Footprints of an
invisible population
Second-home tourism and its
heterogeneous impacts on
municipal planning and
housing markets in Sweden
Andreas Back
Department of Geography Umeå 2020
This work is protected by the Swedish Copyright Legislation (Act 1960:729) Dissertation for PhD
ISBN: 978-91-7855-347-1 ISSN: 1402-5205
Gerum - Kulturgeografi: 2020:3 Copyright © Andreas Back Cover art by Jonny Bylund
Electronic version available at: http://umu.diva-portal.org/ Printed by: CityPrint i Norr AB
Sixty square metres of heaven of Earth A tiny wooden paradise
It’s my own private pinewood Taj Mahal Except for the shape and the size
The cabin!
Where I come to relax The cabin!
Wear the same pants for a week The cabin!
I’ve got skis on my walls In my cabin
Excerpt from the song The Cabin, by Ylvis (2014)
Acknowledgements
In what seems like a long time ago, in a galaxy far, far away, I was handing out flyers in one of the suburbs of Umeå for my then employer. As a way to make the job a bit less dull, I decided to refresh my memory of Karl Marx’s Capital by listening to David Harvey’s podcasts on the subject. Despite my ever-sorer feet, I remember it as an exhilarating experience. In my head, I left the suburban apartment blocks and entered a fascinating world of exchange values, surplus values, commodities, economic crises and whatnot. The fact that Harvey was a geographer thrilled me. Could this really be – geography? If so, then what on Earth was I doing here, mindlessly pushing flyers down letterboxes? I seem to remember that this was when and – very important for a geographer – where I first started dreaming about returning to university to study human geography with the ultimate goal of a PhD. It feels almost surreal to be at the end of that journey, almost ten years later, but there it is.
When speaking of galaxies far, far away, this thesis would not have been the same without the first-rate guidance and patience of my supervisors Roger Marjavaara and Dieter Müller. They are the Jedi Masters of second-home research to whom I have been a Padawan during this work. No matter how stressed or confused I have felt, they have always given me good advice, new ideas and confidence. They have also been the first and foremost to provide constructive (and much needed!) critique all along the way. For that, I am deeply grateful. In addition, I want to thank Kjell Overvåg, University of Agder, for taking the time to read the thesis draft and providing insightful comments. I am also indebted to Brett Christophers and Jan Amcoff at Uppsala University for inspiring discussions, supervision and encouragement during the studies that led me here.
Working with this thesis would never have been the same without the great colleagues at the Department of Geography at Umeå University. Special thanks to comrades Joakim Byström, Desiree Enlund and Emelie Hane-Weijman, but also Aina Tollefsen, Erika Sandow, Guilherme Kenjy Chihaya da Silva, Heather Mackay, Irma Olofsson, Lars Larsson, Linda Lundmark, Madeleine Eriksson, Magnus Strömgren, Marcin Rataj, Rikard Eriksson, Traian Leu and the rest of you. In addition, extra thanks to Lotta Brännlund, Fredrik Gärling and Erik Bergström, for forming the department’s solid bedrock.
One of the best things about being a doctoral student was seeing new places and meeting people. Some of these meetings resulted in memorable anecdotes, while others gave me new friends. Like that reception in New Orleans when I introduced myself to a Canadian researcher who instantly replied ‘Ume-oh? Is that close to Ornskoldsweek? My favourite hockey player is from
Ornskoldsweek!’ Or the many one-liners delivered by C. Michael Hall, like when he responded to the question on how many geographers it would take to get lost, by saying ‘None. Never. We would start by deconstructing the meaning of lostness!’ Like Outi Kulusjärvi, with whom I have had lots of funny and interesting discussions, ranging from Finnish candy, over the woes of writing a doctoral thesis, to the end of capitalism. Like Gijsbert Hoogendoorn, my favourite South African metal-head, second-home researcher and history buff, and Fatima Vally, the most humble person in the world. A very special baie fokken dankie goes out to you two for hosting me in Johannesburg, being generally awesome, introducing me to Sandy Wang and answering my constant questions about South Africa and Afrikaans.
I am particularly grateful for the weekly role-playing sessions and friendship of Adrian Löwander, Björn Palmberg, Fredrik Andersson, Jesper Enbom, Mattias Gunéll, Tobias Andersson and Ulf Johansson. You have been – as ever – excellent friends and comrades during good and bad times these past few years. The same is also true for Caroline Täljeblad, Daniel McCord, Erik Nordlund, Frida Grimm, Helena Back (the PhD mug is finally a PhD’s mug!), Johan
Örestig, Martin Johansson, Martin Kehlmeier and Peter Johansson. Big thanks to Erik Lindenius for the blazoned coat of arms to use as a knighted second-homes researcher, and to Jonny Bylund for the amazing cover art. Special thanks also goes out to my crew of imageflippers who gave me a purpose when I needed it the most.
Finally and most importantly, I want to thank my parents Vanja and Sören, my bonus parents, siblings and cousins for all the encouragement and cheers along the way. Thank you Hanna, for putting up with my never-ending ramblings about work in general and this thesis in particular. I am deeply grateful for your love, wit and companionship. Last but not least, to my magnificent daughters Lo and Tora: you are me and I am you. I began working on this thesis right in between you two were born, so I can’t help feeling like a third belated sibling has finally seen the light of day. This time, however, I was doing the labour.
There is no royal road to science, and only those who do not dread the fatiguing climb of its steep paths have a chance of gaining its luminous summits.
Karl Marx, Capital (1997, p. 15)
Andreas Back
Table of contents
Acknowledgements
Appendices ... 2
I. Introduction ... 4
Background ... 4
Aim and research questions ... 6
Outline of the thesis ... 6
II. Setting the scene ... 9
Permanent and temporary: producing an invisible population ... 9
A case for landscapes ... 10
Defining second homes ... 12
Swedish municipalities: planning, housing, and disparities ... 14
III. Research design ... 20
Methodological considerations ... 20
Quantifying second homes ... 23
Interviewing municipalities ... 24
Combining multiple methods ... 26
IV. Footprints of an invisible population ... 29
A. How can spatial variation in the impacts of second-home tourism be analysed systematically on a national level? ... 29
B. How do Swedish municipalities perceive and manage impacts of second-home tourism? What are the spatial differences? ... 34
C. How is the Swedish housing market impacted by second-home tourism? What are the spatial differences? ... 46
V. Discussion ... 53
Second-home tourism – an umbrella concept ...53
Suggestions for further research ... 54
Policy recommendations ... 55
Concluding remarks ... 59
VI. Sammanfattning (Swedish summary) ... 62
Appendices
Paper I: Back, A. & Marjavaara, R. (2017). Mapping an invisible population: the
uneven geography of second-home tourism. Tourism Geographies, 19(4), 595-611.
Back is responsible for the data handling and writing. The planning and analysis were a joint effort by both authors.
Paper II: Back, A. (2020). Temporary resident evil? Managing diverse impacts
of second-home tourism. Current Issues in Tourism, 23(11), 1328-1342.
Back is responsible for the planning, data collection, analysis, and writing of the paper.
Paper III: Back, A. (forthcoming). Endemic and diverse: planning perspectives
on second-home tourism’s impact on Swedish housing markets.
Back is responsible for the planning, data collection, analysis, and writing of the paper.
Paper IV: Back, A., Müller, D. K., & Marjavaara, R. (forthcoming). The invisible
hand of an invisible population: impacts of second-home tourism on Swedish housing markets.
Back is responsible for the data handling and empirical analysis. The planning, writing, and final analysis were a joint effort by all authors.
Twist the map
You haven’t got it right Geography
I’m gonna show you now Twist the map
You haven’t got it right Geography
I can be your guide
Excerpt from the song Geography, by Frida Selander (2012)
I. Introduction
Background
In a world ‘on the move’ (Sheller & Urry, 2006, p. 207), tourism is part and parcel of many people’s lives. One of the most endemic forms of tourism in the Nordic countries is second-home tourism (Müller, 2007). There are about 1.5-1.8 million second homes in this part of the world, of which roughly 600,000 are located in Sweden (Adamiak et al., 2015; ASTRID, 2015; Hall et al., 2009). The number of second homes per capita in the Nordic countries is among the highest in the world, and a large share of the population there own or use second homes (Adamiak et al., 2015; Müller, 2007). Swedish statistics reveal that about half of the country’s population have access to second homes1 (Statistics Sweden, 2014,
2019b) and use them an average of several months each year (Müller et al., 2010). Although these numbers reveal the sheer scale of the mobility caused by second-home tourism in Sweden and the Nordic countries, this has not garnered a comparable level of attention from the public sector or researchers (Frost, 2004; Müller, 2010, 2011). In fact, although second-home tourism shows a widespread and frequent use of multiple dwellings, governance systems such as civil registries are based on the notion of single permanent dwellings (Czarnecki & Frenkel, 2014; Farstad, 2018; Paris, 2011; Persson, 2011). So, while people’s everyday spatial practices allow shared place attachment and longing for multiple places, administrative systems ground people as belonging in single places (Ellingsen, 2017; Kaltenborn & Clout, 1998). This produces a discrepancy that makes second-home tourists an invisible population in the eyes of administrative systems (Müller, 2007; Müller & Hall, 2003; Pitkänen et al., 2017).
This invisible population has been studied extensively as individuals, with numerous contributions on their motives, activities, emotions, and aspirations as second-home tourists (e.g. Abbasian & Müller, 2019; Aronsson, 2004; Bjerke et al., 2006; Gunnemark, 2016; Halseth, 2004; Hannonen et al., 2015; Kirkegaard Larsen, 2012; Lien & Abram, 2019; Müller, 1999; Müller et al., 2010; Nordin & Marjavaara, 2012). In the present thesis, however, the focus is not on individuals but rather second-home tourists as a collective force with material effects on its surroundings.
The most obvious material effects of second-home tourism are the second homes themselves; anchors fettering mobility in the landscape. However, the physical presence of dwellings is not the only way this invisible population leaves footprints in second-home destinations (Gallent et al., 2016; Pitkänen, 2008).
Literature on second-home tourism is rife with examples of such impacts. Second-home tourism has been found to affect public planning and governance (Hall, 2015; Larsson & Müller, 2019), landscapes and physical environments (Clout, 1971; Pitkänen, 2008), land-use planning (Persson, 2011, 2015), and public services (Osbaldiston et al., 2015). It affects local economies (Bohlin, 1982; Müller, 1999), housing markets (Brunetti & Torricelli, 2016; Di, 2009; Gallent & Tewdwr-Jones, 2000), and access to housing for locals (Hoogendoorn & Marjavaara, 2018), and can cause gentrification (Paris, 2009). Second-home tourism also has effects on civil society (Nordin & Marjavaara, 2012), social capital (Gallent, 2014), and knowledge transfer (Robertsson & Marjavaara, 2015) between visiting and host communities.
There are, however, two deficiencies with these studies that are connected to perspective and context. First, studies on the impacts of second-home tourism seldom investigate them from the local planning authorities’ perspective. Notable exceptions to this are Adamiak et al. (2015), Gomes et al. (2017), Halseth (1998), and Larsson and Müller (2019), but the typical perspective is centred on policies and the effects for locals, local businesses, and the housing supply. Second, inquiries into second-home tourism do not sufficiently address whether, how, and according to what patterns impacts may differ depending on the geographical contexts in which second homes are located.
While there are studies recognising heterogeneous impacts of second homes, such as Adamiak et al. (2015, p. 46), Mowl et al. (2020) and Gallent et al. (2016), such research is usually based on case studies focusing on second-home ‘hotspots’. These are places where the number of tourists is great, the investments sizeable, and the impacts on housing markets evident (e.g. Brida et al., 2009; Gill, 2000; Hoogendoorn et al., 2009; Larsson & Müller, 2019; Marjavaara, 2007). This is not to say that such a focus is unwarranted per se. Such case studies might highlight where second-home tourism has noticeable impacts on local communities. They could also be appropriate when access to national and comprehensive quantitative data is scarce. However, if a large share of second homes are found in more ordinary locations, which is the case in Finland, Norway, and Sweden, for example (Adamiak et al., 2015; Swedish Agency for Growth Policy Analysis, 2012), research conducted on second-home hotspots will result in generalisations that are unrepresentative of the phenomenon at large. These generalisations risk contributing to lopsided research or policy decisions based on incorrect assumptions.
To summarise, second-home tourism in Sweden certainly fulfils Frost’s (2004) description of being a ‘hidden giant’ of tourism. Second-home tourists form an administratively invisible population affecting destinations, in all likelihood in different ways depending on the location. Furthermore, Sweden is exceptional in
both qualitative and quantitative data on the topic. This thesis focuses on housing markets, and the diverse perspectives and strategies of local planning authorities. It contributes to second-home literature through deepened and nuanced knowledge about the impacts and spatial patterns of this type of tourism. Adding to this, the thesis offers perspectives and proposals for stakeholders concerned with second-home tourism, regardless of whether they are planners, policy-makers, or members of the public.
Aim and research questions
The aim of this thesis is to investigate how second-home destinations are impacted by second-home tourism. The impacts investigated are those that are pertinent to local authorities; in this case, Swedish municipalities. Specifically, these impacts involve land-use planning, local economies, infrastructure, public services, and housing markets. Of particular interest is whether there are any spatial variations between different geographical contexts and, if so, what the patterns of these impacts are.
The following research questions are posed:
A. How can spatial variation in the impacts of second-home tourism be analysed systematically on a national level?
B. How do Swedish municipalities perceive and manage impacts of second-home tourism?
C. How is the Swedish housing market impacted by second-home tourism? Through the investigation of these queries, this thesis contributes additional knowledge about the diversity of second-home tourism and its interconnectedness to local contexts, particularly in relation to municipal planning and the housing market.
Outline of the thesis
The thesis consists of six chapters that, step-by-step, present, develop, and discuss the matter at hand2.
The introductory chapter provides a summary presentation of the context, contribution, and aim of the thesis as a roadmap for the reader. The second chapter sets the scene of the research by presenting the invisible population. It
2 See also Figure 3 in Chapter III for an overview of the thesis, the research questions and their
continues by describing the two key concepts used in this thesis, namely landscapes and second homes. This description is necessary for definitional purposes with the intent to understand what is meant by second homes, but also to grasp the conceptual context of the theoretical framework used in this thesis. The chapter also contains a short description of the responsibilities of Swedish municipalities, their role when it comes to housing, and the regional differences within Sweden.
The third chapter describes the methodological considerations and summarises how statistical and interview data has been used in the four papers included in this thesis. The fourth chapter summarises the research gaps and the aims of the thesis, linking to how these are addressed by the results of each research paper. In the fifth chapter, the thesis concludes with a discussion of the results presented in the papers, followed by policy recommendations and suggestions for further research. The final chapter summarises the thesis in Swedish.
Senator Muskie: You say temporary; how permanent is temporary? Mr Collins: That is right, how temporary is permanent?
II. Setting the scene
Permanent and temporary: producing an invisible population
Mobility is an integral part of life and spatial practices in modern societies. People, information, capital, and materials move around the world at faster speeds and in larger quantities than ever before (Adey, 2017; Harvey, 2010; Urry, 2007). Millions of commuters, businesspeople, migrants, refugees, nomads, and tourists move about the world every day, using a multitude of means of transportation (Sheller & Urry, 2006; Urry, 2007). As a widespread form of human mobility, second-home tourism is part of this wider context of ‘mobilized social life’ (Hall, 2005; Urry, 2007, p. 3).As argued by Harvey (1989, p. 6), ‘telecommuting and rapid transport make nonsense of some concept of the city as a tightly walled physical unit or even a coherently organised administrative domain’. However, administrative systems that govern the rights and obligations of the individual in relation to the state are based on immobility, namely the population’s link to permanent dwellings3
(Pitkänen et al., 2017). These systems, such as civil registries, are bulwarks of public administration in Sweden and elsewhere (SOU 2009:75). They are used for many essential state functions, such as taxation and planning, voting, allocation of public funds, and access to public services. In Sweden, the Church originally began keeping a civil registry during the 16th century. The registry’s
purpose soon expanded beyond clerical matters to include the state authorities’ need of recordkeeping for taxation and army conscription (SOU 2009:75). Although much has changed since, the foundational principle of the system is still permanent dwelling. Persson (2011) shows how Swedish legislation and planning practices have evolved over time, struggling to accommodate second-home tourism within existing definitions of permanent and temporary dwellings. The crux of the matter is the definition of permanency, namely ‘how long is permanent?’ (Müller, 2002a, p. 64). There are different definitions, but Swedish legislation stipulates permanent residence as the property where a person has had his or her ‘regular’ place of residence for at least a year (Swedish Tax Agency, 2020c). The goal of this and similar systems is to establish a de jure population through permanent residence in a single location (Ellingsen, 2017; Gallent, 2007; Paris, 2011). Because of the reliance on permanent dwelling, these systems fail to recognise that temporary populations such as second-home tourists are members of the de facto population in a given location during certain periods of time (Hall
3 Illustratively, many languages use words that designate properties and real estate as immovable
objects. For example, Immobilien in German, immobilier in French, fudōsan (不動産) in Japanese,
& Müller, 2018b; Müller & Hall, 2003). In other words, permanent populations are recognised while temporary populations risk being disregarded. This structural mismatch makes second-home tourists an invisible population in second-home destinations (Ellingsen, 2017; Müller & Hall, 2003; Paris, 2011; Pitkänen et al., 2017). Indeed, this invisibility is reflected in the lack of attention to the second-home phenomenon from authorities as well as academia (Back, 2020; Frost, 2004; Müller, 2011).
The invisible population of second-home tourists illustrates a problematic dichotomous separation of permanent and temporary that reflects ‘an artifact of statistical convenience’ rather than real circumstances (Bell & Ward, 2000, p. 104). It boils down to fundamental geographical concepts of space and time. People in a delimited space are recognised as ‘being’, ‘living’, or ‘belonging’ there, by virtue of being present during a fixed period of time defined as ‘permanent’. Conversely, people failing to qualify under the specified time frame are not recognised as ‘being’ in the set space. Hence, they are invisible under certain spatiotemporal delimitations.
Theories on planning, housing and ruralities have to consider the friction caused by such delimitations in order to explain impacts caused by temporary populations (de Souza, 2018; Müller, 2011; Paris, 2009; Woods, 2011). This thesis contributes to such theoretical development by studying the effects of second-home tourism – an invisible population – on public planning and housing markets. In particular, it does so by emphasising the heterogeneous spatiality of these effects.
A case for landscapes
The spatiality of social phenomena can be expressed using different terminologies, but it is arguably apparent that the key spatial concept throughout this thesis is landscape. Although it is certainly a concept that could be discussed at length, it is beyond the scope of the thesis to give an exhaustive overview of this rich field of literature (e.g. Cosgrove, 1984; Duncan & Duncan, 1988; Mitchell, 2002; Olwig, 1996; Widgren, 2015; Wylie, 2007). Rather, the purpose here is to explain the reasoning behind why this thesis employs the term landscape in the theoretical framework of second-home landscapes.
The term landscape is frequently used as a term to describe a view, scenery, or a coherent open area (Antonsson, 2011; Cosgrove, 1984; Hägerstrand, 1991; Tuan, 1990; Wylie, 2007). The word originally hails from Dutch, where landschap illustratively denotes ‘a collection of farms or fenced fields, sometimes a small domain or administrative unit’ (Tuan, 1990, p. 133). Evoking the word landscape can therefore be seen as a way to connect a certain place to its environs more
specifically than using other general spatial terms such as region, area, place, or space (Cosgrove, 1984; Mitchell, 1996; Wylie, 2007). Landscape is frequently used in tourism literature (e.g. Cartier & Lew, 2005; Löfgren, 1999; Urry, 1995), but also in second-home research in describing the geographical contexts of second-home tourism (e.g. Aldskogius, 1969; Halseth, 2004; Müller, 2002c, 2005; Müller et al., 2004; Pitkänen, 2008, 2011; Qviström et al., 2016; Rovira Soto & Anton Clavé, 2017; Stedman, 2006).
Apart from an implied context, landscapes are also subject to processes of change (Antonsson, 2011; Hägerstrand, 1991; Mitchell, 1996, 2008). They are weathered by forces of nature, such as wind, water, flora, fauna, and tectonic movement. Landscapes are also affected by humans, and vice versa. Human labour, culture, capital flows, and investments are affected by landscapes, but also shape and change landscapes for the present and future (Harvey, 2001, 2006, 2017; Mitchell, 1996, 2008, 2013; Olwig, 1996; Smith, 2010). Second-home tourism becomes part of landscapes with varying natural environs and pre-existing histories. Among other things, this history means that agriculture, fishing, and urbanisation have resulted in changing dwelling use from production or primary residence to second homes (Barke, 2008; Bielckus, 1977; Hedenstierna, 1948; Jerling & Nordin, 2007; Lagerqvist, 2011; Langdalen, 1980; Persson, 2011). In other cases, investments in purpose-built second homes have created new housing and supplanted old landscape uses (Paris, 2011). In more recent literature focusing on societal change, this process has been discussed as second-home tourism contributing to a rural-recreational countryside (Halseth, 1998), or a post-productivist countryside (Hoogendoorn, 2010; Müller, 1999) or consumption landscape (Müller et al., 2004).
The sum of this is that it is both logical and practical to use the term landscape in the theoretical framework of this study, rather than other common spatial denominators such as region, area, space, or place. Therefore, the theoretical and spatial framework used in this thesis is called second-home landscapes4.
Figure 1. Second home in Leksand municipality, Sweden. Photo: author.
Defining second homes
The term second home is not the only one available for the subject matter of this thesis. There is a plethora of different terms for second homes, which in itself can be seen as a testament to the phenomenon’s heterogeneity. Some academic and colloquial examples are bachs, cabins, cottages, crofts, dachas, holiday homes, leisure homes, recreation homes, vacation homes, and summer homes (Löfgren, 1999; Marjavaara, 2008; Roca, 2013)5. The term second homes is the most
common term in academic literature.
Looking past semantics, what are second homes and what are they second to? By using the terminology second homes, it can be argued that these homes are placed below or apart from primary (or permanent) residences in an implied hierarchy of dwelling (Gallent, 2007). Using such a hierarchy can impose a ‘sedentarist’ (Ellingsen, 2017; Sheller & Urry, 2006, p. 208) view of second homes as somehow lesser or ‘inessential’ (Müller & Hoogendoorn, 2013, p. 356) as compared to primary residences (Gallent, 2007). Such a definition of dwellings also says very little of second-home tourists’ sense of place or how they value or use these houses
5 As this thesis focuses on Sweden, it may be interesting to note that there are well over a dozen
different Swedish terms for second homes depending on context and dialect, such as bod, fjällstuga,
fritidshus, grosshandlarvilla, land, lantställe, sjöbod, sommarnöje, sportstuga, stuga, torp, and tuskulum.
in comparison to other dwellings (Halfacree, 2011, 2012; Müller, 2002c, 2011; Müller & Marjavaara, 2012; Quinn, 2004). As discussed above, these sedentarist rankings are routinely made in administrative practices (Ellingsen, 2017; Hall & Müller, 2004; Müller & Hall, 2003).
Alternative definitions of second homes serve as criticism of the hierarchical separation of dwellings by, for example, conceptualising them as ‘alternate homes’ (Kaltenborn & Clout, 1998) or parts of ‘multi-house homes’ (Arnesen et al., 2012; Overvåg, 2009). Similarly, Halfacree (2012) has discussed second-home tourism and ‘heterolocal identities’. Gustafsson (2016) has even argued that the many potential uses of second homes beyond leisure make the entire concept problematic.
However, despite its definitional inadequacies, second homes remains the most common term used in academic literature to describe the type of dwelling studied here. This is why it is used in this thesis. A way to reconcile the above criticism with the terminology of primary and second homes used here is to view them as detached from scale, so that the terms do not signify inherent hierarchy but simply dwellings used for different purposes (Hall & Müller, 2004).
Definitions of second homes, such as those used by Coppock (1977b), Hall and Müller (2004), Paris (2011), and Gallent et al. (2016), usually consist of three parts: dwelling, dwelling use, and dwelling location:
i) Dwelling. Second-home tourism can take place in many different kinds of dwellings, which means that second homes are not a certain class or type of building (Coppock, 1977b; Paris, 2014). Detached houses, flats, or even static recreational vehicles can be used for this purpose (Hall & Müller, 2018a; Paris, 2011; Shucksmith, 1983). Most literature delimits the definition of second-home tourism to such dwellings, but there is research that also includes under the second-home umbrella, for example, mobile recreational vehicles or peer-to-peer accommodation networks (e.g. Caldicott et al., 2018; Casado-Diaz et al., 2020). Due to the research aims and availability of data, this thesis focuses solely on detached houses used as second homes. ii) Dwelling use. Second homes are used temporarily for recreational
purposes (i.e. tourism) by owners, guests, or renters (Paris, 2011, 2014). When considered together with the argument above that second homes are not a particular class of building, it means that dwellings are made into a second home through their use. Therefore, whether a building or recreational vehicle is a second home rests in the hands of the user (Müller, 2002b).
iii) Dwelling location. Second homes have to be separate from where their owners or users live, meaning that there has to be some geographical distance between second homes and primary residences (Gallent et al., 2016). The precise distance is not specified in the second-home literature, but as most definitions of tourism argue that tourists leave their everyday environment (Hall, 2005; Hall et al., 2009; Hoogendoorn & Hammett, 2020), the same line of reasoning is employed here. Hence, an additional house next to the owner’s primary residence cannot be defined as a second home, whereas a house outside the owner’s everyday environment can be. This separation rests on the premise mentioned previously, namely that primary residences and second homes are dwellings that are used differently: primary residences for daily life, and second homes for recreation (Gallent, 2007).
From i), ii), and iii), it follows that when the term second homes is used in this thesis, it denotes – unless stated otherwise – detached houses used as second homes. Second-home tourism consequently signifies the act of using second homes as per this definition. The exclusion of flats, caravans, and other dwellings as potential second homes should not be interpreted as speaking in favour of a narrow definition of second homes. Rather, this restriction rests on the aims of the thesis and the availability of data. It is also typical of second-home literature in general, particularly in a Nordic setting (Adamiak et al., 2015; Hall & Müller, 2004; Marjavaara et al., 2019; Müller, 2011).
Swedish municipalities: planning, housing, and disparities
As the research in this thesis has been conducted exclusively in Sweden, it is advisable to provide some elementary information to readers unfamiliar with the context. Although much could be said to offer more information on the many themes this thesis connects to, such as policies for regional development, urbanisation, taxation, and housing legislation, the following section will focus on the three most important topics that anyone unfamiliar with the Swedish context needs to know when reading this thesis. These topics are the Swedish the rural-urban divide, municipalities’ responsibilities for planning, and their relation to the housing market.Administratively, Sweden is divided into regions and municipalities (see Figure 2). Several reforms over the past 70 years have reduced the number of municipalities from 2,281 in 1952 to today’s 290 (Statistics Sweden, 2020c; Öberg & Springfeldt, 1991). These reforms are closely related to economic restructuring, urbanisation, and expanded responsibilities for planning and
public services (Antonsson & Jansson, 2011; Hedlund, 2017; Håkansson, 2000; Lundholm, 2007; Öberg & Springfeldt, 1991). This is the context in which second-home tourism has been situated as being part of changing rural production landscapes into post-productivist landscapes of amenity use (Hoogendoorn, 2010; Müller, 1999; Müller et al., 2004).
While Swedish municipalities vary in terms of size and population, their responsibilities are the same (Syssner, 2018). For example, Arjeplog municipality is 14,494 km2, roughly the size of Montenegro, but with a population of about
2,700 or only 0.2 inhabitants per km2 (Statistics Sweden, 2020c). On the other
end of the scale is Stockholm municipality, with 972,000 registered inhabitants in a space of 215 km2, meaning 5,139 inhabitants per km2 (Statistics Sweden,
2020c). The five most populous municipalities – Stockholm, Gothenburg, Malmö, Uppsala, and Linköping – together have about 2.3 million registered inhabitants, which is nearly a quarter of Sweden’s population. This is also equal to the combined population of the 188 least populated municipalities (Statistics Sweden, 2020c). These demographics and, by extension, economic disparities between rural and urban areas have widened in recent decades (Amcoff, 2020; Håkansson, 2000; Syssner, 2018), with sparsely populated municipalities without large tourism sectors faring the worst (Hedlund, 2017).
Municipal responsibilities are extensive, including land-use planning, infrastructure, water and sewerage, education, elderly care, social services, housing, culture, and sports facilities (Boverket, 2016, 2018a, 2018b; SKR, 2017). Most municipal services and responsibilities are financed by municipal income tax, but also through the redistribution of state funds and fees for services rendered. A further source of income is the municipal property fee, introduced in 2008 in lieu of the abolished property tax (Swedish Tax Agency, 2020b). About 85% of the total municipal income comes from income taxes and state funds (SKR, 2019). All these sources of income are based on the civil registry and its link to primary residences. This implies that there is a constant systemic pressure to maximise the number of registered inhabitants, prompting virtually every municipality to make population growth their highest priority (Niedomysl, 2004; Syssner, 2016, 2018, 2020). Moreover, it is important to note that several areas of municipal planning and services are related to second-home tourism, while the largest sources of income are connected to permanent residence. There is therefore a risk of a mismatch between income and actual expenditures (Hall, 2015; Langdalen, 1980; Paris, 2011).
A wide range of responsibilities and planning tools connect Swedish municipalities and the housing market: overall planning for housing provision, land-use planning, construction, ownership, and renting (Boverket, 2018a, 2018b). In general terms, Swedish housing can be divided into four tenure types: private ownership (semi-detached or detached single family houses), housing associations (semi-detached houses or flats), privately owned housing companies (flats), and public housing companies owned by municipalities (flats) (Borg, 2018; Christophers, 2013). Almost all municipalities own and operate one or several public housing companies (Grander, 2018). The private ownership and housing association dwellings are sold on the open market. It is particularly important to note that, unlike Denmark (Hjalager et al., 2011) and Norway (Rye & Berg, 2011) for example, legal obstacles to any of these tenure types being used as second homes are virtually non-existent. This makes second homes a ubiquitous part of Swedish housing (Marjavaara & Müller, 2007; Müller & Marjavaara, 2012; Nyström, 1989). As such, second homes are part of Swedish municipalities’ planning and housing schemes by virtue of their pervasiveness. This relationship has proven to be the cause of considerable friction, for example when there is a mismatch between planning statutes and actual dwelling use (Persson, 2011) or when public services are planned based on the demand of permanent residents rather than that of the actual population (Larsson & Müller, 2019).
To summarise, Swedish municipalities have been given increased responsibilities in relation to their residents for planning, public services, and housing over the past 70 years. At the same time, urbanisation and economic restructuring have
deepened disparities, particularly between urban and sparsely populated areas. This makes second-home tourism an area of interest for Swedish municipalities; partly because of its impact on public services, housing markets, and planning efforts, and partly because of its use as a vehicle for economic development and population growth. Incidentally, this theme appeared several times in the interviews conducted for the thesis (Back, 2020, forthcoming).
Whoever has lived in a cottage for a weekend or a month or a summer can look at a picture of it only with longing, must feel the slap of water on canoe and the tug on the fishing line, and even smell the burning birch log. Such a house is not simply a house. It is the concrete emblem of a whole class of experiences, experiences that are, at their best, among the most pleasurable and recreative in life.
III. Research design
Methodological considerations
This thesis espouses a pragmatic view on methodology, whereby the research questions are central to the choice of methods (Tashakkori & Teddlie, 1998). With this view of methodology, reformulating the research questions from a methodological perspective offered a way to specify what objectives the chosen research methods had to accommodate (see also Figure 3 below):
a. Construct a theoretical framework capable of systematic analysis of spatial differences on impacts of second-home tourism. This framework is needed to perform (b) and (c) below.
b. Gather information on how municipalities perceive and manage impacts of second-home tourism on public planning and housing markets. Analyse geographical differences using the theoretical framework in (a). c. Describe and analyse how housing markets are impacted by
second-home tourism, focusing on the heterogeneous geography of these impacts using the framework in (a).
As can be seen in Figure 3 below, the methodological objectives (a, b and c) were derived from the research questions (A, B and C) and corresponded to the research papers (I, II, III, and IV). Neither the research questions nor the methodological objectives automatically entailed a particular choice of methods. Rather, different methods were considered, rejected, and adopted during the research process (Dunning et al., 2007).
To begin with (a), the theoretical framework had to be systematic and national in scope. In addition, it had to be possible to map in order to facilitate spatial analysis. A theoretical framework based on qualitative data would not have been sufficiently reliable. It would have been particularly unruly to reliably synthesise and transfer findings to a national spatial analysis. On the other hand, using statistical data to create the theoretical framework meant more reliable and generalisable results. There were therefore clear advantages to using statistical data for Paper I.
As for (b), the type of information required meant accessing data directly from municipalities. Such data could be extracted in many different ways, such as through written documentation (e.g. Persson, 2011), surveys (e.g. Adamiak et al., 2015), and interviews. Interviews with municipal civil servants were chosen as
the primary way to get specific and rich data on a broad range of topics in Papers II and III. This was complemented with correspondence via e-mail. Documents such as surveys, municipal plans, and newspaper articles were not used. This was partly due to time constraints, and partly because an overview by Frykholm (2017) showed that the absolute majority of municipalities lacked specific plans for second-home tourism. In order to compensate for this limitation and ensure rigour, all interviewees were asked whether their municipalities had specific plans or policies for second-home tourism (Morse et al., 2002). The overwhelming majority stated that their municipalities did not have such documents, but rather general plans for tourism or the hospitality industry. Lastly, (c) could have been investigated through case studies, interviews, or other quantitative data from municipalities or the real estate sector. However, the data used was unique in that it was national and georeferenced, and covers all sales of detached houses made on the open market. Viewed from this perspective, alternatives would ultimately have lacked the novelty and explanatory power of the statistical data chosen for Paper IV.
Further methodological considerations connected to the sources of data and their combination in this thesis are discussed below.
Quantifying second homes
The dataset used in Paper I encompassed all detached houses in Sweden in 1997 and 2012, totalling some 2.4 million primary residences and 660,000 second homes. The data was retrieved from ASTRID at Umeå University, a database consisting of georeferenced public data (ASTRID, 2015). Variables that were particularly pertinent for the analysis in this paper were the exact location of each house, its construction year and assessed property value, and the exact location of its owner’s primary residence (see appendix or Back & Marjavaara, 2017). Paper IV relied on a georeferenced dataset covering all purchases of detached houses in Sweden 1999-2017, retrieved from Statistics Sweden (2019a). The dataset covered about 1.4 million cases. The variables used in this paper concerned the location of each purchased house, whether it was used as a primary residence or second home, and its price (see appendix or Back et al., forthcoming).
Both datasets had their shortcomings. In the case of Paper I, one of the variables used was the location of owners’ primary residences in Sweden. This meant that second-home owners living outside Sweden were excluded from the study. The proportion of such second-home owners is greatest in certain parts of southern Sweden (Statistics Sweden, 2020b), which means that the results in these areas might have been affected by this limitation in the data. A minor issue involved instances of missing or faulty information in the dataset, meaning that cases were excluded from the study because they lacked information for all six variables used. These cases were spread geographically, however, which means that their exclusion did not skew the results.
A more crucial limitation of the statistical datasets concerns their coherency in defining second homes (Baggio & Klobas, 2011). The way to define second homes in public records changed in 2014, which means that this occurred after the period studied in Paper I (1997-2012) and during the time investigated in Paper IV (1999-2017). Since both datasets built on public records, this change of definition directly related to the data. However, the definition in both cases rested on Swedish Tax Agency records from the civil registry, ensuring methodological consistency and reliability (Baggio & Klobas, 2011). This meant that both datasets defined second homes as detached houses without registered inhabitants in the civil registry. While this notion is perfectly in line with the view of second homes as products of dwelling use, it hinges on the assumption that houses labelled as second homes in public records are actually used as such.
However, these deficiencies were by far outweighed by the obvious benefits of the data. There are no better sources for national statistics than public records such
as those used here. Furthermore, there are no indications of systemic data errors (such as incentives for tax evasion). Bearing this in mind, the overall reliability of both datasets must be considered high (see also Boschma et al., 2008; Brouder & Eriksson, 2013; Byström & Müller, 2014; Marjavaara & Müller, 2007; Müller, 2006).
Interviewing municipalities
Selecting municipalities
Papers II and III were based on interviews and e-mail correspondence with respondents representing 20 Swedish municipalities6. The municipalities were
selected using statistical data from the georeferenced database ASTRID (2015) in conjunction with the second-home landscapes from Paper I (Back & Marjavaara, 2017). Out of the 20 municipalities, 16 were selected to represent the four second-home landscapes of Paper I (four municipalities per landscape). These 16 municipalities were selected as they had the largest proportions (99 to 45%) of second homes in one of the four second-home landscapes (ASTRID, 2015; Back & Marjavaara, 2017). As such, they were extreme cases in the data, meaning that the groups of municipalities were as dissimilar as possible. An additional four municipalities with a rather even distribution of second homes in all second-home landscapes were included for comparison. The municipalities in this group had about 20 to 30% of their second homes in each second-home landscape. It should be noted that the use of extreme cases does not entail that the second-home landscapes are a static model of categorical differences. On the contrary, the basis for the model and the results of the interviews underline the dynamism of second-home landscapes. Rather, the use of extreme cases was connected to the fact that municipalities were the primary source of data, while the second-home landscapes of Paper I did not follow municipal borders. Selecting extreme municipalities was a way to ensure that the respondents addressed views concerning the second-home landscape in question.
Alternative methods for selecting municipalities, such as randomised selection or selection based on previous knowledge, were rejected based on the aim of the thesis. While randomised selection could have made it possible to test the validity of the second-home landscapes, the small number of cases would have meant a risk of the selection being a collection of more or less similar municipalities. This would have made the aim to research second-home heterogeneity essentially moot. Similarly, a selection based on previous research could very well have provided good data on interesting and preferably contrasting cases. On the other
hand, second-home research has traditionally favoured well-known tourism hotspots, which means that such a selection method would have risked being biased. By extension, it would have obstructed the intention of the thesis to move beyond case studies in obvious second-home locations.
Selecting interviewees and analysing interview data
Several departments in a Swedish municipal organisation share responsibilities connected to second-home tourism. This means that there is no pre-determined role in a municipality with the most expertise on the subject. For this reason, the initial contacts with the municipalities were made through general channels so that the municipalities themselves could choose the most informed respondents. A total of 24 respondents were interviewed via telephone and an additional 13 were contacted through e-mail. There was at least one telephone interview per municipality. Out of the 37 respondents, 24 were women and 13 men. Most respondents worked with land-use/general planning (18 respondents) or social services (12). The remaining held positions in infrastructure and sewerage, tourism, or rural development (7). Altogether 22 respondents held managerial positions in their respective departments. The number of respondents per municipality was admittedly small, and could be seen as constituting a risk for bias (Rubin & Rubin, 2011). On the other hand, the data-driven selection method, the number of municipalities studied, and the geographical spread of municipalities in each group served to compensate this deficiency.
The interviews were conducted between November 2017 and March 2018. Summaries were made in direct connection to each interview, based on notations and the interview recordings. This was in order to evaluate the interview data and the conversation, and to determine whether additional information was needed (Rubin & Rubin, 2011). When the information was incomplete, further data was collected through interview or e-mail contact with the same or other respondents. As many of the municipal administrations were small, the first interviewee could often recommend whom to contact for further information (Stratford & Bradshaw, 2016).
A scheme of broad thematic topics was used in the interviews. These topics were based on previous literature on impacts of second-home tourism, with an emphasis on public planning and housing. General topics were discussed at the beginning of each interview in order to warm up and create a good atmosphere for the continued conversation, while more controversial questions were handled at the end (Dunn, 2016). Thanks to the richness of the interview data, the material could be split and used in two separate papers. Data concerned with second-home tourism and planning was treated in Paper II (see Back, 2020), whereas themes
related to second-home tourism and housing were handled in Paper III (see Back, forthcoming).
Although the interviewees were only able to give an overview of the topics discussed, the semi-structured format of the interviews made it possible to pose follow-up questions and make improvised adjustments (Dunn, 2016). It allowed the interviews to bridge several topics according to the respondents’ train of thought, and made it possible to devote more time to topics they felt were particularly important for their municipality (Rubin & Rubin, 2011).
The purpose of the interviews was to obtain the information that the interviewees could provide on the thematic topics. In other words, they were topical interviews whose focus lay on what the interviewees conveyed rather than why they did so (Rubin & Rubin, 2011). Consequently, the analyses in the papers were not based on objective truth but rather on the respondents’ interpretations. This means that when the thesis states that something ‘is’ in a certain way in the municipalities based on the data collected from respondents, it does so based on the respondents’ interpretations of how it ‘is’.
The interview data was analysed based on the information that the interviewees had provided for each interview topic. All responses for each topic were analysed separately for each group of municipalities. The analysis consisted of three parts. First, it synthesised the information each interviewee had conveyed on the topics. This involved conceptual discussions and the respondents’ own summaries of themes, but also numerous concrete examples from each interviewee’s job or municipality that illustrated interview topics. After this, the synthesised information was compiled on a poster summarising the entire body of interview data for every municipality and topic separately, for easy overview and to simplify further analysis. Second, it analysed the information that all interviewees in each group of municipalities had conveyed on each topic in order to find a converging perspective for each second-home landscape. Third, it identified instances in which interviewees diverged on topics within groups of municipalities, or converged between groups of municipalities. Converging information was treated as views common for each second-home landscape.
Combining multiple methods
The study rests on statistical datasets and interview data, but these have not been applied in isolation from each other. Rather, they were used in conjunction to achieve the objectives of the study (Winchester & Rofe, 2016). There were several advantages to this multi-methods approach.
First, the results of Paper I, concerning the second-home landscapes, were used for methodological development, selection of cases, and analysis in Papers II, III, and IV (Bryman, 2016). Second, the results of the interviews in Papers II and III gave further depth and comprehension to the findings in Papers I and IV (Doyle et al., 2009; Dunning et al., 2007). Third, the results in Papers II, III, and IV also spoke back to reinforce and enrich the second-home landscapes that were the result of Paper I. This complementary combination and dialectical relationship between both the methods and the papers meant more comprehensive and credible results (Bryman, 2016; Paolucci, 2011).
To summarise, the different types of data and research results were used in dialogue to achieve confirmation and comprehension in a way that would not have been possible using only one type of data source (Doyle et al., 2009).
One of the most striking features of the development of second homes is the variety of interests involved and the extent to which these interests are actually or potentially in conflict.
IV. Footprints of an invisible population
A. How can spatial variation in the impacts of second-home
tourism be analysed systematically on a national level?
As can be seen in the introductory chapter of this thesis, the research aims are concerned with the impacts second-home tourism has on municipal planning and housing markets, but also – and perhaps most importantly – whether there are variations in these impacts between different geographical contexts. In order to conduct such spatial analyses, a theoretical framework for the geography of second-home heterogeneity is needed. Therefore, the starting point of this thesis is the framework, second-home landscapes, within which the spatial differences have been investigated and analysed in Papers II, III, and IV.
Studying second-home tourism through a geographical lens is certainly nothing new to the field. Without delving too deeply into the impacts of second homes on destinations and local communities, at present it is enough to emphasise that the literature suggests that there are such impacts. There are also numerous studies revealing how the impacts of second-home tourism depend on spatial context. A pioneer of spatial studies on second-home tourism was Ljungdahl (1938), who noted how public transportation affected urbanites’ opportunity to exit Stockholm and visit second homes in certain parts of its archipelago, which in turn impacted the evolution of second-home settlements. This highlighted the importance of distance and accessibility for second-home use. Many later studies have investigated second-home owners in different places and the relationship second-home tourism establishes between destinations and population centres. Lundgren (1974) and Jansson and Müller (2003) showed how urban areas have recreational hinterlands, but also connections to more distant recreational areas. Similar arguments have also been put forward by, for example, Gallent et al. (2016). Put simply, this means that people either own second homes in close proximity to their primary residences, or farther away in particularly attractive, amenity-rich, areas (Müller, 2002a; Müller et al., 2004). Both distance and accessibility affect the frequency of use (Gallent et al., 2016; Müller, 2002c). A number of studies have investigated the relationship in different parts of the world, such as Canada (Svenson, 2004; Wolfe, 1951), France (Clout, 1971), the Netherlands (Dijst, 2005), Spain (Casado-Diaz, 2004), Sweden (Müller & Hall, 2003; Müller & Marjavaara, 2012), and the United Kingdom (Rogers, 1977). Distance and accessibility are not the only factor affecting the distribution, use, and impact of second homes. Another important piece of the puzzle, hinted at above, is how places vary in their recreational qualities and attractiveness for
second-home tourism. An early contribution in this field was Aldskogius’ (1969, p. 5) groundbreaking spatial modelling of ‘vacation house settlements’, focusing on the patterns of growth of second-home tourism and how this relates to the ‘recreational landscape’ and its ‘recreational place utility’. Another contribution in the same vein was Robertson (1977), who investigated the spatial patterns of second-home owners’ investment decisions in Australia. More recent work has used assessed property values as a way to measure places’ differing attractiveness for second-home tourism (Marjavaara & Müller, 2007).
To summarise, most previous research on the uneven geography and impacts of second-home tourism has focused on two factors. Both of these relate to two sides of the same coin, namely where second homes are located. First, there has been an interest in where second homes are in relation to population centres and owners’ primary residences, and how this impacts frequency of use and, in turn, the destination. Second, there have been quite a few studies on why second-home owners choose to invest in certain places, what recreational qualities these places have and, particularly, how the uneven distribution of second-home demand affects housing markets. These factors can approximately be translated into a question of distance and attractiveness, succinctly summarised by Müller et al. (2004).
Müller’s work on ‘relative attractiveness’ (2006, p. 336) relates the prevalence of second-home tourism to population distribution and geographical distance. As such, it takes the first steps towards a conceptual bridge between distance and attractiveness. The same is investigated for a Swedish case in Müller (2005) and hinted at for Finland in Adamiak et al. (2015, p. 46). Apart from these studies there have been few, if any, that have conducted systematic spatial analyses that unify distance and attractiveness into a theoretical framework; particularly for analysing how the differing impacts of second-home tourism might be distributed spatially, especially on a national level. This research gap is addressed in Paper I.
Paper I. Mapping an invisible population: the uneven geography of second-home tourism
This paper targets the research gap identified above by analysing the heterogeneous geography of second homes. It does so by mapping the composition of the Swedish second-home stock in accordance with the theoretical typology proposed by Müller et al. (2004). The point of departure is the statement in Müller et al. (2004, p. 16) that a ‘major factor influencing the impact on rural change that has not been sufficiently discussed in previous studies is the composition of the second home stock’. The challenge posed by Müller et al. (2004, p. 16) consists of two parts. One is the ‘composition of the second home stock’, while the other is the impacts of said second-home stock ‘on rural change’.
The underlying assumption for the paper is that if the composition of the second-home stock is heterogeneous, impacts of the invisible population on second-second-home destinations should also be heterogeneous. The paper maps the geography of these differences.
Figure 4. Second-home typology and quotes from Müller et al. (2004, p. 16). Adapted with permission.
The typology in Müller et al. (2004) divides second homes into four discrete types: 1) converted weekend second homes, 2) converted vacation second homes, 3) purpose-built weekend second homes, and 4) purpose-built vacation second homes (see Figure 4). The division is made using two axes: distance between primary residence and second home (weekend – vacation), and the attractiveness of the area where the second home is located (converted – purpose-built). Short distances imply weekend homes, while longer distances will mean vacation homes. Lower attractiveness result in converted homes, while higher attractiveness mean purpose-built homes. The typology also connects the different types to certain geographical contexts, such as distance to urban areas, whether the place is attractive for second-home tourism, the natural environs, etc. (Müller et al., 2004).
Using georeferenced data on primary residences and second homes from 1997 and 2012 (ASTRID, 2015), the paper transfers the typology to second-home landscapes (see Figure 5). Each second-home landscape is dominated by one of the four types of second homes from Müller et al. (2004). The variables used in this mapping are the construction year of second homes, the distance between second homes and owner’s primary residences, change in number of second homes, change in assessed property values, second homes’ share of total dwellings, and second-home density. Although the terminology of second-home landscapes been used previously, its precise definition has varied from discursive conceptualisations of second-home spaces (Pitkänen, 2008; Vepsäläinen & Pitkänen, 2010) to material contexts or preconditions for second-home tourism (Marjavaara & Müller, 2007; Müller et al., 2004). This paper adheres to the latter use of the term.
It should be noted that the second-home landscapes are not to be interpreted as unitary or static units, whereby all second homes in a certain type of physical geography are the same. As can be seen in Figure 4 above, both axes in the typology are continua, meaning that the second-home landscapes contain all types of second homes but to varying degrees (Müller et al., 2004). Indeed, the basis of the theory and variables used in the paper relate to preconditions for second-home tourism that are dynamic and changeable, due to mobility, migration, labour markets, capital flows, and economic development. This means that the different second-home landscapes should be read as something constantly in flux. This dynamism is further emphasised in Papers II, III, and IV. As theorised by Müller et al. (2004), the mapped second-home landscapes (Figure 5) cover rather distinct geographical contexts7. A north-south divide
separates the converted weekend and converted vacation second-home landscapes. The paper traces this difference primarily to population density, migration, and economic restructuring. The northern two-thirds of the country, where the converted vacation landscape is prominent, has some 15% of Sweden’s population. Purpose-built weekend and purpose-built vacation landscapes are predominately close to cities and amenity-rich destinations, such as mountain, lake and coastal areas. These attractive locations of re-resourced rural lands (Overvåg, 2009, 2010) are accessible from urban areas, or are sufficiently amenity-rich to merit longer travel (Jansson & Müller, 2003; Müller et al., 2004).
7 Not included in the paper are the absolute number of second homes per second-home landscape.
These are listed below, from the years 2012 and (1997), rounded to even hundreds. Converted weekend: 255,400 (279,800).
Converted vacation: 63,400 (54,900) Purpose-built weekend: 98,900 (107,600) Purpose-built vacation: 249,900 (221,100)
The second-home landscapes as presented in this paper are an important contribution to second-home literature, planning, and policy-making. As demonstrated above, previous research has identified distance and attractiveness as important drivers when it comes to the impacts of second-home tourism. The second-home landscapes combine both distance and attractiveness in one framework. By conducting a spatial analysis at the national level using quantitative data, the paper offers a systematic analysis of the spatial patterns of the impacts of second-home tourism. The second-home landscapes also provide contexts in which case studies on the impacts of second-home tourism can be situated. This allows for systematic generalisations and comparisons of case studies across regions and countries. Although the geography of the second-home landscapes in Paper I is specific to Sweden, the logic and method of the analysis can be transferred to other countries using similar data.
Although the above might seem like the contribution of second-home landscapes is simply practical, the main point of the paper is indeed theoretical. Based on the assumption that different contexts for second-home tourism also give rise to different impacts, the paper calls for a more nuanced understanding of the phenomenon. It proposes that second homes should be seen as an umbrella concept based on dwelling use. This emphasises second-home tourism as an everyday practice in a multitude of contexts with varying impacts.
B. How do Swedish municipalities perceive and manage
impacts of second-home tourism? What are the spatial
differences?
As was made clear already in the introduction of this thesis, second-home tourism impacts landscapes where it takes place – the tourism destinations. What impact refers to in this context are the effects caused by second-home tourism: physically, economically, socially, and culturally (Pitkänen, 2008). Previous research has investigated and revealed a multitude of different impacts that have to be addressed by planning. Here, planning is defined as an activity undertaken by private or public entities to plan ahead for the future (Hall, 2015; Paris, 2019). Tourism is a human activity that prompts planning (Gunn & Var, 2002; Hall, 2008; Inskeep, 1988). More specifically, the planning studied in this thesis is primarily the planning undertaken by municipal authorities in relation to a number of different areas of responsibility, connected to previous research. These can be broadly summarised as 1) impacts on land-use planning, 2) impacts on public services, 3) impacts on local economies, 4) impacts on community, and 5) impacts on housing provision and housing markets. Although these five areas of research are interrelated, they are presented separately below to provide a more accessible overview.
