Housing tenure preferences of young one-person households in the Netherlands : a generational perspective

(1)

Housing tenure preferences of

young one-‐person households in

the Netherlands

A generational perspective

Selma Romkes 10192123

MA Human geography University of Amsterdam Master thesis 15 August 2016 Supervisors Mw. prof. dr. D. Manting Dhr. dr. W.R. Boterman Second reader Dhr. prof. dr. S. Musterd

(2)

Preface

Amsterdam, 18 June 2016

Being a member of generation Y, I am constantly accused of having all kinds of personal traits. It seems to me that the opinions of members of generations on members of other generations hardly ever go hand in hand with positive connotations. According to opinions on the Internet, being a member of generation Y makes me spoiled, indecisive and unable to finish whatever I started.

In February 2015 I registered for the master Human geography at the University of Amsterdam. In my motivation letter I wrote the following words:

“During my bachelor Interdisciplinary Social Sciences I endured a broad interest in different fields of study. My primary focus has been both on the housing market and urban history. I am very much interested in challenges concerning urban space, especially within the housing field. During my masters I hope to further develop my knowledge on housing preferences and housing markets.”

Writing this preface and looking back on my thoughts and aims from a year ago, gives me the insight that I unconsciously combined all the fields of study I hoped to explore during my masters, into this thesis. I thank everyone that helped me accomplishing this.

I finished what I started.

(3)

Abstract

This quantitative longitudinal research examines the stated housing tenure preferences of young one-person households in the Netherlands between 1985 and 2015 from a generational perspective. Academic and popular writings from different countries point towards changing household formation and housing preferences among millennials. Using data from the Dutch housing research it is attempted to contextualize young one-person households within time in comparison with preliminary generations and reveal differences within their societal context. Using longitudinal and binomial logistic regression analyses, this research shows that young person households cannot be treated as one homogenous group. Rather, young one-person households from different generations show to be very different in terms of their housing tenure preferences. Over the years, the preference for homeownership has decreased from 1985 until 2015. Most young one-person households from generation Y prefer renting a home.

Key words: housing tenure preferences, one-‐person households, generations, generation Y,

individual context, spatial context, societal context

(4)

Introduction

Motive

Housing studies increasingly stress the way young people form households and how these households move on the housing market. The behaviour of members of generation Y - born between 1980 and 1995 - is considered to be very different from the way in which former generations moved on the housing market. It is widely adopted that former generations followed a certain life cycle in which most households went through comparable life events accompanied by corresponding forms of tenure with owner-occupation as a desirable end goal (Wulff, 2001; McKee, 2012).

However, in many countries homeownership rates are falling and it is increasingly questioned to what extent millennials prefer rental dwellings over owner-occupied housing (McKee, 2012). In the UK, academics and the media even speak of ‘generation rent’ (Lund, 2013) referring to young people that are unable to access the housing market (Boterman et al., 2013). However, the context of the Dutch housing market and society is different and the Netherlands does not have such a homeowners-society history as the UK. Renting in the Netherlands is nothing new. However, the promotion of homeownership is an important aim in Dutch policy (Esveldt & De Jong, 2013). Nonetheless, the housing tenure preferences of young people in the Netherlands are a not much investigated topic and it is unclear how the housing preferences of young people relate to the way in which they form households and the generation they belong to.

At the same time, in one of the demographic scenarios for the future described in ‘Cahier Demografie: Toekomstverkenning, welvaart en leefomgeving’ De Jong (2015) argues that the life course of households in the Netherlands changes as a result of increasing welfare. He describes that young people want to live independently and that solo living is a common way of living for young adults. In 2015, 27% of all one-person households in the Netherlands were members of generation Y (CBS, 2015b). This means that not only the way of tenure may

be changing, but also that young people are increasingly living in different types of households. Besides, not only young people are considered to prefer renting a home over buying one. It is argued that one-person households have the same preference (Faessen, 2002).

In this study these two phenomena are combined into one research on the housing tenure preferences of young one-person households belonging to generation Y. In conducting

(7)

the research, specific attention is given to individual, spatial and societal contexts that are believed to shape housing tenure preferences. Especially the last one, the societal context, is of particular interest because it are the larger societal structures and events that shape the identity of generations (Mannheim, 1970; Wyn & Woodman, 2006). Including these contexts helps building towards a generational perspective in which the housing tenure preferences of young one-person households are analysed through the lens of their generation.

Question and research data

As mentioned above, the purpose of this research is to examine the housing tenure preferences of young one-person households by placing them into a generational life course perspective. Therefore, the following research question is formulated: What are the housing tenure preferences of young one-person households of generation Y seen from a generational perspective, giving attention to individual, spatial and societal circumstances?

The data used to answer this question are derived from the Dutch housing survey data from 1985 until 2015. Two different analyses will be used to analyse these data. On the one hand, a bivariate longitudinal analysis shows how the housing tenure preferences of young one-person households developed over the last 30 years. On the other hand, a binomial logistic regression model shows the effect of the individual, spatial and societal circumstances on the chance that a young one-person household prefers buying a home, given the other circumstances that may influence this preference.

Relevance

“For most individuals, housing is the largest consumption and investment item of their lifetime and, as a result, housing satisfaction is an important component of their quality of life.” (Eteca-Amestoy & Vera-Toscano, 2007, p. 1).

The citation above shows the importance of housing in people’s lives. Gaining insight in the housing tenure preferences in relationship to the society in which people live, may be of great importance in policymaking - especially since the Dutch government promotes homeownership - but may also be important in predicting the housing preferences of future generations. According to De Jong et al. (2008) sudden increases or decreases in housing demand have a huge impact on the tension on the housing market since the existing housing supply is very inflexible. Knowing people’s housing tenure preferences within context makes

(8)

it easier to understand these sudden increases or decreases in the demand for certain types of housing and may contribute to a healthier housing market.

Reading guide

In order to give a comprehensive overview of this research, this thesis is divided into seven chapters. This reading guide provides a short summary of the content of every chapter.

The following chapter (1) will discuss the most important theories in order to understand this research within the academic discussions in the housing field and within demographic studies. The theories from the housing studies discussed in this chapter, mainly focus on housing tenure preferences and choice. In the part on demographic studies, the focus will be on studies that focus on generation Y and one-person households.

Subsequently, the second chapter (2) will contribute to the attempted generational perspective of this research by outlining a historical overview of the Dutch societal context. Within this overview special attention will be given to cultural, economic and political trends and events that shaped the lives of generations in the Netherlands in the last decennia.

The third chapter contains the research question, which leads to the formulation of a few sub-questions that will form the foundation for this thesis. Moreover, the formulated questions will lead to several hypotheses - corresponding to the sub-questions - that are tested later on in the analysis.

In the fourth chapter (4), the operationalization of concepts will be discussed after which these will be combined into a conceptual model. The fifth chapter (5) gives more insight in the choice of methods within this research. The research strategy, research design, units of analysis, data and data analysis will all be discussed in that section.

In the sixth chapter (6), first a univariate analysis provides insight in the research population after which the results of the longitudinal and binomial logistic regression analyses will be outlined in order to contribute to a more comprehensive view of the housing tenure preferences of young one-person households in the Netherlands.

Finally, in the seventh (7) chapter, it will be attempted to answer the formulated research question based on the analyses that were conducted. Additional to the conclusions, the discussion contains an evaluation of the research. In this part the limitations of the research are discussed after which suggestions for future research are given, based on the outcomes and limitations of this research.

(9)

1. Theory

1.1 Housing studies

As mentioned earlier in the introduction, this research focuses on the housing tenure preferences of young one-person households on the Dutch housing market. This chapter outlines the different concepts and theories that are needed to conduct the research by examining existing academic research in the housing field. First, it is important to go back to the roots of housing preferences and choice and look at the housing market itself.

In his chapter on the features of the housing market Hugo Priemus (1983) argues that roughly said, the housing market can be divided into two distinct, but interrelated components. The first component is the property market, in which supply, demand and transfer of ownership rights are situated. The second component encompasses the residential service market in which user rights are requested, offered and transferred (Priemus, 1983; Ekkers & Helderman, 2002).

The first part of this literature chapter focuses on housing studies and is divided in three parts. The first part focuses on housing supply; the second part on housing demand with a particular focus on housing preferences and choice; and the third part highlights homeownership as a choice of tenure.

1.1.1 Housing supply

Within academic research two distinct approaches are used to examine the housing supply (DiPascale, 1999). The first approach is the reduced-form estimation (DiPascale, 1999) in which the price of a dwelling is seen as a function of supply and demand factors. The second approach, the structural approach (DiPascale, 1999), has its roots in both the investment literature and in urban geography. Within the structural approach the overall supply is a derivative of new housing construction, which is seen as a function of shifts in price and investment. Three types of decisions are formulated that influence the shape of the supply: (a) production decisions are determined by developers of new constructed dwellings; (b) homeowners make adaptation decisions and (c) policy decisions are made by the (local) government (DiPascale, 1999). In other words, the decisions made by developers, homeowners and policy makers influence the type of dwellings that are available on the housing market. This makes it particularly interesting to look at people’s behaviour on the housing market and whether or not people are or want to become homeowners.

(10)

1.1.2 Demand: Housing preferences and choice

Above, it is abstractly described that the housing market consists of both a supply and a demand side. So far, it is clear that there are two approaches to understand the housing supply, which is, seen from a structural point of view, influenced by decisions of all kinds of actors that are active on the housing market. However, the housing market is mainly about people and the choices they make. To give a complete overview, it is important to look at the approaches for housing demand as well. Therefore, it is relevant to examine the way in which people actually move on the housing market; what are their preferences and choices during their life courses?

In the academic world, decision-making of people in the field of housing is a much-described topic. Choices between buying and renting a home are not only based on housing preferences (Andersen, 2010). In her article ‘Young People, Homeownership and Future Welfare’ Kim McKee (2012) argues that larger structures in the society concerned, shape the housing opportunities for households. Consequently, housing choices are to a large extent constrained by these structures because people have limited opportunities (McKee, 2012). In line with this argument, here it is argued that it is important to not only look at demand and supply of the housing market alone when conducting research on housing choices but that it is of particular interest to look at the underlying social, economic and demographic trends and changes that are strongly interconnected with these constraints and chances on the housing market.

In her study on urban residential preferences of middle-class families Lia Karsten (2007) argues that there are three approaches to explain the demand on the housing market: an economic, demographic and time-geographical explanation. The traditional way of looking at residential preferences and choice is through an economic and demographic approach and is based on the idea that the preferences of households for certain dwellings are dependent on income, knowledge of and experience on the housing market, personal taste and priorities of the household (Wong, 2002). A lot is written on how households vertically move on a so-called housing ladder in which household formation and certain events in life coincide with moving to and living in particular types of dwellings (McKee, 2012; Wulff, 2001). In other words, this so-called “life cycle model suggests that people pass through a predetermined order of stages including pre-marriage, marriage, childbearing, empty nester stage and widowhood, and that these stages go hand in hand with fairly specific housing patterns (from rental to owned; from flat to house; from small to large dwelling)” (Wulff, 2001, p. 472).

(11)

Both Doling (2012, as cited in Ronald & Elsinga, 2012, p. 35) and Xu et al. (2015) describe a strong economic aspect of the life cycle model by outlining that a person goes through roughly three stages in life in which demand for owner occupied dwellings is determined by income: the first stage occurs at a young age, when the income is not yet sufficient enough to cover the entire costs of consumption; this changes in the second stage when people earn their own money, make savings and purchase a dwelling as soon as their income allows them to have a mortgage; the last stage occurs with retirement when most people paid of their mortgage and live of their savings (Doling, 2012, as cited in Ronald & Elsinga, 2012, p. 35; Xu et al., 2015). Perhaps the economic life cycle model may partly explain the housing careers of some households, but as Andersen (2010) points out, other – non-economic – factors are very important in determining residential choices of households.

To conclude, whether or not the life cycle model takes an economic form or not, it implies the existence of differences between age groups and the corresponding life events. However, this does not necessarily imply that the life cycle model is also able to grasp generational differences, since it is stated that every household more or less has the same housing career. Nevertheless, in the introduction it was stated that millennials seem to behave different on the housing market. How can their housing career be explained when it seems so different from the housing career of former generations?

In analysing the academic literature on housing choice, it becomes clear that the life cycle model is not only much cited and acclaimed. Many academics rather criticise the framework for analysing contemporary housing market behaviour by its simplistic view on people’s demand for housing (Molgat & Vézina, 2008; Wulff, 2001). Here it is argued that, indeed, the life cycle model does not completely grasp the behaviour and tenure choice of young people that move on the contemporary housing market and that the model may be out-dated. Historical and institutional changes in Dutch society and on the Dutch housing market, such as deferred marriage and a declining social housing supply have changed how people shape their life courses (Thorns, 2012, as cited in Ronald & Elsinga, 2012, p. 201).

Indeed, contemporary research found that the way young people move on the housing market today, often cannot be described as a linear housing ladder, but rather as a multitude of housing pathways, which are often far from linear and more chaotic in essence (Hochstenbach & Boterman, 2014; Wulff, 2001). This new life course model is able to frame changes in the way people move on the housing market (Xu et al., 2015). Wulff (2001) adds that an “advantage of the life course framework over the earlier life cycle model is that it highlights both the dynamic nature of households and the fact that stages in a life course model vary in

(12)

timing, duration and arrangement.” (p. 472). Although the revised, life course model provides opportunities for generational comparison, it is very important to realize that generations are not homogenous groups in which all members act the same (McKee, 2012). In other words, within generations large varieties of households and housing tenure preferences of these households exist.

A practical way of looking at this variety of households’ preferences for certain types of housing, is by looking at it through a geographical lens (Karsten, 2007). Most time-geographical research on housing preferences focuses on families and/or gender differences. Sometimes, researchers within this strand also point towards the social ties of households. As Karsten (2007) argues, “this social engagement is one reason why households with children tend not to move as often as singles and childless couples” (Karsten, 2007, p. 85) who are considered to be much more flexible. Looking at young one-person households in this way, the fact that one-person households are both living alone and have a young age may have large implications for their housing tenure preferences. These households are simply more flexible and less tied to a certain geographical place. The question is however, to what extent their social ties actually determine their tenure preference.

1.1.3 Homeownership or rent: a choice of tenure

The factors that influence tenure choice are much described in housing literature. Esveldt & De Jong (2013) argue that most people want to improve their living comfort by moving. Usually this means a larger home of better quality in a living environment that meets more of the households’ requirements. On average in the Netherlands, owned dwellings are larger than rental dwellings and mostly of better quality. Furthermore, homeownership is often believed to provide for long-term financial security. However, the financial risks grew in the last years due to the economic crisis of the late-2000s, which make homeownership more risky (Esveldt & De Jong, 2013).

Andersen (2011) found “three factors that are decisive for preferences for homeownership: long-term financial considerations; short-term financial considerations; and possibilities for use and disposal of the dwelling.” (Andersen, 2011, p. 204). According to Andersen (2011) motives for renting are a little less financially based than the motives for buying a home and point towards different lifestyles. He argues that this is mostly true for households without children and especially, for singles. However, although Andersen (2011)

(13)

does take lifestyles into account, these are solely practical and financial reasons. The approach does not take ideological structures into account.

In order to gain more insight in the choice of households to live in an owner occupied dwelling it is important to reveal the ideological structures behind homeownership (Ronald & Elsinga, 2012). Karsten (2007) argues “the goods that people own can be considered to be an expression of the self. Housing is one form of consumption through which the self can be expressed” (Cooper Marcus, 1995, as cited in Karsten, 2007, p. 86). Moreover, owning a home can provide for a feeling of status of the household (Esveldt & Elsinga, 2013).

Although ideological structures already give more insight in the way in which people choose a certain type of tenure, the approaches so far do not really look at the different groups in society and how these are shaped by the larger structures within this society. In addition, household formation and tenure choice are closely connected (Byrne, Duffy and FitzGerald, 2014). Therefore, the second part of this literature chapter will examine demographic studies with a particular focus on the importance of a generational life course perspective in forming households and residential preferences.

1.2 Demographic studies

Earlier in this chapter, it is described that the housing market consists of a supply and a demand side. Furthermore, it is explained that there are ideological structures underpinning housing tenure choice and preference. In the paragraph on housing demand it became clear that several approaches can be used to examine housing preferences and housing choice of households. The life course perspective was considered to be of most interest for this research.

In her oration on the importance of demographic analysis in urban planning, Dorien Manting (2013) argues that research from a life course perspective and the effect of changes in the extent and composition of the population is important in urban development and policy. Here it is argued that it is not only important in formulating policy but that, consequently, a demographic aspect is also essential in research on the housing market and housing choices because it are the households that are vital for the housing market. After all, policy is based on research.

Based on this argument, this literature part addresses demographics. Since this research focuses on a particular generation (generation Y) it is important to have a clear conception of what the term generation encompasses. However, as mentioned earlier in the

(14)

introduction, this research does not target generation Y as a whole. Looking at a particular group within this generation, namely one-person households, narrows the focus down. As a consequence of this focus, here it is attempted to contribute to a geography of living alone.

1.2.1 Generation Y

Within the social sciences, youth in the form of generations is not often explicitly discussed. Instead, youth studies conceptualise “youth as primarily a transnational stage or phase of life” (Wyn & Woodman, 2006, p. 495). However, as Edmunds & Turner (2005) and Wyn & Woodman (2006) argue, generation studies are increasingly getting attention within the structural functionalist strand of sociology. Demographic change and its repercussions in politics are of particular interest. Academics focus on how different societal structures influence inter-generational inequality and conflict (Edmunds & Turner, 2005; Wyn & Woodman, 2006).

Already in 1983 David Kertzer reviewed the different uses of the term generation by social researchers. He argued that there are roughly four distinct usages of the term, each grounded in different disciplines within the social sciences. Within the social anthropology generation encompasses (global) kinship relationships. Demographers however, mostly examine generation as a cohort. Generation in this sense, points towards people’s life cycle from a young to an old age (Kertzer, 1983). Although viewing generation from this perspective has advantages for operationalization of the concept, it lacks in examining societal structures (Edmunds & Turner, 2005). Other social scientists do take these societal structures into consideration by referring to a generation as a life stage in which conflict with other generations arises out of the different responses of differently aged people on societal events. Less common in the social sciences, but more common in history, is the perception of a generation covering multiple cohorts in which a large historical event defines a certain generation. An example of a generation as an historical era is the so-called ‘pre-world war II generation’ or the ‘baby-boom generation’ (Kertzer, 1983).

Probably the most cited sociologist within the generation literature is Karl Mannheim (1970). Mannheim stated that people that have approximately the same age in the same period of time live through the same societal conditions that shape who they are and how they behave (Mannheim, 1970). A focus on this social-historical type of using generation in social research “shifts the emphasis from the assumption of linear development in which youth is a phase towards adulthood, to locate young people within the political, economic and cultural

(15)

processes that both frame and shape their generation, and the meaning and experience of ‘youth’ in distinctive and enduring ways.” (Wyn & Woodman, 2006, p. 495).

This influence of societal processes is also visible in both academic literature and more popular writings on generation Y. Often, generation Y is characterized as a generation that has to deal with a more unstable labour market, flex working and job insecurity. However, sometimes the media also make statements about members of generation Y that are not willing to work a long-term job and prefer the insecurity of short-term labour. Jamieson & Simpson (2013) also discuss globalization and modern technology as societal factors that influence the way in which young people behave and form households. They argue that being connected with other people at any moment of the day, through for example smartphones, may be a force that contributes to young people forming one-person households (Jamieson & Simpson, 2013).

1.2.2 Living alone: the emergence of the one-‐person household

Since the Second World War household sizes in Western European countries are decreasing, which caused a new type of household to arise: the one-person household (Wall, 1989). The reason behind the emerge of the one-person household is often sought in the individualisation theory of modern sociologists like Zygmunt Bauman (2001), stating that modern western society changes the way in which people form relationships and treat communal life. “Individuals become increasingly disembedded from traditional forms of support and constraint (e.g. family, ethnic group, neighbourhood), and are located instead as individuals in social, legal and welfare systems that foster and encourage independence.” (Beck and Beck-Gernsheim in Hughes, 2014, p. 707).

Almost all authors that write about one-person households point out that - the same as for generations - they cannot be treated as one homogenous group of people. Instead, there are multiple groups “each with its own preferences and perspectives.” (Faessen, 2002, p. 350). Often a distinction between old and young one-person households is made because these groups are considered very different in their housing choices (Faessen, 2002; Moore Fisher & Graham, 1974). Faessen (2002) continues his argument by stating that despite the differences in preferences, the majority of one-person households seem to have a preference for renting.

As Jamieson and Simpson (2013) state in their book on living alone, doing research on one-person households is of particular interest because a lot of “confusion of popular stereotypes and messages about this growing trend” (Jamieson & Simpson, 2013, p. 1) of

(16)

living alone exists. In addition, there is a large discussion on whether or not the worldwide rise in one-person households is an effect of globalization and the associated trend of individualization (Jamieson et al., 2009; Jamieson & Simpson, 2013).

One of the most common prejudices about one-person households has its roots in the life cycle model and is based on the presumption that it is a temporary household formation (Jamieson & Simpson, 2013). However, although Jamieson and Simpson (2013) do state that this and other prejudices exist, they do not succeed in explaining that living alone can also be a long-term way of living. Molgat and Vézina (2008) cover this unknown field by outlining three different ‘representations of living alone’, namely: (1) living alone as a youth lifestyle; (2) living alone as a transitional period; (3) living alone as a way of life for the long term.

Another interesting concept that can be derived from the articles of Jamieson and Simpson (2013) and Molgat and Vézina (2008) is what here is called the geography of living alone. Although the authors do not exactly use this concept it seems logical and interesting to combine statements from these authors under this name because it is stated that the geographical place has a context-bound, reciprocal relationship with the types of households that live in that place. Jamieson and Simpson (2013) outline three European and one American family system that have certain implications for one-person households of which the north-western European system and highly urbanized places are considered the most suitable for living alone (Jamieson & Simpson, 2013; Molgat & Vézina, 2008). Molgat and Vézina (2008) are taking this argument even further by stating that some geographical places are contributing to living alone, although it remains unclear how that process works.

Nonetheless, the geographic aspect of one-person households seems of particular interest in consideration of the widespread trend of people that live alone, but also because some authors point toward a reciprocal relationship between one-person households and geographical space.

(17)

2. Dutch societal context: a historical overview

In the previous chapter a lot is already written on generation Y and how this generation differs from other generations by behaving in a different manner. It became clear that this difference lies not only in how the members of this generation form households, but it is also widely believed that they have different housing (tenure) preferences. However, since this is a research focussing on the Netherlands, and considering the attempt to answer the research question from a generational perspective, it seems important to go more in depth about the environment of generation Y in the Netherlands and especially, the societal events that they face and shape who they are.

Therefore this chapter encompasses a historical overview of the main societal trends in the Netherlands over the last 30 to 40 years. This overview is based on the social and cultural reports of ‘The Netherlands Institute for Social Research’ (SCP), in which - since 1974 - the SCP reviews Dutch societal trends every two years (SCP, 2009). The chapter is structured according to the argument of Wyn & Woodman (2006), discussed in the theoretical chapter, namely that it is important to view people within the cultural, economic and political processes within society (Wyn & Woodman, 2006). Therefore, these will form the framework of the chapter. Within the descriptions of the different contexts, demographics and housing play an important role.

2.1 Cultural societal context

Already in 1994 the SCP wrote about individualisation within Dutch society. As used by the SCP, individualisation means the growing autonomy of the individual in relation to his or her direct environment. This process is considered to occur in relation to the (social) ties of which an individual is part. Furthermore, it relates to the moral order in which the individual, is embedded through those ties. In 1998 the SCP reports that the time between 1970 and 1998 is characterised by the “penetration of individualisation within the last and strongest bastion of traditional society, the family” (free translation, SCP, 1998, p. 3). In this period of time, people started to conceive marriage and the family in a different way. Where most people in the 1970s used to get married and have children in their early or mid-twenties, by 1998 this was far less common. Additionally, not only the overall opinion on deferred marriage had changed, also divorce was now seen as more ordinary. In short, the prevailing norms and values about marriage and family in Dutch society changed significantly (SCP, 1998).

(18)

Of great interest for this study is the remark that individualisation caused an increasing amount of household types. The traditional stages of household formation that almost all people followed in a strict order were now replaced by more diverse trajectories associated with individual preferences. The most remarkable result of individualisation, according to the SCP (1998), is the increase in one-person households. From 1970 until 1998 the amount of one-person households in the Netherlands doubled. However, as the SCP (1998) notes, individualisation finds a natural boundary because of the fact that almost everyone wants to live with a partner at a certain stage in adulthood (SCP, 1998).

Between 1970 and 1995 the Dutch population increased dramatically with 19%. Migration surplus played a big role in this increase. Not only the population increased dramatically, the amount of households increased even more strongly with 63%. This was partially due to a changing demographic composition with more elderly living in the Netherlands, but also, and perhaps even more, to the earlier-described individualisation trend, which meant an increasing appreciation for household independency. Besides, an increasing amount of young people that left the parental house was going to live alone, instead of living with a partner (SCP, 1998).

In 2000 the SCP reported that Dutch people turned themselves away from traditions and wanted the freedom to live their lives as they want to. The formation of new types of households was an ongoing trend in the early 2000s (SCP, 2000; 2002; 2004). Another topic that was of great concern in the early 2000s was the integration of ethnic minorities in Dutch society. Where native Dutch civilians thought that ethnic minorities did not adjust enough to the Dutch culture, ethnic minorities felt less accepted (SCP, 2004). Already in the second half of the 20th century, ethnic minorities started to segregate in space. This was still visible in the early 2000s. The migrant population in the Netherlands was a young population that was especially living in the four big cities (Amsterdam, Rotterdam, Den-Haag and Utrecht). In 2004 the SCP wrote that the negative effects of ethnic segregation in the Netherlands became more visible. Degradation of the living environment and high crime rates were of great concern (SCP, 2004).

Interesting is the big role of ICT in the SCP reports from 2004 onwards. Informatisation of society became a large trend in the last decennium of the 20th century and this was further expanded in the early 2000s (SCP, 2004; 2006). The trends in ICT usage were very visible in a lot of aspects of society, not only in the economic and political field, but also in the cultural field. ICT usage sharply increased in throughout the beginning of the 21st century. ICT usage became an important aspect of leisure activities, especially among young

(19)

people (SCP, 2006). As a result, in 2006 the SCP concluded that the Netherlands was turning into a knowledge society.

In the housing field, the mid and late 2000s showed an increasing focus on liveability of neighbourhoods. Increasingly, local authorities were concerned with safe, clean and undamaged neighbourhoods that were considered to improve this liveability. More attention was placed on the role of civilians and the interrelationships and social cohesion between neighbours. Another way of addressing liveability was by the intervention on housing supply. Relatively cheap housing was demolished and replaced by more expensive housing (SCP, 2006; 2008). Often, the demolished housing was social rental housing, replaced by housing for homeowners (BZK & CBS, 2015). Besides, public space became a place of intervention (SCP, 2006; 2008).

By 2012, the largest social issue in the Netherlands was the responsibility of civilians within society. Civilians became more assertive and the SCP wrote about fading traditional norms and values. As mentioned before, this trend already occurred in the second half of the 20th_{century but by 2010 this would become a major item on the political agenda, with}

repercussions visible within society (SCP, 2012). In 2014, civic responsibility still played a big role in the social and cultural report, but the tone was slightly different. Now, the focus was more on differences in society and how these inequalities could be overcome (SCP, 2014).

2.2 Economic societal context

From the mid-60s onward, until about 1980, economic inequality in the Netherlands decreased. Lower income households experienced a relative increase in their income and an increasing amount of households belonged to the middle class. Social awareness and solidarity with disadvantaged groups in society became important factors in the dominant ideology. A countermovement to this ideology emerged with the economic crisis in the late 70s. Emancipation movements and solidarity could not provide the welfare to decrease. In contrast, people accused these movements and ideology of being partially responsible for decreasing welfare. The countermovement propagated corporate identity as an alternative for solidarity. From this moment onwards it was morally allowed to proliferate personal success.

From 1985 onwards income inequality increased. Welfare recipients were increasingly less prosperous than working people. Between 1985 and 1996 this inequality continuously

(20)

rose. This trend was not only visible on people’s bank accounts. Increasing inequality started to precipitate in space when the population with a relatively low income (often immigrants) were increasingly concentrated in certain areas. Especially the four largest cities in the Netherlands had a lot of citizens depending on public benefits. This was not actually the result of increasing inequality, but rather a result of a “specific spatial dynamic” (SCP, 1998, p. 14), namely suburbanisation. The government tried to reverse this process of selective suburbanisation and immigration by triggering the regeneration of cities through public housing and economic incentives (SCP, 1998).

Public housing in the Netherlands, in the form of subsidised rental dwellings, resulted in a housing supply of high quality. However, economic welfare was wider spread than housing supply, which meant that on the one hand people with a relatively low income were dependent on individual housing allowance, and on the other hand, people with a relatively high income were living in housing that did not meet their welfare needs (SCP, 1998).

In the early 2000s the SCP (2000; 2002; 2004) reported a strong increase in economic prosperity. Economic development was attributed to the influence of the information and communication technology that was considered to guarantee a persistent economic growth. In 2000, the SCP noted that in no other country in the European Union, the relationship between income and homeownership was so tight as in the Netherlands. In the early 2000s, the social rent sector was still large and this might be a possible explanation for the tight relationship between income and type of tenure (SCP, 2000; 2002).

By 2014, the SCP wrote that housing was an important indicator for differences within society. Differences in economic capital were ratified when different population groups were socially and spatially segregated. Although not much, the highest and lowest incomes in the Netherlands were converging. However, it was concluded that income differences formed a parallel with other forms of capital (for example social capital), which made differences more significant. Dutch society existed increasingly of different worlds, with income as largest effector (SCP, 2014).

Besides income, another important aspect of the economic context is labour. Technological development was a very influential trend in labour productivity from the mid until late 20th century. The influence of this development proceeded to a great extent through the production process and was therefore visible in labour productivity that increased with 71% between 1970 and 1998. However this increased gain in labour productivity was especially visible in the more capital-intensive sectors and was limited by the increasing interest in the more labour intensive service sector (SCP, 1998).

(21)

Between 1971 and the beginning of the 21st century, the employment rate, after a downfall in the 70s, rose substantially. As mentioned before, there was a major shift from production labour to service labour, which meant an overall increase in the level of labour. However, the amount of people with a full time job strongly decreased between the 1970s and late 1990s. This was partially due to the greater flexibility of employment and for the other part due to the increasing popularity of part time jobs (SCP, 1998).

Overall, the late 20th century showed a dramatic increase in employment and this trend continued into the early 2000s (SCP 1998; 2000). The unemployment rate in the Netherlands was one of the lowest of all European countries. Especially the labour participation of women further increased. Many profited of a relatively strong economic expansion. In other words, the extra work was distributed over many jobs. A likely explanation is the tempered development of wages. Besides, the amount of part time jobs increased (SCP, 2000; 2002; 2004). Although the early 2000s knew a high employment rate, this was strongly divided along ethnic lines. In 2004, only half of the non-western migrants had a job. To a large extent, this could be attributed to low employment rates among non-western origin women and the relatively young age of the migrant population. If non-western migrants had a job, these were often low paid jobs (SCP, 2004).

Again, a substantial trend in the early 2000s was the increase in labour in the relatively new ICT sector. Informatisation was strongly connected with flexibility of working hours and working space. The Internet made it increasingly possible to work in other places than the office (SCP, 2004). According to the SCP, the development towards a knowledge society was mirrored in the increase in ICT wise employees. Besides, the overall education level of the Dutch population was increasing and this made it possible to handle the stronger information flows of the 21st century (SCP, 2006).

Another trend that became visible in the mid and late 2000s was the increase of members of ethnic minority groups in better positions on the labour market. Migrants increasingly populated the middle class. Their education level and income increased. Besides, more migrants were now active in politics, arts and cultural sectors. This was a huge change in comparison with their position in the 90s and before. Moreover, the amount of women with a non-western background in a high labour position rose substantially (SCP, 2006; 2008; 2010).

Besides the improved position of ethnic minority groups and women on the labour market and the increased flexibility, public debate on labour did not only have positive connotations throughout the 2000s. Especially after 2010, the position of young people on the

(22)

labour market was increasingly mentioned. Increased flexibility would cause people to have higher job insecurity. People increasingly had short term labour contracts (SCP, 2010; 2012).

2.3 Political societal context

Until far after the Second World War, the government was “an authority that was hierarchically placed above civilians, but in which these civilians could participate through suffrage” (free translation, SCP, 1998, p. 9). After this era, politics and governance were more characterised by negotiation. This new system implied the existence of more or less equal partners whose rights and duties are constantly under discussion. This model was visible within the relationship between government and civilians but also and maybe even more, within the relationship between public organisations.

The 60s and 70s were the heydays of democracy and political participation in the Netherlands. With retention of the participation democracy, the government changed its focus in the 80s and 90s. From the 80s onward, the government proliferated itself as an organisation surrounded by other equivalent organisations. The new negotiation culture went along with a rehabilitation of the market. This phenomenon went hand in hand with upcoming neoliberalisation in the 80s and was strongly reflected in the actions of the government. Government departments started to present themselves as companies that deliver products to civilians. Society turned more and more into a corporate environment.

The early 2000s can be typified as a time of political satisfaction. The overall Dutch population showed trust in national politics. As a result, political interest was relatively stable between 1970 and the early 2000s. As the SCP notes in the social cultural report of 2002, especially freedom of speech was important for Dutch civilians. Civil organisations were considered to be very important as intermediary institutions with regard to the democratic functioning of Dutch society. The organisations formed a manner in which civilians could influence national politics.

In the report of 1998 the SCP already wrote about individualisation. In 2004, this again gained a lot of attention in the social cultural report. The SCP reported that traditional organisation structures were changed into networks by individualisation, informatisation and informalisation. These networks, was noted, were sometimes completely virtual which was a new phenomenon. According to the SCP there was a trend in which personal traits and interests determined membership to certain organisations instead of background and birth.

(23)

According to the SCP political organisations where changing by modern technologies and media, which made political expression and mobilization easier (SCP, 1998; 2004).

Informalisation and democratisation of society were expressed politically within the shift from government to governance, but were also expressed by increasing public accountability and transparency. This made policy development in the Netherlands more interactive. Again, this was encouraged by the increasing access to and use of the Internet and modern media (SCP, 2004).

In the mid 2000s, the SCP reviewed the role of religion within civil society and politics. They concluded that an increasing amount of the Dutch population did not go to church and that this had repercussions in the way people voted. Politics and religion became increasingly independent. However, this was not particularly a new trend since this was already visible in the 20th century (SCP, 2006).

By 2012 public debate had shifted towards a different role of the government. Increasingly a focus was placed on civic participation. In other words, civilians were expected to be responsible and the shift from government to governance that already took shape in the early 2000s became starker. Civic participation, as was believed, would cause a greater legitimacy of policy and a decreasing gap between government and civilian. According to the SCP (2012) this trend was not only initiated by the government, the civilians themselves were ready for more responsibility and wanted more freedom. Increasing civic responsibility was often presented as a remedy for two social issues, namely the crisis of the welfare state and fading norms and values within society (SCP, 2012).

In 2014, the SCP wrote not only on increasing cultural and economic differences within Dutch society. Also the political insights and feelings of Dutch civilians seemed more divergent than ever before. Society would divide into groups that were increasingly different from each other in moral, cultural and political sense. In the SCP report of 2014 winners and losers of globalisation are mentioned. People with a good economic position were considered to be winners and people that have a worse economic position were considered losers. As mentioned before, this economic position was increasingly associated with other types of capital and had a large repercussion in political satisfaction (SCP, 2014).

(24)

3. Research question

Previously, it became clear that young people in the Netherlands increasingly live in one-person households. At the same time, although the government promotes homeownership, young people are believed to choose rental dwellings over owner-occupied housing. Despite the increase in one-person households among young people, the residential preferences of this group are not yet extensively examined in Dutch academic research and especially not within a generational perspective. In England, the media speak of ‘generation rent’, but considering the differences between the Dutch and English housing markets and their historical circumstances, it is not clear if this term is also applicable for young people in the Netherlands. Do Dutch young one-person households prefer renting or buying a dwelling?

Therefore, the purpose of this research is to examine the housing preferences of young one-person households by placing them into a generational life course perspective. This means that the assumption is that households do not follow a certain linear housing path such as the life cycle model suggests. Instead, within this perspective, there is specific attention for the individual, spatial and societal context that shape how young one-person households move and behave on the housing market and how this may differ among different generations. Therefore, the following research question is formulated: What are the housing tenure preferences of young one-person households of generation Y seen from a generational perspective, giving attention to individual, spatial and societal circumstances?

3.1 Sub-‐questions

In order to answer the above stated research question, the following sub-questions are formulated:

a. What are the housing tenure preferences of young one-person households in the Netherlands and how did they evolve in time?

b. What is the influence of individual circumstances on the housing tenure preference of young one-person households?

c. What is the influence of spatial circumstances on the housing tenure preferences of young one-person households?

d. What is the influence of societal circumstances on the housing tenure preferences of young one-person households?

(25)

3.2 Hypotheses

Corresponding with the above-formulated sub-questions, 4 (directional) hypotheses are composed:

H1: Young one-person households from generation Y are less likely to prefer homeownership

over rental dwellings than young one-person households from former generations.

H0: Young one-person households from generation Y do not differ from one-person

households of former generations in their housing tenure preferences.

H2: Individual features of the household influence the housing tenure preferences of young

one-person households.

H0: Individual features of the household do not influence the housing tenure preferences of

young one-person households

H3: Spatial features of the households’ living environment influence the housing tenure

preferences of young one-person households.

H0: Spatial features of the households’ living environment do not influence the housing tenure

preferences of young one-person households

H4: Societal features influence the housing tenure preferences of young one-person

households.

H0: Societal features influence the housing tenure preferences of young one-person

households.

(26)

4. Concepts and operationalization

In this chapter, first the most important concepts of this research will be defined and explained. Second, these concepts will be broken down into dimensions, variables and indicators in order to make the concepts measurable.

4.1 Definitions

Housing tenure preferences

In this research, housing tenure preferences refer to so-called stated tenure preferences. Housing tenure preferences encompass the ‘relative attractiveness’ (Jansen et al., 2011) of a certain type of dwelling (rent or owner-occupied) for a particular household. This makes stated housing preferences distinctive from housing choices or revealed housing preferences, which “refer to actual behaviour” (Jansen et al., p. 2). “The most important difference between housing preference and housing choice is that preference is a relatively unconstrained evaluation of attractiveness.” (Jansen et al., 2011, p. 2).

Generation

Throughout this thesis the term generation refers to people that have approximately the same age in the same period of time and therefore live through the same societal conditions that shape who they are and how they behave (Mannheim, 1970). In order to distinguish between different generations the generation theory of Aart Bontekoning (2014) is used. The key point of his generation theory is that successive generations are focused on the evolution of the social system where they are part of. Within this evolution new generations replace the old societal patterns created by older generations with new patterns (Bontekoning, 2014). In compliance with Bontekoning (2014) the table below (table 1) shows the generations that can be distinguished in the Netherlands.

(27)

Table 1: Generations in the Netherlands

Generations Born between

Pre-war generation 1905-1920 Silent generation 1920-1935 Protest generation 1935-1950 Connecting generation X 1950-1965 Pragmatic generation 1965-1980 Authentic generation Y 1980-1995 Conscious generation 1995-2010

Source: Own adaptation of Bontekoning, 2014

Generation Y – millennials

Members of generation Y (in this research also referred to as millennials) are born between 1980 and 1995 (Bolton et al., 2013). In compliance with Bolton et al. (2013) a key characteristic of millennials is that they are exposed to technology frequently, “which has advantages and disadvantages in terms of cognitive, emotional and social outcomes” (Bolton et al., p. 247). In addition, “members are now experiencing an era of economic uncertainty and violence (…) and the worst global recession since 1929.” (Bolton et al., 2013, p. 247).

One-‐person household – living alone

In this research the definition of a one-person household as used by Statistics Netherlands (CBS, 2016) is used. A single person is “someone who provides for his or her own housing and essentials in a private, non-commercial manner.” (CBS, 2016). A single person forms a one-person household. Persons that live with others on one address but carry a household by themselves are also considered to be a one-person household. Singles can have all kinds of marital statuses. An example is when people live alone while waiting to finalize a divorce (CBS, 2016). In the operationalization the different typologies of one-person households will be further examined.

(28)

4.2 Operationalization

Above, the main concepts of this research were already briefly defined and explained. In order to make the concepts in this research measurable, an operationalization of the concepts is necessary. In table 3 first, the dimensions of the different concepts in this research are outlined. Second, these dimensions are being further divided into variables. Finally, indicators are defined to indicate the amount, extent or nature of the formulated variables.

Housing tenure preferences

A crucial division that is made in research on housing preferences and of great importance for the way in which housing preferences are examined, is the division between revealed and stated preferences (Andersen, 2011; Coolen & Hoekstra, 2001; Floor & Van Kempen, 1997; Timmermans et al., 1994). In his article on motives for tenure choice Andersen (2011) explains that both the revealed and stated preferences methods cause methodological problems. The methodological problem of the revealed preferences method seems somewhat obvious, namely “the actual housing situation does not necessarily reflect underlying preferences.” (Andersen, 2011, p. 185). According to Andersen (2011) the problem with the stated preferences method is that it does not look at how realistic the housing preferences of people are: some people do not take the possible constraints on the housing market into account. In the definition section, it was already explained that there is a sharp difference between residential preferences and residential choice. Because this research examines the preferences, despite the methodological pitfalls, it addresses the stated preferences of young one-person households.

One-‐person households

Earlier in the theoretical framework, it was mentioned that young one-person households cannot be treated as one homogenous group. Instead, a large variety of young one-person households exist. Based on academic literature it would be logical to make a division between different types of young one-person households by looking at their marital status. However, the amount of people that is married and/or divorced in the age category concerned is very small. Therefore, another way to make a division between different types of young one-person households is to make a division based on age. In the theory on generation Y it already became clear that the current stage of stage of millennials is the stage towards adulthood. The expectation is that generation Y can best be divided into two age categories that are outlined

(29)

in the table below (table 2). Although these age categories are not equally divided in terms of years, it is argued that they are equally divided in terms of lifestyle. The first group is relatively young and consists to a large extent of students. Besides, this age group is more moving inclined than the other age category (De Groot et al., 2008). The second group however, is relatively older and members of the group are more likely to be graduated, to have a full-time job and less moving inclined.

Table 2: Young one-‐person households by age group Age groups

20-24

25-34

Individual context

The individual context of young one-person households is divided into three dimensions: a socio-demographic, a socio-economic and a housing situation dimension that reflects the households’ current housing situation. The socio-demographic, socio-economic and housing situation dimensions encompass a few basic variables that are common in housing research, namely gender, age, education level, income, housing and tenure situation. Academic literature on young people living alone points towards a specific separation of living and personal relationships for one-person households (Jamieson & Simpson, 2013). Therefore, a logical variable within the socio-demographic dimension would be marital status. However, the amount of young one-person households that is married, divorced or widowed is very small and therefore the variable marital status is unusable.

Spatial context

The spatial context for young people that live alone encompasses the residential area of the household divided into four housing market areas. According to Clark & Dieleman (1996) the local housing market influences the way in which people form households and their residential preferences and therefore it is important to look at the local housing supply and residential area. Another variable that would have contributed to the spatial context of young one-person households is urbanity rate of the residential environment. However, the variable urbanity rate was not available within the dataset for all years of interest.

(30)

Societal context

Based on the argument in the generation study of Wyn & Woodman (2006) the societal context of young one-person households is divided into three dimensions, here reduced to one due to measurement reasons: a time dimension. Therefore, the year in which the Dutch housing survey was conducted is converted into the categorical variable year.

(31)

Table 3: Operationalization

Dimensions Variables Indicators

Individual context Socio-demographic Gender Man

Woman Age 20-24 25-34 35-44 45-54 55-64 65-74 75+ Socio-economic Level of education Low

Middle High

Income Low

Middle High

Housing situation Tenure Owner-occupied Rented dwelling Housing type Single-family dwelling

Multi-family dwelling

Spatial context Residential area Housing area by province

combinations North (Groningen, Friesland, Drenthe) East (Overijssel, Gelderland) West (Noord-Holland, Zuid-Holland, Utrecht, Flevoland) South (Zeeland, Noord-Brabant, Limburg)

Societal context Time Year in which the housing survey was conducted 1985 1990 1994 1998 2002 2006 2009 2012 2015

Tenure preference Stated Tenure preference Prefers to live in an owner-occupied dwelling Prefers to live in a rental

dwelling Source: own adaptation of WBO & WoON (1985-2015)

(32)

4.3 Conceptual model

Above, the operationalization of the concepts in this research was outlined in a table (see table 2). In order to visualize the investigated relationship of these concepts conceptual model below (figure 1) was created.

Figure 1: Conceptual model

(33)

5. Methods

This chapter explains the methods that are used to conduct this research. In the first part of this chapter, the research strategy and design will be discussed. Here, the different reasons for the choice of method are outlined and explained. Second, the units of analysis will be examined. In this part, more practical information on young one-person households in the Netherlands will be given to underset the research and understand it within the demographic context of the Netherlands specifically. Subsequently, the third part of the chapter focuses on the data and analyses that are used. Here, more will be explained about the Dutch housing survey and why it is particularly suitable for researching the stated housing preferences of young one-person households. The fourth and final part of this chapter briefly discusses the data preparation that was done prior to the actual analyses.

5.1 Research strategy and design

This research examines the stated housing tenure preferences of young one-person households in the Netherlands. As Hooimeijer (1994, as cited in Jansen et al., 2011) points out, there are many ways to examine housing preferences. What method is chosen to measure housing preferences is very much context-bound and depends on the purpose of the research. In addition, the outcomes of research on housing preferences are different for each method but are not better or worse than other methods. Therefore, choice of method in housing preferences research is based on the research question and “the information in which one is interested” (Hooimeijer, 1994, as cited in Jansen et al., 2011, p. 12).

In order to study the tenure preferences among young one-person households aged between 20 and 34, a quantitative research strategy is used. Alan Bryman (2008) describes that a quantitative research strategy involves a deductive approach in which a hypothesis is, or multiple hypotheses are tested. In order to examine a generational effect on housing preferences of one-person households of generation Y, a longitudinal design is used. This means that the housing preferences of young one-person households are researched not only for members of generation Y, but also for preliminary generations.

Where a cross-sectional design would only be able to reveal patterns on one moment in time, this “longitudinal design allows some insight into the time order of variables” (Bryman, 2008, p. 49) and therefore is able to reveal differences between generations in order to contribute to a comprehensive overview of generation Y and the unique circumstances that

Housing tenure preferences of young one-person households in the Netherlands : a generational perspective