology construction or a different range of variables accounting for gender, politics, and the role of public services (Bambra, 2007). The most fundamental critiques on Esping-Andersen’s work include the range of countries used to construct the typology and the lack of consideration given to gender as a form of social stratification. Additionally, the decision to organize the classification around the study of social transfers such as pensions and sickness and unemployment benefits has been debated as well and the idea of a static composition of the welfare state regime has been questioned. After the de-commodification approach of Esping-Andersen (1990), scholars tried to use different criteria to construct welfare state typologies including basic income (Leibfried 1992), poverty rates (Ferrera 1996; Korpi, Palme 1998) and social expenditure (Bonoli 1997; Korpi, Palme 1998). Summaries of these approaches are well presented in the article of Nadin & Stead (2008).

In the original Esping-Andersen’s typology only 18 OECD countries were included in the analysis, leaving out all South European countries, except for Italy, which was classified under the corporatist model. Following the critiques on Esping-Andersen’s classification several researchers suggested distinguishing another welfare state type, organizing South European countries into a new strongly family-based welfare state regime (Leibfried, 1992; Barlow & Duncan, 1994; Ferrara, 1996). This model was called Mediterranean and includes among others the countries Italy, Spain, Portugal, Malta, and Greece. The main difference of this model is in its dominant position of family, and the high number of welfare tasks that are carried out by the family with minimum interference of market and state. The Mediterranean model is also characterized by a strong insider-outsider problem, which means that insiders receive very strong protection, but outsiders receive weak protection. For example, pension schemes provide generous protection to core sectors of the labor market (usually public sector employees, white collar-workers and private wage-earners of medium and large enterprises), but weak protection to those located outside of the core sectors (Ferrera, 1996 in Aiginger & Leoni, 2009). The insider-outsider dynamic extends to other spheres of social protection, like short-term risks deriving from sickness, maternity and temporary unemployment. Besides, young people who are still in education or entering the labor as well as families with small children can count on little support from the state (Aiginger & Leoni, 2009). The labor market in this model is segmented; old and young people’s employment rate is relatively low. Public expenditure is unbalanced with high expenditure on pensions and very low on social assistance (Urbe et al., 2012). Table 1 below summarizes the most important differences between the four welfare state regimes deducted from Esping-Andersen’s typology and expanded with the modern research based on the critiques on the original typology.

(9)

Table 1. Welfare state regimes’ characteristics. Source: Hoekstra, 2013 Liberal welfare state model Social-democratic welfare state model Corporatist (conservative) welfare state model Mediterranean welfare state model De-commodification: extent to which a regime promotes an acceptable standard of living independent of one’s market value

Low High Medium Low

Stratification:

whether the welfare state increases or decreases differences between groups of civilians

Increases

differences Decreases differences Reproduces existing differences

Reproduces existing differences

Income distribution

and poverty Large income differences, high incidence of poverty Small income differences, low incidence of poverty Medium income differences, medium incidence of poverty Large income differences, high incidence of poverty

Unemployment Low Low High High

Arrangement between state, market and family Dominant position of market parties Dominant position of the state Important position of the family, considerable influence of private non-profit organizations Dominant position of the family Representative

countries (EU) UK, Ireland Sweden, Denmark, Finland Belgium, Germany, France, Austria Italy, Spain, Portugal, Malta, Greece

(10)

2.2. Housing system in welfare state classification. Models of Kemeny and Hoekstra

Even though welfare state research has a long tradition, housing has been playing a minor role in its output. While the welfare state regime classification of Esping-Andersen was developed with reference to pension systems and sickness and unemployment benefits, housing was omitted from the study because of its ambiguous role within the wider welfare state. Housing researchers have made many attempts to incorporate housing systems into Esping-Andersen’s classification or develop original typologies of housing regimes.

It is generally acknowledged that a welfare state is supported by four main pillars: social security, healthcare, education, and housing, because they are characterized by considerable state influence and de-commodification (Hoekstra, 2013a). According to some scholars (e.g., Harloe, 1995; Kemeny, 2001), the housing pillar has a specific position and significantly differs from security, healthcare, and education. The reason is that housing is a “tradable commodity”, which occupies a central position in the capitalist economy; it is characterized by high capital intensity and in some cases, a strong private-sector involvement. Unlike social security, healthcare, and education, which are usually provided by the state and funded from taxes, housing provision often relies on the private sector (Fahey & Norris, 2009). For these reasons Torgersen (1987) called housing “a wobbly pillar under the welfare state” (in Hoekstra, 2013).

Housing researchers have made several attempts to either incorporate the housing system in the welfare state classification of Esping-Andersen (Hoekstra, 2003) or create an original typology of housing regimes (Kemeny, 1995). In his study of 7 countries Kemeny made an attempt to explain why different housing systems emerged and developed a two-fold typology of dualist and unitary (rental) housing markets. The initial premise of his model was the idea that the power configuration of ‘market-state-family’ may vary in different sectors, which would allow housing to diverge from the wider welfare regime (Kemeny, 2015). According to him, various types of housing markets arise from whether these power relations are mediated through corporatist institutions that try to balance different interests (Stephens, 2016).

The dualist rental housing system is characterized by two polarized rental tenures (cost and for-profit) and a strong preference of housing policies for the owner-occupied sector. The private rental market is profit-oriented, largely unregulated and is based exclusively on market principals with minimal security. The social sector is constructed as a cost renting system, where housing is provided in the form of a publicly controlled, strongly regulated command economy and functions only as a safety-net for the poor (Lennartz, 2011). This way of organization of two tenures prevents competition between them, which makes the dualist rental regimes unattractive. The state thus seeks to promote home-ownership as a norm. This market type is typical for Anglo-Saxon countries.

In the unitary rental system, landlords compete with for-profit firms on the open market. The integration of non-profit firms into the open market is possible through their financial maturation using subsidies and rent controls as state tools to promote competition. The main assumption is that the integration leads to competition between all types of landlord having a dampening effect on the market rent levels. This integration results in increased robustness of rental markets to external shocks and an improved quality standard of dwellings, which makes the rental sector an attractive alternative to home-ownership (Kemeny, 1995 in Lennartz, 2011). Germany, the Netherlands, and Scandinavian countries are representatives of the unitary system.

One of Kemeny’s core ideas is that housing is embedded in social structures and significantly affects the structuring of society and the welfare state. He claims that a dominant position of certain tenure has important implications for a wider welfare state regime. For example, in systems with prevailing ownership, wealth is redistributed over the lifestyle from the young to

(11)

the old. Younger households are burdened by high mortgage payments, thus are not able to pay high taxes to fund welfare state provision. In societies with prevailing renting the housing costs are equally distributed over the lifecycle of young people, who are able to pay taxes to the welfare provision. Systems built around homeownership are rooted in ideologies of privatism and individualism. The second – with a prevailing renting sector - is grounded in collectivist countries (Lennartz, 2011).

Studies seeking to employ Esping-Andersen’s classification to the housing field faced unavoidable questions as how to apply it to the subject. According to Mark Stephens (2016) the most common approach is to simply assume that a housing system reflects a broader welfare regime, but “only if regime theory reflects causes”. For instance, if power configurations between market, state and family in a country favor a social democratic regime, we might expect the same configuration in a housing system (Stephens, 2016: 23).

If the relationship of the state-market-family triangle cannot be determined and associated with particular pattern of commodification and stratification, it might be possible to make an assessment by other means by adapting Esping-Andersen’s typology to housing (Stephens, 2016: 24). Like it was done by Joris Hoekstra, who made an attempt to explain the differences in housing policies, housing market development and outcomes, by combining Esping-Andersen’s and Kemeny’s models and applying as many variables of housing systems as possible (Hoekstra, 2005; Elsinga & Hoekstra, 2005; Hoekstra, 2009). Hoekstra (2003) adapted the original Esping-Andersen’s definition of de-commodification as “the extent to which households can provide their own housing, independently of the income they acquire on the labor market” (Hoekstra, 2003: 60). In his method, de-commodification is represented by housing subsidy, price regulation, and stratification - by the rules applied to social housing allocation. In other words, he assumes that a social democratic housing regime is characterized by large production subsidies, a highly regulated market and housing allocation on the basis of need. Conversely, liberal housing regimes would mostly allocate housing to the market, with a small share of social housing to be reserved for the poor; production subsidies would be small or non-existent and regulation would be minimal. Hoekstra’s housing and welfare state regime classification is presented in Table 2.

Esping-Andersen’s classification of welfare states is almost 30 years old, and even since Kemeny’s (1990) and Hoekstra’s (2003) classifications circumstances have evolved. In recent years most welfare states have undergone a profound restructuring, influenced by demographic change, financial problems, and globalization.

Nowadays the process of welfare state rollback has spread across the European countries that historically had strong universality and equality principles in social and economic policies (Dewilde & Ronald; 2017). The model of the ‘male breadwinner’ described in ‘The three words of welfare capitalism has broken down, and labor markets have become more ‘feminized’ (Stephens, 2016). Full employment in the traditional sense has ended given way to growth in part-time work and zero-hours contracts. Demographics, including rise in the number of lone parents, single person households and an ageing population represent further significant changes. The rise in number of immigrants in previously ethnically homogeneous countries, as well as growing student mobility and internationalization of education have presented a new potential line of stratification.

(12)

Table 2. Housing and Welfare state regime classification. Source: Hoekstra, 2010 Welfare state regime (Esping- Andersen) Liberal Social-democratic Conservative Mediterranean Ideology, political structure and type of rental system (Kemeny) Privatist ideology, non-corporatist political structure and dualist rental system Collectivist ideology, corporatist political structure and unitary rental system Not included in Kemeny’s theory and typology Share of

homeownership High Mixed Mixed High

Quality and appreciation of apartments

compared to single-family dwellings

Low Low Low High

Housing outcomes with regards to the rental market Typical for dualist rental system Typical for unitary rental system Typical for unitary rental system Not included in the analysis Housing satisfaction of homeowners compared to tenants

High High Mixed High

Housing systems have changed too, followed by increasing commodification, privatization and financialisation. In Europe social renting no longer attracts financial subsidies to support new constructions. Home-ownership has risen even in countries where renting was meant to provide an attractive alternative to ownership. With growing home-ownership mortgage debt goes in the same direction. ‘Financialisation’ has become an unavoidable issue (Aalbers, 2016; García-Lamarca & Kaika, 2016), and access to housing has become obstructed for young people. Generational divides bring another line of stratification to the system (Stephens, 2016). Hyperactivity of capital in the post-crises era made real estate markets an attractive area to store wealth not only for owners but also for international investors. In the UK, for instance, the government started to scale down social housing in the 1980s through ‘right-to buy’ scheme and reduced new social housing provisions, instead catering to low-income households through a wider range of public, non-profit, and private intermediaries. Private landlords played an increasing role in accommodating low-income households on housing allowances. Similar transfers of state-market responsibilities were later adopted in other Western European countries such as Sweden and the Netherlands (Dewilde & Ronald; 2017).

In her recent article “Do housing regimes matter? Assessing the concept of housing regimes through configurations of housing outcomes” Caroline Dewilde (2017) found out that in today’s world “the configurations of housing outcomes have become less associated with the features of housing regimes but rather with welfare state type and the country’s economic affluence” (Dewilde, 2017: 1). I therefore focus on the effect of the welfare state regime rather than housing regime on housing outcomes.

(13)

2.3. Student housing market

The student housing market is a substantial and growing niche in the public and private rental sector. A niche market is a market in which supply has become adapted to meet the needs of a specialized group and which displays reluctance to meet demand from another source (Rugg, Rhodes, & Jones, 2002). As a niche real estate market student housing market has many particular aspects within the overall housing market. It is characterized by its dominating rental nature, type of developers and operators, type of contracts and rental time lease, types of dwellings, models of investment and building, and market resilience.

Describing the features of student housing markets, I will primarily focus on the power of interest groups such as non-profit landlords, for-profit landlords (private developers) and educational institutions rather than on the role of political parties in housing systems. In the student housing market, the non-profit landlord group is represented by national or municipal social housing corporations. The for-profit landlord group is represented by private developers including purpose-build student accommodation (PBSA) developers and small private landlords). Purpose built student accommodation (PBSA) is accommodation that is specifically built for occupation by students with consideration to their specific needs.

The student housing market is a rental market where a dominant number of accommodations is rented out to students for a fixed period of time. The duration of contracts differ depending on the landlord. For-profit landlords, especially small landlords, tend to offer more flexible contacts offering one-month to permanent lease. Private PBSA developers tend to set up the contracts for a period of six-month or one-year. Non-profit landlords usually offer a one-year rental lease agreements, with or without possibility to extend. The universities have rental lease times similar to non-profit landlords.

Social and private student housing markets in Western Europe significantly differ in the type of accommodations they provide but the quality of both is quite high. Private PBSA developers have a tendency to offer more luxury housing, supplying a great number of furnished studio flats or en-suite accommodations of hotel-type rooms with bathrooms and flat-screen TVs (Forrest, 2016). The public sector also builds a vast number of studios but of a cheaper price and slightly lower quality, as well as private rooms with shared kitchen and shared/or private bathroom.

The investment and building models created for student housing providers differ per country. In the liberal welfare state types student-housing market is purely market-driven and governments do not provide any financial or non-financial incentives. In social democratic or conservative countries, the state may support construction of student housing with interest-free loans (e.g., Denmark and Finland), construction subsidies (e.g. Sweden) or public subsidies toward student rents (e.g., housing allowances in the Netherlands).

The student housing market is closely related to the welfare state regime, because students are a vulnerable low-income group and mostly rely on welfare services, especially in housing provision and financial support. Becoming a student is a reason why many young people move to a new city and leave the parental home. The majority of full-time students does not participate in the labor market and makes no income, relying instead on parental support or governmental loans/subsidies.

The provision and development of student housing has become a challenge for many university cities as a result of international mobility, the ongoing expansion of higher education institutions and growing student population. However, over the last decade student housing

(14)

market has turned from the pioneer niche sector into a reliable mainstream real estate market in Europe and worldwide.

Demand for student housing is predicted to remain strong through 2020 because of the forecasted increase in enrollment rates and population growth. Positive demand trends in the student real estate market generated a growing supply interest from private investors and developers. For example, during 2015 the investments into global student accommodation market has seen rapid expansion especially in the UK, US, Netherlands and Germany (Savills World Research, 2015). Investors consider student housing as a profit-making activity for several reasons. First of all, student housing is characterized by high returns because of the high turnover, very low vacancy and collection loss. Near campuses an occupancy rate of 100% is not uncommon, and students can combine their purchasing power through shared accommodations. Therefore, the investments in student housing are often recession-proof (Stribling, 2007).

Second, the surge in interest for private investments meets the demographics. People born between 1980 and 2000, the so-called echo boomers or millennials are now of university age. They already outnumber the baby boomers and account for 2.3 billion in the world (Goldman Sachs Global Investment Research, 2016). The high percentage of millennials as well as the new economic reality (economic crisis, globalization, unemployment) has led to growing student enrollment rates in the world. The world’s gross enrollment ratio accounted for 34.5% in 2014 (latest available data), compared to 29.5% in 2010 and 19% in 2000 (World Bank, 2016). Although in Europe the enrollment rates show significant national differences, the majority of European countries indicate an increase in enrollment rates during the last decade. According to Eurostat, the most significant growth in the percentage of tertiary education students in the total population of 20-24 years old were noticed in Belgium, Spain and Netherlands (Eurostat, 2017). Third, economists have noticed the non-cyclical economic nature of tertiary education, meaning that in good economic times a university degree provides an important credential in the job market, and in difficult economics times people attend university in order to temporarily avoid a challenging job market (Levy & Tucker, 2006).

To conclude, increasing influx of tenants and potential for steady revenues even in times of recession attract private developers and investors to enter student market niche in real estate. Welfare states reduce funding for housing provision, which makes it impossible to meet the demand without private sector participation. Pursuing their main goal – to make profit – private student housing accommodations rarely have affordable price.

(15)

2.4. Housing outcomes. Affordability and accessibility of student housing market

Housing regimes are essentially unobservable, but they can be evaluated by the outcomes they produce. Housing outcomes refer to adequacy of housing in terms of its physical size, quality of units, quality of neighborhood, costs, physical and market access to housing (Stephens, 2016: 26). The research of Dewilde (2017), in which she investigated the correlation between housing regimes and housing outcomes in 15 Western European countries, concluded that differences in the level and form of state intervention and the social production of housing shape typical patterns of housing outcomes. The quantitative analysis findings revealed that more state intervention (particularly in the form of a regulated unitary rental market) resulted in good housing conditions and low housing costs (Dewilde, 2017: 14).

Like other housing, student accommodation is evaluated by the number of indexes of its appropriateness: size, cost (affordability), quality of dwelling, physical accessibility (location), market accessibility, furnishing, number of roommates, quality of shared space and so on. In this research the study will draw on affordability and market accessibility. These two outcomes were chosen because they reflect the configuration of power between market and state in the most realistic manner. Dewilde (2017) concludes that more state intervention results in better housing outcomes, particularly low housing costs. However, extensive state interventions might have an opposite effect on another housing outcome – accessibility, because they hinder private actors involvement in housing provision. This potential contradiction is the main reason I focus on the configuration of affordability and accessibility outcomes. A detailed operationalization of the concepts is given in Chapter 3.3. 2.4.1. Affordability

A vast literature has been devoted to the concept and measurement of housing affordability (Thalmann, 1999; Lerman & Reeder, 1987; Stone, 1990, 1993; Bogdon & Can, 1997). The term affordability has been one of the topical policy indicators to the central concern of housing policy and thus the key theme of contemporary housing studies.

Many different approached have been developed to measure housing affordability. In a general sense affordability is a challenge each household faces in balancing the cost of its actual or potential housing and its non-housing expenditures within the constraints of its income (Stone, 2006). The most commonly used concept of affordability is called the cost-income ratio (or rent-income ratio) and estimates the share of household’s income that is devoted to housing costs. A high cost-income ratio indicates low affordability of housing and low cost-income ratio – high affordability (Kadi, 2014).

Because this approach is relative and context-sensitive, different ratio standards of assessing housing affordability problems have been set. Affordability ratio standard is the highest level of cost-income ratio above which housing is considered unaffordable. For example, in the United Stated the affordability ratio standard is set at the level of 30% of income. In Europe housing is considered unaffordable (overburdened) when the total housing costs represent more than 40 % of household’s disposable income (Eurostat, 2017). The cost-income ratio approach to affordability has faced many critiques for the relativity and subjectivity of the ratio standard (threshold). The problem arises when it is not clear how to define this standard, and whether the same standards can be used in different contexts or for different types of household. Stone (2006) argues that affordability is not a characteristic of housing but the relationship between housing and people. Therefore, in addressing the affordability issues it is important to think of the questions: affordable for whom? On what standard of affordability? And for how long?

(16)

These questions were partially considered in the fundamental affordability model by Leerman and Reeder (1987). In their article the authors developed quality-based definition of housing affordability adding the adequacy standards to the cost-income criterion. Their definition of affordability applies to households renting minimally adequate but decent, safe, and sanitary housing for less than a specified percentage of income (30%) (Leerman & Reeder, 1987).

Another established concept of housing affordability was developed by Thalmann (2003) in response to the conventional rent-to-income indicator. In his model Thalmann proposes a set of indicators based on the residual income approach introduced by Grigsby and Rosenburg (1975). Those indicators allow to identify when a household’s housing rent differs from the ‘standard’ and when households pay more for their housing rent than market prices (Thalmann, 2003). For the housing policy these indicators help to determine who needs general income support and who needs specifically housing aid. In this thesis, however, the standard cost-income affordability ratio is used.

2.4.2. Accessibility

Affordability as a measure for housing outcomes is not sufficient on its own because it does not capture the dimension of housing access. It might be a possible situation that housing is highly affordable at the market but in such shortage that a household does not have an access to it. Access to decent housing for vulnerable and low-income groups (such as students) depends on an adequate supply of affordable housing (Edgar et al, 2002: 59). The number of available affordable dwellings may be limited, because of the low mobility within the affordable accommodations or the lack of new constructions of such dwellings. In order to face this limitation of an affordability measure, the parameter of housing accessibility is included in the research.

The majority of literature about housing accessibility uses this term to describe physical accessibility – access to housing that enables independent living for people with disabilities. Another widely used understanding of accessible housing is related to its physical location – spatial accessibility. In this study accessibility is understood in accordance to Neuteboom & Brounen’s (2001) definition as the joint result of the prevailing market conditions and specific household characteristics. Accessibility is a notion that covers two sides of the market: demand and supply. The demand side relates to affordability and the supply side relates to competitiveness (Neuteboom and Brounen, 2011: 2232). Accessibility is therefore defined as the availability of affordable dwellings considering the students’ competitiveness for this type of dwellings.

In his article, Richard Sendi argues that “the issue of housing accessibility is increasingly becoming acute as a consequence of the gradual abandonment of welfare state policies that occurred in many European countries during the last decade” (Sendi, 2011: 1). He suggests that the widely used concept of housing affordability represents “a failure by the housing research community to recognize fully the essence of the problem” and therefore the emphasis in housing research should shift from housing affordability to accessibility (Sendi, 2011: 4). The generally limited availability of affordable and accessible accommodation for students presents the actual bottleneck. This results in students either staying longer with their families, taking on more debt, living in substandard or illegal housing and not attending university. The study on affordable and accessible student housing goes beyond housing research. It contributes to broader debate about higher education policy development, university strategies on attracting domestic and international students, and young adults’ pathways to the labor market and home ownership.

(17)

2.5. Hypotheses

Based on this literature review on welfare state regimes and the outcomes they produce I expect that systemic differences in the levels and forms of state intervention in the student housing market shape different patterns of housing outcomes. Executing empirical research on different types of welfare regime, I argue that they provide different levels of affordability and accessibility on student housing market. The following hypotheses have been developed:

H1: Social democratic welfare regime is characterized by the dominant position of the state in state-market-family relations. A dominant position of the state results in more extensive policies and regulations (e.g. housing allowances, rent regulations, social housing) and more generous social benefits such as free education and student subsidies, which provide higher level of affordability of student housing market.

H2: A dominant position of the state resulting in extensive housing rules and regulations hinder private actors from entering the housing market because it is difficult for them to operate within tight regulations. As social housing provision is limited, the welfare state cannot ensure a sufficient level of student housing supply on the market, which causes housing shortages. Therefore, social democratic welfare states are characterized by a lower level of student housing market accessibility.

H3: Liberal welfare states are characterized by the dominant position of the market in state-market-family relations. When market actors have a stronger position than the state in student housing provision and are not bound by housing rules and regulations, they set high rent prices. At the same time, a welfare state with a dominant market does not provide sufficient social benefits to supports students’ incomes. Therefore, in countries with liberal regimes the level of housing affordability is relatively lower.

H4: In countries with a liberal welfare regime, student housing provision is largely market-driven and the state’s rules and regulations are minor or not existent. This creates an attractive market conditions for international investors and developers, thus an increasing supply and more housing options for students. Therefore, in liberal welfare states the student housing market is characterized by a higher level of housing accessibility.

H5: The transitional regime combines the attributes of liberal and social democratic welfare regimes, thus provides moderate levels of affordability and accessibility.

(18)

3. Methodology

This chapter describes the methodology of the research. The first part explains the research design. The second part justifies the case study selection. The third part presents the conceptual framework and operationalization of the dependent and independent variables.

3.1. Research design

In order to empirically test my theoretical expectations, I use a multiple-case study research design. This type of design is common within housing research, particularly those investigating the role of a welfare regime in urban housing markets (e.g., Kadi, 2014; Kadi & Musterd, 2014; Ronald & Elsinga, 2012; Lennartz, 2016; Ruonavaara, 2008). Welfare regime theory provides a ready comparative framework that justifies a selection of cases, in which each new country represents a type of welfare regime (Druta, 2017).

I analyze typical case studies within opposing welfare regimes to determine the differences in their levels of affordability and accessibility of student housing. The number of case studies (3) does not coincide with the number of welfare regimes (4) because the research is restricted to the two opposing regimes (liberal and social democratic) and the one in the middle – transitional regime. In the countries with liberal welfare regime student housing market is the least regulated and the share of market actors in student housing provision is the highest. While in the countries with social democratic regime it is the opposite: the student housing market is highly regulated and the share of student housing public provision is the largest. As a third case study, I analyze a transitional model – a welfare regime, which is placed between the liberal and social democratic. The opposing cases allow illustrating the social phenomena in a very pronounced way; and the transitional model has common characteristics with two extreme cases, demonstrates enough commonality to allow for its comparison with these cases. Within the opposing regimes I select a typical case studies – countries that best represent these welfare regimes.

3.2. Case study selection

The research draws on data collected in three European urban contexts: Sweden (social democratic regime), the United Kingdom (liberal regime) and the Netherlands (transitional regime). The choice of these contexts was driven by both theoretical and practical considerations.

The United Kingdom is one of the two representatives of liberal welfare regime on the European continent (together with Ireland). It is an extreme case that represents a student housing market with no social housing options for students with the dominant majority of student accommodation provided by private actors. Student housing provision is purely market-driven and the developers are not obliged by tight rules and regulations. The UK student housing market is especially interesting because despite the uncertainty from the UK’s vote to leave the EU, international investments in student housing are still going strong (The Class of 2020, 2017). Besides, the new Higher Education Bill that allows universities to raise tuition fees, the rent strikes and student protests in London sparkes my interest to analyze housing affordability and accessibility situations in this liberal country.

Out of three European countries with social democratic regime (Sweden, Denmark, and Finland), the Swedish student housing market is the most regulated in Europe, which creates unattractive conditions for international investors to enter the market and leaves the largest share of student housing provision up to municipal organizations. Nowadays the student housing market in this country is undergoing gradual changes as the problem of accessibility is getting rather extreme. The majority of university cities in Sweden have long waiting lists of 50

(19)

weeks on average to get a student accommodation. The new projects Stockholm 6000+ and Gothenburg 7000+ are aimed at collaboration of national and municipal governments, private and public developers to work together on improving the situation.

Even though the Netherlands’ welfare regime was defined as conservative corporatists in Esping-Andersen’s comparative welfare research, later researchers referred to it as a transitional/hybrid regime (Hoekstra, 2013). Looking at the Dutch student housing market, it is considered transitional between social democratic and liberal regimes. The Dutch student housing market has a generous social housing sector that provides the largest share of purpose-built student accommodations. However, nowadays the Dutch student housing market is in a transitional phase, moving towards a liberal model (Savills, 2016: 2). This is manifested by the government’s actions on cutting public spending on student support (in September 2015 government introduced a new loan system that replaces the grant system) (Savills, 2016: 5). Even though it is still in its infancy, the student housing private market is maturing and investment volumes increase. In order to attract and keep the investors, the government eases rules and regulations. Municipalities enable zoning changes to stimulate the transformation of vacant office building into temporary projects and the national government reduced the minimum surface size for new PBSAs from 18 to 15 m2_{(Savills, 2016: 6) and made some}

changes in the point system.

3.3. Conceptual framework and operationalization

The conceptual framework below illustrates the relationship between the independent and the dependent variables in my model. The independent variable is the ‘welfare regime’. It has three attributes: liberal, social democratic and transitional. The dependent variable in the model is ‘student housing outcomes’. This variable has two attributes: affordability and accessibility. The independent variable ‘welfare regime’ is defined from the literature review and measured with qualitative data collected during the expert interviews. I operationalized it by focusing on two aspects deducted from the thematic analysis: • Welfare states’ interventions stimulating accessibility of student housing (the supply side of student housing market): 1. Public subsidies and/or direct government investment; 2. Government loans; 3. Specific legislation governing the investment and operation of student housing (e.g. tax incentives). Attribute 2: Accessibility

‘Queue time’ or ‘waiting lists’ for affordable (social) housing

Attribute 1: Affordability

Proportion of ‘average student living expenses’ to ‘average rent’ INDEPENDENT VARIALBE Welfare regime Liberal Social democratic Transitional DEPENDENT VARIABLE Student housing outcomes Figure 1. Conceptual framework

(20)

• Welfare states’ interventions stimulating affordability of student housing (the demand side of student housing market):

a. Public subsidies towards students’ rent (housing allowances);

b. Public subsidies towards students’ income (government education or/and maintenance grants)

c. Rent regulations; d. Tuition fees.

The comparative case study research approach allows me to identify the cases where these indicators are present/strong and where they are absent/weak and to relate these to the strength/presence of the dependent variable.

The dependent variable ‘student housing outcomes’ is measured with quantitative data collected from student housing market reports and interviews. I use the income-to-rent ratio concept of affordability, which estimates the share of household’s income that is devoted to housing costs (Kadi, 2014). As students do not usually have stable income, the average amount of student living expenses is used instead. Data about the average student accommodation rent in the country is available online and provided in the following chapter. In sum, affordability is measured as a proportion of ‘average student living expenses’ to ‘average rent’. Affordability is a market concept related to the capacity to pay. Accessibility on the other hand is a humanitarian concept, which represents the ability of households to get access to housing. People who can pay for housing can get an access to it, and those who cannot afford to pay, cannot get an access to housing. These are the relationships between the two concepts, thus the affordability variable is included in the accessibility concept. In this research, therefore, accessibility is defined as availability of affordable dwellings considering the students’ competitiveness for this type of dwellings. To evaluate accessibility of student housing in Sweden and the Netherlands, I used quantitative data on the ‘queue time’ or ‘waiting list’ for affordable (social) housing. In the UK I used the ‘bed ratio’ parameter to evaluate accessibility for the private PBSAs and qualitative data from expert’s interview to estimate the queue time for university housing.

(21)

4. Data and Analysis

This chapter presents data and analysis on the research. In the first part I provide the data description explaining why certain experts were selected for the interview. In the seconds and third part I explain my data collection and analysis methods. Finally, the analysis procedure is presented, which is based on the 6-stage suggested by Braun and Clarke (2006).

4.1. Data description

The study uses a purposive sampling method with an experts sampling approach (see Table 3). Purposive sampling method, also known as judgmental, selective, or subjective, “relies on the judgment of the researcher when it comes to selecting the units” (Black, 2010). This type of method is the most appropriate because in every country student housing sector is a small niche with a limited number of experts, who are easy to identify.

The experts were identified based on theoretical and practical considerations. While doing this research I collaborated with The Class of 2020, a think-tank platform of student housing in Europe, who provided me with contacts of their partners in each of the three countries. The results of this thesis will be used for a The Class of 2020 blog articles and the Annual Trends Report on emerging trends and development in student housing in Europe. In order to increase the validity of the collected data, I included different types of experts in my sampling: experts from social housing organizations, private commercial actors and experts from umbrella associations, who do research on student housing market, and have very broad knowledge of it within the country or internationally. I chose this combination of experts because it allows me to identify the power configurations between the state and the market, and to get in-depth information about both public and private sector success or struggle in providing affordable and accessible student housing. In the Netherlands, I interviewed the strategic advisor for public affairs at DUWO Martjin van der Linden, the CEO and founder of The Student Hotel Charlie MacGregor, and the foundation manager of The Class of 2020 Wouter Onclin. DUWO is a social housing association and the biggest provider of affordable student housing in the Dutch market. Today, DUWO lets approximately 30,000 homes in the regions of Amsterdam, Delft, The Hague, Leiden as well as in Deventer, which was added at the start of 2015 through a takeover of the student housing there (DUWO, 2017).The Student Hotel is a new and already internationally successful project by British investors. It is a highly innovative yet controversial development, because it follows the hotel business model and does not comply with student housing rules and regulations, setting higher prices for its properties. However, expensive prices are justified by the high level of product: trendy design, large modern communal areas, hospitality, and flexible contracts. In the Netherlands, The Student Hotel has its properties in Amsterdam, Rotterdam, The Hague, and Eindhoven, but the model is currently expanding overseas to Spain, France, Portugal, Italy, Germany, and the UK. The Class of 2020 is a leading think tank organization on student housing and internationalization of education in Europe. It is an umbrella organization for a vast international network of partners, which help The Class produce high-quality research and share best practices on the student housing market. Additionally, they organize conferences and bring experts together to work on up-to-date issues.

In Sweden, I talked to Susanne Malmgren, the Director of Student Accommodation from national organization Academiska Hus; Mira Grunevald, the director of Business development at private company ByggVesta and Martin Johansson, the Secretary General at Studentbostadsföretagen, which is translated in English as the Swedish Student Housing Federation.

(22)

Academiska Hus is the only state-owned property company in Sweden with 65% of the market share for university properties. They collaborate with universities and colleges to develop, build, and manage campuses, university buildings, and accommodations. Academiska Hus has a leading role in the realization of the new student housing strategy Stockholm 6000+, which is to provide 6500 new student units in the Stockholm area. BuggVesta is a knowledge-intensive private developer and manager, who focuses on sustainability. ByggVesta recently started to operate on the student real estate market but already has succeeded in Linköping and Stockholm. Studentbostadsföretagen is an umbrella organization for Swedish student housing providers. Its members hold approximately 75% of all student accommodation on Swedish housing market. Collecting, creating, and conveying information from and to their member companies, Studentbostadsföretagen has the best knowledge of the overall market.

Among the UK actors I interviewed are the Head of Accommodation Services at a public university, the University of Sheffield Ian Jones and the member of CUBO (College and University Business Officers) umbrella association Richard Kington. No social housing corporations exist on the student housing market in the UK. The only public actors there are public universities. The universities are publicly funded and regulated (the government regulates their tuition fees, student funding and student loans, research assessments, etc.), although the government does not own its assets. CUBO is a professional association for senior managers in higher education with responsibility for strategic development, management, and administration of institutions’ commercial businesses, including residential portfolio (CUBO, 2017). CUBO includes members who represent a significant part of Higher Education business and therefore is a knowledgeable participant in the research.

Unfortunately, I did not succeed in interviewing a private developer in the UK student housing market. I have been in contact with the largest private developer Crosslane, but their PR-department did not agree to provide information. Instead I interviewed the Residential Service manager at the University of Lincoln Michael Ball, who is a former Chairman of ASRA umbrella organization and has a broad understanding of the student housing market as a whole. ASRA is the leading representative organization for student accommodation professionals throughout the UK and Ireland. They are a network of people working within student accommodation in the education, public and private sectors (ASRA, 2017).

(23)

Table 3. Experts sampling

Type of

organization Organization and interviewee in

the Netherlands

Organization and

interviewee in Sweden Organization and interviewee in the UK

Public organization DUWO Martijn van der Linden, Strategisch Adviseur Public Affairs Date: May 3d Place: Office DUWO, Campus Uilenstede Academiska Hus Susanne Malmgren, Director of Student Accommodation Date: April 28th Interview conducted by phone University of Sheffield Ian Jones, Head of Accommodation Services Date: May 15th Place: Conference “Practical approaches to student residence life” at TSH Collab Wibaustraat 131, Amsterdam Private developer The Student Hotel Charlie MacGregor, CEO & Founder Date: May 11th Place: The Student Hotel City, 129 Wibautstraat, Amsterdam ByggVesta Mira Grunewald, Director of Business Development Date: 5th_of May Interview conducted by phone Student housing association The Class of 2020 Wouter Onclin, Foundation Manager Date: May 5th Place: The Class of 2020 Office Weesperstraat 61, Amsterdam Studentbostadsföretagen Martin Johansson, Secretary General Date: April 25th Interview conducted over Skype CUBO Richard Kington, Director, Accommodation Catering and Events at University of Edinburgh Date: May 4th Interview conducted over Skype ASRA Michael Ball, Residential Services Manager at University of Lincoln Date: May 18th Interview conducted by phone