What makes people stay longer in the densifying city? Exploring the neighbourhood environment and social ties

Abstract

Residential mobility, the movement of households within urban areas, is an important issue for urban planning and development. However, little is known on how residential mobility intentions are shaped in densifying cities. Drawing on quantitative and qualitative survey material from the densifying inner city of Oslo, Norway, we investigate potential drivers of intentions to live longer in a neighbourhood of a densifying inner city. Findings show that dwelling ownership, dwelling size, perceived neighbourhood safety, and socializing with friends and neighbours locally are all associated with intentions to live longer in a neighbourhood. Residents of older compact neighbourhoods are found to participate more frequently in activities locally and seem to have stronger local social ties than residents of newly densified neighbourhoods. Moreover, residents’ insights suggest that newly densified neighbourhoods are often inadequately designed or developed in terms of physical attributes, and this may contribute to possible moving intentions. These findings shed further light on the challenging path towards liveable urban densification.

Keywords:

• compact city
• neighbourhood quality
• intentions to move
• urban liveability
• social relationships
• social cohesion

Introduction

Residential mobility, the movement of households within urban areas, is an important topic for both the individual and the city (Clark, Deurloo, & Dieleman, 2006; Coulter, van Ham, & Findlay, 2016; Oishi, 2010). Citizens may change residential location to adapt to changes in personal or household circumstances or to seek a better quality of life. Residential mobility is also important to consider in urban planning and development. The city and its characteristics should offer residential opportunities that cover residents needs especially in the current age of high mobility. What drives residents to stay put in a neighbourhood is particularly important for urban planning and policy as intentions to stay put in a neighbourhood may indicate a sense of belonging and needs satisfaction both in the short- and long-term. This would therefore suggest that urban planning fulfils one of its major goals which is to enable a good quality of life in cities (Mouratidis, 2021b).

Several cities worldwide are becoming denser to accommodate new residents and, at the same time, be less harmful to the environment at least compared to sprawled development (Andersen & Skrede, 2017; Næss, Saglie, & Richardson, 2020). Urban densification involves the construction of new additional dwellings within built up areas resulting in increases in population and building density. Existing neighbourhoods are being redeveloped or new neighbourhoods are being developed replacing industrial zones or brownfield land. These changes entail residential mobility but also a modification of the urban structure and neighbourhood characteristics. Understanding what drives residents’ intentions to live longer in neighbourhoods in densifying cities can provide important input on how to shape new compact areas that better cover residents’ needs. However, hitherto little knowledge exists on this topic.

In this paper, we investigate possible drivers of intentions to live longer in neighbourhoods of a densifying city. Recently, studies such as Mouratidis (2019) as well as Habibi & Zebardast (2022) indicate that ‘compactness’ might afford residential comfort (see Miller, 2008). Concluding his discussion of how the built environment influence subjective well-being, Mouratidis (2018c) argues that there is a need for more research on potentially ‘influential neighbourhood characteristics’ (p. 36). Following Mouratidis’ call, we systematically investigate how residents understand the different social, material, physical and cultural elements or factors conditioning people’s intentions to stay in a neighbourhood in a city under densification. We also aim to provide insights into whether and how local social ties and participating in activities locally – indicating social (Forrest & Kearns, 2001) or community (Blandy, 2008) ‘cohesion’ – may affect residents’ thoughts about leaving or staying in their neighbourhood.

The main research question explored in this study is ‘how do neighbourhood and dwelling characteristics as well as activities and social ties in the neighbourhood relate to intentions to live longer in a neighbourhood in a densifying inner city?’ To provide answers to the research question, we draw on quantitative and qualitative data from several neighbourhood surveys in Oslo, Norway. Oslo is a suitable case for this research as densification and compact urban development are at the core of the city’s growth strategy. First, we conduct a quantitative analysis that explores whether and how participation in social and other activities locally, feelings of neighbourhood safety, and dwelling size and ownership relate to intentions to live longer in a neighbourhood. Following this quantitative analysis, we zoom in on the role of neighbourhood design and residents’ experiences living in a newly built ‘compact’ neighbourhood. This is done with a qualitative analysis of a strategic sample of answers to open-ended questions drawn from the survey distributed in an urban transformation area.

Literature review

Researchers have examined patterns of residential mobility within cities (Booi, Boterman, & Musterd, 2021), personal factors affecting decisions to move or stay (Clark & Lisowski, 2017; Galster & Turner, 2017; Karsten, 2007), and implications of residential mobility (Clark et al., 2006; Winstanley, Thorns, & Perkins, 2002). An additional line of research examines the neighbourhood – in terms of neighbourhood and dwelling characteristics and/or life in the neighbourhood – as a driver of residential mobility and intentions to move or stay (Permentier, van Ham, & Bolt, 2009). Intentions to move or stay are important as they may indicate future residential mobility behaviour (de Groot, Mulder, & Manting, 2011) but also shed light on whether there is a sense of belonging in a particular area and how well an area covers residents’ needs. In this paper, we explore neighbourhood and housing characteristics, activity participation, and local social ties as possible drivers of intentions to live longer in a neighbourhood.

An important concept that can help explain the intentions to move or live longer in the neighbourhood is place attachment (Lewicka, 2011). Being attached to a particular place may indicate staying longer at that place (Permentier et al., 2009), therefore, drivers of neighbourhood attachment are likely to also be drivers of intentions to living longer in a particular neighbourhood (also Wessel & Lunke, 2021). Time living in a neighbourhood increases place attachment (Lewicka, 2005; Mouratidis, 2020) and this is possibly due to local social relations, memories, and a sense of continuity in a place (Lewicka, 2011). Place attachment is linked with local social capital (Lewicka, 2005). Those who are more attached to their area are more likely to have strong local social ties.

Activities and social interaction at a local scale are key for shaping what can be called ‘home’ (Porteous, 1976), bringing a sense of belonging and a sense of staying put at a particular place. Local social ties are conducive to neighbourhood satisfaction (Clark & Lisowski, 2018; Mouratidis & Yiannakou, 2022) which has been linked to staying longer in the neighbourhood (Clark et al., 2017). Social interaction with family, friends, and neighbours who live in the neighbourhood as well as using the facilities of the neighbourhood are associated with intentions to stay in the neighbourhood (Clark et al., 2017; Permentier et al., 2009). The frequency and intensity of local social contacts can indicate social capital in the neighbourhood (Paldam, 2000). Socially and ethnically heterogeneous communities may have less frequent interactions and lower social capital (Putnam, 2007). When neighbours differ in respect to values or ways of life, such diversity has also been linked conflicts between residents (Gans, 1961), Thus, contemporary societies need to find ways to strengthen social ties and social integration in these communities (Putnam, 2007).

The neighbourhood environment may shape social ties and in turn place attachment and intentions to stay in a neighbourhood. Residents of inner-city neighbourhoods may have larger social networks and engage more often in social activities, at least is certain contexts (Mouratidis, 2018a), supporting some of the ideas by Putnam (1995), but, on the other hand, they have considerably weaker social ties at a local scale and lower local social capital (Mazumdar et al., 2018; Mouratidis & Poortinga, 2020). However, little is known on how living in an older compact versus a newly densified neighbourhood may influence local social ties and activity participation (if time living in the neighbourhood is the same in both cases). We could hypothesize that local activities are more intense in pre-existing compact neighbourhoods and thus one may be more likely to find social and other activities to participate in.

The neighbourhood environment may also contribute to intentions to stay in a neighbourhood via the mediating role of neighbourhood safety. The feeling of safety in a neighbourhood is a key element of urban livability as it is strongly linked to neighbourhood satisfaction, neighbourhood happiness, and overall subjective well-being (Mouratidis, 2019; Mouratidis & Yiannakou, 2022). Therefore, we could reasonably expect that the absence of this key element, in other words feeling unsafe in a neighbourhood, may be a driver of residential mobility.

Housing characteristics are also possible drivers of intentions to live longer in a neighbourhood. Home ownership predicts place attachment due to increased control that homeowners have over their future, while renters may not be willing to develop strong roots to a place due to future uncertainty (Lewicka, 2011). Renters are also more likely to move out as they are more mobile, i.e. they may terminate their lease and find a new home. Dwelling size is a key contributor to housing satisfaction (Wang & Wang, 2020) and a strong indicator of staying put in a neighbourhood (Permentier et al., 2009). Conversely, living in a dwelling that is not large enough to cover household needs leads to housing dissatisfaction and can be a key reason for moving out when personal and contextual circumstances allow to do so.

The world is experiencing an era of rapid urbanization, city growth, and in many cases urban densification. Although there is an increasing body of research on socio-spatial phenomena in cities under densification, the role of neighbourhood and housing characteristics, local activities, and local social ties in intentions to move or stay in a neighbourhood of a densifying city has not been much explored. This paper will explore this topic, focusing on a relatively small European capital that has been under intense urban densification, the city of Oslo, Norway. This case provides us with a relatively different spatial, cultural, and societal context than most previous studies on residential mobility, moving intentions, and place attachment (Centner, 2021). Findings can offer input on how to improve the sense of belonging, social cohesion, and livability in such dynamic urban contexts.

Methodology

The case of Oslo

Norway, including Oslo, could be assumed to constitute a favourable macro condition affording the residents a general positive outlook or securing well-being (Centner, 2021, pp. 23-24). Norway is often described as a generous and ‘high-trust’ welfare state (Bergh & Bjørnskov, 2011). It is an affluent country, with low social inequalities, and ranks first globally in the Human Development Index. The country has low levels of crime (Vrålstad, 2018). Here, the life-expectancy is high (United Nations Association of Norway, 2020b). The Norwegian society is characterized by high levels of gender equality, and it provides its citizens generous maternity/parental leave (e.g. United Nations Association of Norway, 2019). The residents have access to clean water, and with Centner (2021, p. 24) ‘[i]mpressive access to nature’, and are seldom negatively impacted by pollution – also true for people in urban areas (e.g. United Nations Association of Norway, 2020c). The country has relatively low levels of both unemployment (United Nations Association of Norway, 2020a) and income inequalities (but see Hansen & Toft, 2021). With Midtbøen & Nadim (2022), Norway also offers its citizens a favourable opportunity structure, illustrated by the ‘high intergenerational [social] mobility among majority natives’ (Midtbøen & Nadim, 2022, p. 100). Consequently, we can expect relatively positive assessments of residential satisfaction or residential well-being by the residents, also in the largest city, namely Oslo (Adriaanse, 2007; Emami & Sadeghlou, 2021; Mouratidis, 2021a, p. 10). At least, we can expect to see high levels of residential well-being (see e.g. European Commission, 2016) for the numerical majority of the city’s residents who are homeowners, particularly those that live in more homogeneous and ‘middle-class’ areas (Andersen & Skrede, 2017; Vassenden, 2014).

In Norway, researchers, planners, and public officials became preoccupied with securing a more ‘sustainable urban development’ following the Brundtland commission’s report Our Common Future in 1987 (Næss, Saglie, & Thorén, 2015, p. 37). In short, the solution was to turn to a compact urban development, or densification as the main strategy, as proposed by the Government in 1992 (Næss et al., 2015, pp. 39, 42, 44). Today’s inner city of Oslo has become relatively dense, even if there are districts that are also dominated by villas with private gardens within the metropolitan area. Policymakers, such as politicians and public planners, argue and demand a more compact city (Andersen & Skrede, 2017). As this policy vision is compatible with the profit-motives of private developers and property owners (Andersen et al., 2020), the ‘visions’ are to a large degree realized. Of course, the willingness of companies to rent office space and of citizens to buy or let apartments in such projects, help explain why ‘the density increase has been substantial’ in Oslo (Næss, 2021, p. 1). To be more specific, between 2000 and 2020, the population density in the larger Oslo area increased by 22.5% or ‘from 31.3 persons per hectare in 2000 to 38.4 in 2020’ (Næss, 2021, p. 5). Moreover, in the inner zone, including also the ‘urban parts’ of Oslo where you find the ‘bar scene’, the opera, the main train station and the central business district, population density has increased even more. Whereas the population of the inner zone was 152 000 in 2000, it had increased to 228 000 in 2020 (Næss, 2021, p. 13).

Research strategies

Between the spring of 2017 and the fall of 2019, we distributed the surveys to current and former residents in four different neighbourhoods in Gamle Oslo, (located in Oslo’s inner zone, see subsection 3.3 below) and with a total of 2353 respondents. For the quantitative analysis, we selected only current residents of the four neighbourhoods and only those who answered the question corresponding to the main dependent variable examined here (i.e. intentions to live in longer in their current neighbourhood). The final sample of the study is N = 1403 residents. To recruit respondents, a main strategy was to put a printed invitation to the online survey in every mailbox within the different neighbourhoods. We have also been present at different arenas (such as the local metro station) and distributed leaflets inviting people to do the survey for instance using their smart phone riding the metro to work in the morning. As some of the neighbourhoods are more ethnically diverse (the others being more ‘homogenous’), we recruited research assistants (some Norwegians and others with a background that corresponded to the nationality of one of the major ethnic groups in the area) to knock on doors, visit mosques and so forth to invite people to participate. The research assistants also had iPads to be able to assist people to fill out the survey when they met. In the more ‘disadvantaged’ neighbourhood, we also put up flyers about the survey in local stores, public offices, kindergartens, the public health centre and so on. Moreover, we established a temporary office in one of the housing cooperatives where people could come and talk to us, as well as getting information about the survey.

Study area neighbourhoods

Data for the quantitative analysis come from survey responses of residents of four neighbourhoods located in the inner-city of Oslo (Figure 1). These neighbourhoods are Grønland, Tøyen-Kampen, Ensjø, and Sørenga. The first two – Grønland and Tøyen-Kampen – are old compact neighbourhoods of Oslo. Ensjø and Sørenga are new neighbourhoods developed in parts of the city that are under densification. Ensjø is a former strictly industrial area that is being redeveloped into a new residential area, while Sørenga is a former container dock by the Oslo fjord that has been redeveloped into a new residential area. Aerial and street photographs of these neighbourhoods are displayed in Figures 2 and 3.

Figure 1. Map showing the location of the four neighbourhoods in Oslo.

Figure 2. Aerial view of the four neighbourhoods.

Figure 3. Urban spaces in the four neighbourhoods. Source: authors’ collection.

Table 1 presents some basic characteristics of the four neighbourhoods. We observe that the old compact areas are denser than the new neighbourhoods. This is because the two new neighbourhoods belong to areas that are still under redevelopment and densification. Ensjø is still only partially residential (with relatively high densities), while a large part contains industrial and logistics land uses. Thus, the population density of the whole Ensjø is relatively low. Sørenga is fully developed, but its surrounding area is still under densification. So, although the density of Sørenga is relatively high, the surrounding area has a much lower population density. The district (Gamle Oslo) that all these four neighbourhoods belong to traditionally accommodates poorer residents, but we still observe a substantial socioeconomic division within the district itself. As Table 1 shows, residents of the two newer neighbourhoods have, on average, higher incomes and higher levels of education than residents of the two older neighbourhoods. Crime rates in Grønland are the highest among the four neighbourhoods. The crime rate in Sørenga is also quite high but this could be because this refers to the crime rate for the whole subdistrict where Sørenga belongs to.

Table 1. Characteristics of the case neighbourhoods.

	Neighbourhood type	Area size (hectares)	Population	Population density (persons/hectare)	Low-income households with children (%)	Low education for ages 30-59 (%)	Crime rate (offenses/per 1000 inhabitants)
Sørenga	New and densifying	14	1957	140	10 %	13 %	247
Ensjø	New and densifying	125	6341	51	9 %	14 %	65
Grønland	Old and compact	39	9522	246	36 %	27 %	308
Tøyen and Kampen	Old and compact	86	17067	198	28 %	21 %	132

Notes: Sørenga’s statistics on crime rate, low-income households with children, and low education for ages 30–59 refer to the subdistrict (delbydel) Bispevika. Statistics for Tøyen and Kampen were calculated based on data for subdistricts Nedre Tøyen, Enerhaugen, and Kampen. Data source for area size: authors’ analysis with geographic information systems (GIS). Data source for population: data from Statistics Norway 2017–2019 and authors’ analysis with GIS (for Sørenga). Data source for offenses: unpublished data from Oslo police. Offenses include all types of reported cases during the period 01 October 2016–31 December 2016. Data source for percentage of low-income households with children and low-education for ages 30–59: Oslo Municipality - living conditions indicators 2020–2021.

Variable descriptions

Descriptive statistics for all variables used in the study are shown in Table 2. To assess the residential mobility intentions, we evaluated residents’ intentions to live longer in their neighbourhood. Survey participants were asked ‘How many years do you expect to live in… (name of current neighbourhood)?’ on a scale: ‘less than 1 year’, ‘1-3 years’, ‘4-10 years’, ‘more than 10 years’, and ‘rest of my life’. The survey also asked whether residents perform a series of activities locally in their neighbourhood. The question was phrased as ‘Do you spend your free time/participate in any of the following activities in… (name of neighbourhood)?’. The answer options were: ‘sports, ‘cultural activities’, ‘local religious and spiritual organizations’, ‘local ethnic organizations’, ‘local civil society organizations’, ‘local political organizations’, ‘meeting friends and neighbours locally for informal activities’, and ‘no, I do not participate in activities in… (name of neighbourhood)’. To assess the feeling of safety in the neighbourhood, participants were asked ‘How often do you feel unsafe in … (name of neighbourhood)?’ on a scale: ‘often’, ‘sometimes’, ‘rarely’, and ‘never’. Time living in the current neighbourhood was assessed with the question ‘How long have you lived in … (name of neighbourhood)?’, measured on a scale: ‘less than 1 year’, ‘1-3 years’, ‘4-10 years’, and ‘more than 10 years’. Participants were asked whether they own their dwelling, and this was coded as a dichotomous variable. To assess the size of participants’ dwelling, the question asked was ‘How large is the dwelling you currently live in?’. Dwelling size was assessed in square meters. The following sociodemographic variables were also obtained via the survey: age (measured in years), gender, born in Norway, parents born in Norway, and household size. In Table 2, it is observed that the number of responses may differ for each variable. This is due to missing data for some of the questions. We chose to keep incomplete responses as some statistical analyses and statistical models in the subsection 4.1 below do not include all variables.

Table 2. Descriptive statistics.

Variables	N	Min/Max	Mean	s.d.
Dependent variable
Intentions to live longer in neighbourhood	1403	1/5	3.28	1.20
Activities in the neighbourhood
Doing sports locally	1215	0/1	0.13	0.34
Participation in cultural activities locally	1215	0/1	0.18	0.39
Participation in local religious and spiritual organizations	1215	0/1	0.03	0.18
Participation in local ethnic organizations	1215	0/1	0.01	0.08
Participation in local civil society organizations	1215	0/1	0.12	0.33
Participation in local political organizations	1215	0/1	0.08	0.27
Meeting friends and neighbours locally for informal activities	1215	0/1	0.43	0.49
No participation in activities locally	1215	0/1	0.40	0.49
Neighbourhood and dwelling variables
Feeling unsafe in neighbourhood	1395	1/4	3.15	0.87
Time living in current neighbourhood	1403	1/4	2.64	0.96
Owns dwelling	1400	0/1	0.87	0.34
Dwelling size (square meters)	1374	25/240	72.30	26.36
Neighbourhoods (resident of…)
Ensjø	1403	0/1	0.21	0.41
Sørenga	1403	0/1	0.12	0.32
Grønland	1403	0/1	0.32	0.47
Tøyen and Kampen	1403	0/1	0.35	0.48
Sociodemographic variables
Age (years)	1403	13/105	41.44	13.67
Female	1403	0/1	0.55	0.50
Not born in Norway	1391	0/1	0.12	0.33
Parents not born in Norway	1367	0/1	0.17	0.37
Household size	1401	1/10	2.17	1.16

Note: N for different variables differs due to missing data.

Qualitative data collection

Qualitative-oriented studies of neighbourhood preferences and local ways of life have demonstrated the complexity or ambivalence of people’s residential histories, local experiences, and practices of neighbouring (Andreotti, Le Galès, & Fuentes, 2013; Cornelissen, 2022; Savage, Bagnall, & Longhurst, 2005). To shed light on such potential ‘complexities,’ we qualitatively analyse the statements provided in the open-ended sections to questions such as ‘Do you have any thoughts about your current neighbourhood?’, ‘What do you like about your current neighbourhood?’, ‘What do you not like about your current neighbourhood?’. However, because of space limitations, we have focused on one neighbourhood. Moreover, to contribute to the literature on neighbourhood transformations (Andersen et al., 2020), and to the studies of the compact city/densification (e.g. Mouratidis, 2018b; Mouratidis, 2019), we further limit our sample to residents and out-movers of Ensjø, a newly redeveloped area of apartment buildings in one of the city’s largest transformation projects.

Results

Quantitative results: drivers of intentions to live longer in the neighbourhood

The first part of the quantitative analysis explores possible contributing factors to intentions to live longer in the neighbourhood. Since the dependent variable is ordinal, we used ordered logistic regression analysis. Table 3 shows the results of this analysis. Three models were developed. The first model includes neighbourhood and dwelling variables together with sociodemographic variables. The second and third models examine the role of participation in activities in the neighbourhood. The second model includes a dichotomous variable ‘No participation in activities locally’ in addition to neighbourhood, dwelling, and sociodemographic variables. The third model includes a series of dichotomous variables on participation in various activities in addition to neighbourhood, dwelling, and sociodemographic variables.

Table 3. Ordered logistic regression models of intentions to live longer in neighbourhood.

Variables	Intentions to live longer in neighbourhood
	Model 1	Model 2	Model 3
	B (95% CI)	B (95% CI)	B (95% CI)
Neighbourhood and dwelling variables
Feeling unsafe in neighbourhood (ref = never)
Often	−1.178*** (-1.673, −0.683)	−1.203*** (-1.708, −0.697)	−1.214*** (-1.721, −0.707)
Sometimes	−0.919*** (-1.222, −0.616)	−0.961*** (-1.284, −0.639)	−0.917*** (-1.242, −0.592)
Rarely	−0.281* (-0.513, −0.049)	−0.336** (-0.589, −0.083)	−0.333* (-0.587, −0.078)
Time living in current neighbourhood (ref = more than 10 years)
Less than 1 year	−0.350 (-0.743, 0.044)	−0.029 (-0.459, 0.400)	0.073 (-0.361, 0.507)
1-3 years	−0.601*** (-0.922, −0.280)	−0.384* (-0.747, −0.020)	−0.236 (-0.604, 0.132)
4-10 years	−0.442** (-0.738, −0.147)	−0.368* (-0.683, −0.053)	−0.303 (-0.620, 0.014)
Owns dwelling	0.765*** (0.436, 1.094)	0.616*** (0.264, 0.967)	0.597*** (0.244, 0.950)
Dwelling size (square meters)	0.008*** (0.003, 0.013)	0.007* (0.001, 0.012)	0.006* (0.001, 0.012)
Activities in the neighbourhood
Doing sports locally			0.072 (-0.261, 0.405)
Participation in cultural activities locally			0.031 (-0.272, 0.334)
Participation in local religious and spiritual organizations			0.235 (-0.413, 0.884)
Participation in local ethnic organizations			−0.113 (-2.303, 2.076)
Participation in local civil society organizations			0.517** (0.150, 0.884)
Participation in local political organizations			−0.022 (-0.440, 0.396)
Meeting friends and neighbours locally for informal activities			0.688*** (0.450, 0.925)
No participation in activities locally		−0.642*** (-0.871, −0.414)
Sociodemographic variables
Age (years)	0.070*** (0.060, 0.081)	0.080*** (0.069, 0.092)	0.085*** (0.073, 0.097)
Female	0.222* (0.018, 0.426)	0.168 (-0.053, 0.389)	0.158 (-0.064, 0.380)
Not born in Norway	−0.189 (-0.641, 0.263)	−0.186 (-0.667, 0.294)	−0.308 (-0.791, 0.174)
Parents not born in Norway	0.072 (-0.325, 0.468)	0.000 (-0.418, 0.419)	0.071 (-0.347, 0.490)
Household size	−0.041 (-0.147, 0.066)	−0.025 (-0.137, 0.087)	−0.027 (-0.140, 0.086)
Pseudo R-square
Nagelkerke R-square	0.349	0.377	0.389

Notes: *p < 0.05, **p < 0.01, ***p < 0.001. 95% confidence intervals are shown in parentheses.

Results in Table 3 indicate that neighbourhood and dwelling attributes are associated with intentions to live longer in the current neighbourhood. Specifically, feeling of safety in the neighbourhood, longer time living in the neighbourhood, dwelling ownership, and dwelling size are all associated with intentions to live longer in the current neighbourhood. The association between time living in the neighbourhood and intentions to live longer in the neighbourhood becomes weaker when activities in the neighbourhood are included in Models 2 and 3, possibly because the longer one stays in a neighbourhood the more likely it is to participate in activities locally. Findings from Models 2 and 3 indicate that participation in activities locally is associated with intentions to live longer in the neighbourhood. Conversely, those who do not participate in activities locally expect to live for a shorter period in their current neighbourhood. The activities that are associated with intentions to live longer in the neighbourhood are: ‘meeting friends and neighbours locally for informal activities’ and ‘participation in local civil society organizations’. Meeting friends and neighbours locally displays the strongest associations, suggesting that local social cohesion is linked to residential mobility intentions. Finally, among sociodemographic variables, it is age that stands out for its association with intentions to live longer in the neighbourhood. Older participants intend to live longer in their current neighbourhood.

The second part of the quantitative analysis compares residents’ intentions to live longer in their neighbourhood for the different neighbourhoods examined in the study. Again, ordered logistic regression analysis was used due to the ordinal nature of the dependent variable. Models include sociodemographic variables and variables on dwelling size, dwelling ownership, years living in the neighbourhood, as the aim is to detect differences in intentions to live longer in a neighbourhood independently of these attributes. Table 4 displays results of the analysis. Findings indicate that residents of the neighbourhood of Tøyen-Kampen intend to live longer in their neighbourhood compared to residents of the other neighbourhoods of the study. Residents of the other neighbourhoods – Ensjø, Sørenga, and Grønland – display similar intentions to stay in their neighbourhood. This was assessed with additional models not presented here to reduce the size of the tables.

Table 4. Ordered logistic regression models comparing intentions to live longer in neighbourhood for different neighbourhoods.

Variables	Intentions to live longer in neighbourhood
	B (95% CI)
Neighbourhood (ref = Tøyen and Kampen)
Ensjø	−0.510*** (-0.795, −0.224)
Sørenga	−0.635*** (-1.019, −0.252)
Grønland	−0.610*** (-0.861, −0.358)
Neighbourhood and dwelling variables
Time living in current neighbourhood (ref = more than 10 years)
Less than 1 year	−0.350 (-0.743, 0.044)
1-3 years	−0.601*** (-0.922, −0.280)
4-10 years	−0.442** (-0.738, −0.147)
Owns dwelling	0.684*** (0.355, 1.014)
Dwelling size (square meters)	0.011*** (0.006, 0.016)
Sociodemographic variables
Age (years)	0.075*** (0.064, 0.086)
Female	0.171 (-0.032, 0.373)
Not born in Norway	−0.198 (-0.334, 0.454)
Parents not born in Norway	0.060 (-0.334, 0.454)
Household size	−0.053 (-0.158, 0.052)
Pseudo R-square
Nagelkerke R-square	0.340

Notes: *p < 0.05, **p < 0.01, ***p < 0.001. 95% confidence intervals are shown in parentheses.

To try to understand why residents of the neighbourhoods of Tøyen and Kampen intend to live longer in their neighbourhood, we compare key factors that may shape these intentions across the different neighbourhoods (Table 5). These key factors are based on findings from the analysis in Table 3. Residents of Tøyen and Kampen are the ones who most participate in local activities and who seem to have the strongest local social ties. Local activity participation and local social cohesion are possible contributing factors to living longer in the neighbourhood according to findings in Table 3. It is also observed that residents of older neighbourhoods – Grønland, Tøyen and Kampen – participate more in local activities and seem to have stronger local social ties than residents of newly developed neighbourhood of Ensjø. Residents of Grønland display substantially lower feelings of safety in their neighbourhood, which is in line with the higher crime rate shown in Table 1, and this may, to some extent, explain why intentions to live longer in their neighbourhood are not as high as in Tøyen and Kampen.

Table 5. Key differences between neighbourhoods.

	Neighbourhood type	Feeling of safety (mean)	No participation in activities locally (%)	Participation in local civil society organizations (%)	Meeting friends and neighbours locally for informal activities (%)
Sørenga	New and densifying	3.65	N/A	N/A	N/A
Ensjø	New and densifying	3.27	54 %	9 %	27 %
Grønland	Old and compact	2.79	40 %	10 %	46 %
Tøyen and Kampen	Old and compact	3.23	32 %	17 %	50 %

Note: N/A = Data not available.

Qualitative results: neighbourhood design and life in the densifying city

While the previous section reported the key patterns from our surveys, we now take a closer look at the survey responses of Ensjø residents, some of whom live in a large-scale and ‘compact’ urban transformation project while others live in older parts of Ensjø. Moreover, in this section, we also examine the answers for those who have left Ensjø, in total 29 respondents. Among the survey responses from Ensjø residents, we randomly selected some of the respondents and look closer at the open-ended questions to see if the respondents’ statements may nuance our findings from the quantitative analysis of the data (above). The qualitative analysis also provides insights into the role of neighbourhood design by elaborating on residents’ assessments of physical attributes of their neighbourhoods. Analysing this material, some key themes emerge.

Our first observation is that some seem to be firmly attached to their place of residence and/or neighbourhood – some seemingly because of ‘deep roots’ locally, which parallels what we observed in the quantitative analysis – thus reporting that they will live here forever. However, others are more hesitant or uncertain. Those that have ‘deep roots’ have lived in the area for many years, and they differ in this respect to those who have moved into the newer apartment buildings near these more established neighbourhoods at Ensjø. For some of the more ‘uncertain’ residents, living in a new and modern apartment does not serve as a ‘neighbourhood anchor’ (Billingham & Kimelberg, 2013, p. 99). That is, these residents do not consider this apartment or Ensjø as their future home, at least in the longer term. So, while ownership might indicate strong place-attachment, other factors also come into play.

Thus, our second observation is that for the same person or family, some factors seem to contribute to place-attachment while other factors appear to erode place-attachment. This means that a person may consider that some certain elements make her neighbourhood a nice place to live, while other specific factors might contribute to push her out.

Looking only at the responses of those that have moved out from Ensjø, the respondents wrote down different reasons for leaving the area: broke up with a partner; did not want to look straight into the wall of an adjacent apartment building/did not like the density; wanting a larger dwelling; wanting to live somewhere that was more ‘green’; wanted their child to attend a different school; wanted a better place to raise children; wanted to live somewhere more quiet; wanted to live closer to family/kin; or, as one man in his sixties explained: he and his partner left because ‘the neighbours made a lot of noise’. He went on to describe Ensjø as a place with no sense of community, ‘because of all the renters’, whereas his new ‘quiet’ suburban neighbourhood afforded such a community. Another male respondent left his new apartment at Ensjø after having lived there for a couple of years. His family moved to an older neighbourhood in the Tøyen-Kampen area, into the same building that some of their friends had also moved into. He explained that his family really disliked Ensjø which he described as being a ‘dead, impersonal area suffering from really bad urban planning as well as from speculative and simple architecture’. Ensjø was not a place for children to grow up. However, the Tøyen-Kampen area offered just that, ‘a nice place’ for their children. The respondent elaborated and explained that whereas Ensjø was not a community, their new neighbourhood was a safe place with ‘close ties’ between the many families, a place characterized by ‘a strong local patriotism’. This supports our statement on the significance of social cohesion in the quantitative sections above.

Statements from another respondent also point to the importance of social relations but indicate the relevance of transformation and scale as well. The respondent is a middle-aged man who had bought a new apartment at Ensjø in the early 2000s and had put down roots with no intention of moving. He praised the block where he lived as a ‘community’ and great place to live with wonderful neighbours’ – people he interacted with weekly. But because of housing constructions close by, the larger area of did not constitute a ‘neighbourhood’ or a ‘community’.

More generally, elements that seem to strengthen place-attachment among the current Ensjø-residents, include feeling safe when being outside in the neighbourhood; having friends nearby; the geographical proximity to amenities such as public transport, stores, and places to drink and eat. Environmental and neighbourhood elements that seem to weaken place-attachment, are: first, noise from traffic and construction of new buildings; second, lack of parking options; third, things invoking danger or feelings of insecurity, such as traffic and unlit places, and the presence of young men/male teenagers congregating; fourth, a lack of facilities for older children, including teenagers; and fifth, high residential instability rates. The design/architecture, density and size of apartment buildings were also material elements residents stressed as contributing negatively to the quality of the neighbourhood. For instance, one male resident living in an apartment building built before the Second World War, that is, in one of the older neighbourhoods, was happy about the establishment of new public and commercial amenities following the large-scale transformations nearby, but he thought many of the new apartment buildings were of ‘poor architectural quality’. Also, many residents said that they would have preferred more open/green spaces close to their own home, instead of so many apartment buildings. This was a point made by some of the residents in both new and older neighbourhoods. For instance, a woman living in one of the older areas, said that ‘within our area, it is a nice place to grow up, but across the street [where larger apartment buildings have been/are being built] we miss places to meet. There is too much traffic, too many high-rises, and a shortage of public greenspace’.

Moreover, while social networks and social interactions clearly contribute to place-attachment for many residents, neither having friends in the neighbourhood or participating in local organizations seem to be a universal factor when analysing the residents’ own statements on how it is to live at Ensjø and whether they want to stay put. For instance, while a woman reports that she wants to live in her ‘new’ neighbourhood forever, she does not participate in any local organizations. However, she does have neighbourhood friends, people whom she meets weekly. Another female respondent had lived for many years in the established neighbourhood, and even though she liked it here and considers her neighbourhood a good place to grow up for her child – the woman feels that she belongs here – the respondent does not socialize much with people in her immediate neighbourhood. Also of relevance, is that some residents in the newer residential areas referred to what they perceived to be high residential instability rates, and that this prevented the establishment of a local community.

Another observation is how residents are dissatisfied with the private companies responsible for constructing the new residential areas as well as with the city as an urban planner. Some of the people that had bought an apartment in the transformation-area at Ensjø, stated that developer and/or the city had not fulfilled their promises as these had been formulated in the plans or marketing prior to the residents purchasing an apartment. One quote can illustrate such frustrations:

The city authorities should take more responsibility by taking care of the citizens’ interests, by making more general plans, by providing places to meet and green areas prior to housing construction. Moreover, new-build areas need more public amenities less these areas become “dead”. We need more arenas for teenagers and children where they can be creative. Do not make the same playground everywhere.

We also see that for some of the families living in apartments, a larger dwelling – often a house with a small garden – is a preferred type of dwelling. As the inner parts of the city do not have a large amount of such housing, many will therefore have to move out, either to the outer city or to the suburbs outside of Oslo to fulfil this ‘need’. Of course, moving out, cannot be reduced to one factor alone (housing), as other factors influence such decisions as well. For instance, one male respondent explains that his family will leave their new apartment. He is relatively dissatisfied with the Ensjø-area but says that they wanted to move primarily because they would like a house and want to live closer to his parents – in an area with lower housing costs compared to the capital.

We see that our respondents’ satisfaction, well-being, comfort, and potential plans for leaving their current neighbourhood are linked to physical factors, such as density and the height of the surrounding buildings, as well as to traffic and quality of design and architecture. While policymakers want more density, it seems that at least some residents even in newer projects, projects that could be said to contribute to densification, are less happy or satisfied with this density as well as its design. Of course, policymakers could change urban planning strategies – for instance compete with the ‘suburban dreamscape’ (Skrede & Andersen, 2021), but this would come into conflict with the ‘environmental sustainability’ policy that translates into compact urban development. Like Skrede & Andersen’s (2022) study of how residents in Oslo react emotionally to their densified neighbourhoods, we also see how emotions are a relevant factor. As Drozdzewski & Webster (2021, p. 5) note in their review of the literature on the neighbourhood, belonging, and we would also add alienation, are often something that is felt – as in a sense of belonging to the neighbourhood or feeling alienated due for instance to demographic or physical changes in the neighbourhood (e.g. Dench, Gavron, & Young, 2006). Moreover, place attachment and a sense of belonging are for some positively influenced by having a social network or social relations locally, for others, community cohesion seems to be less a critical factor influencing their feelings of belonging. As both social relations and emotions can play a role in influencing place attachment and (thus) residential satisfaction, this may pose another challenge for planners as ‘community’ and feelings are difficult to plan for.

Discussion and conclusions

Findings from this study – based on quantitative and qualitative data from surveys in neighbourhoods in the densifying inner city of Oslo, Norway – indicate that neighbourhood and dwelling attributes as well as activity participation at a local level are all related to intentions to live longer in a neighbourhood of a densifying inner city. Perceived neighbourhood safety, dwelling ownership, and dwelling size are found to be associated with intentions to live longer in a neighbourhood. Residents’ input also suggests that newly densified neighbourhoods are often insufficiently designed or developed in terms of physical built environment (e.g. public spaces, parks, greenery, amenities, and building design), and thereby contributing to possible moving intentions. Besides physical factors, activities performed at a local level were found to be related to intentions to stay put in a neighbourhood. Residents who performed activities locally reported intentions to live longer in their neighbourhood. On the contrary, those who did not perform any activities locally reported that they are more likely to move out of their current neighbourhood. The activities more strongly linked to intentions to stay put in a neighbourhood were meeting friends and neighbours locally and to a lesser extent the participation in local civil society organizations. The relationship between social activities and intentions to live longer in a neighbourhood could be bidirectional; participating in social activities may strengthen place attachment and make people stay in their neighbourhood, but, also, those who intend to stay longer in a neighbourhood may be more willing to develop local social connections and participate in activities locally.

Residents of older compact neighbourhoods were found to participate more frequently in activities locally and seem to have stronger local social ties than residents of newly densified neighbourhoods. This may be, at least partially explained, by the more years living in the neighbourhood which has also been found to be a driver of intentions to stay put in the results of the present study. However, even when we control for time living in the current dwelling, it was residents of an old compact neighbourhood (Tøyen-Kampen) who were found to intend to live longer in their neighbourhood.

What could additionally explain this finding is place attachment. Place attachment may influence intentions to live longer in a neighbourhood (Permentier et al., 2009). Although we partially address this by including time living in a neighbourhood – which strongly predicts place attachment (Lewicka, 2005; Mouratidis, 2020) – in our quantitative analysis, there might still be some place attachment inherent to older neighbourhoods independently of time living there. If this is the case, we could expect place attachment to grow in newer neighbourhoods in the future, which might also result in increased neighbourhood satisfaction and intentions to stay longer in these neighbourhoods. We need to note however that place attachment does not only depend on time living in a particular neighbourhood and the physical attributes of the neighbourhood; it also depends on local social conditions and ties (Andersen et al., 2020). Changes in social conditions and social ties can occur any time, contributing to an erosion of place attachment (Watt, 2010) and leading to changes in intentions to stay in a neighbourhood.

Overall, our findings support earlier research indicating the importance of neighbourhood and dwelling physical characteristics as well as neighbourhood social cohesion in needs satisfaction and decisions to stay or move (Clark et al., 2017; Forrest & Kearns, 2001; Lewicka, 2005, 2011; Mouratidis, 2021b; Mouratidis & Yiannakou, 2022; Permentier et al., 2009). However, the findings also contribute to existing knowledge with additional input on the role of neighbourhoods, social ties, and activity participation in densifying cities. Newly (re)developed neighbourhoods often do not fulfil expectations in terms of neighbourhood design and dwelling features. Local social relationships also tend to be weaker and local activity participation tends to be more infrequent. The qualitative analysis indicates that residents find a lack of community in newer neighbourhoods as a contributing ‘push factor’, some also stressing how they relocated to a more established residential area with close ties between neighbours. Therefore, to achieve a more liveable, equitable, and inclusive urban densification, it is critical to pay more attention to neighbourhood and dwelling design in new or redeveloped neighbourhoods aiming at better covering residents’ daily needs and enhancing local social cohesion.

There is still a need for more research on how people relate to ‘space’ and to each other as neighbours, especially if we want to be more certain about the mechanisms or conditions (Small, 2004), affording neighbourhood satisfaction, comfort or happiness, and (thus) neighbourhood stability. Often, social scientists are asked, expected, or indeed want to present policy recommendations based on their findings – as this recent Oslo-based study by Wessel & Lunke (2021). Our main suggestion is for planners and policymakers, but also other scholars, to recall Sampson’s (2012) argument that the specific urban or housing policy change informed by social science can be expected to change the very social world the intervention was meant to improve. In other words, ‘the rules of the game change and [often unintended and sometimes unwanted] social processes spring into action’ (Sampson, 2012, p. 381). For instance, mandating housing or tenure mix to achieve neighbourhood social mix and thus city-wide desegregation (Wessel & Lunke, 2021, p. 146), may in fact lead to middle-class flight from the policy-intervened neighbourhoods and thereby increased segregation. It follows that whereas social scientists may be good at interpreting or analysing the situation as it is, or rather was, there is no guarantee that we can predict the future. In the words of Erickson (2004, p. 194), we lack the ‘God-like’ abilities to see or know everything, especially the future. Having said this, a recent comparative study of a southern and a northern European city, namely Thessaloniki (in Greece) and Oslo, concludes by noting that ‘[i]t is evident that urban environments that promote safety, lack of noise, place identity and place attachment, and local social connections, contribute to liveability at a neighbourhood scale’ (Mouratidis & Yiannakou, 2022, p. 11). Our findings do, in large part, corroborate such a claim, with a nuancing caveat. We need to first interject that Mouratidis & Yiannakou (2022) analysed survey data from closed-ended questions whereas in this paper, we have also interpreted respondents’ own statements provided in open-ended questions. This might partially explain our caveat. That is, while people are happy or satisfied with their dwelling and neighbourhood ‘for now’ (when we did the survey) some might be conditionally happy/satisfied. By this we mean that residents knowing that they can, and probably will, leave the neighbourhood (for instance in about three years) seem to make people ‘ok’ with some factors (e.g. perceived local school quality) that might would have impinged on their happiness/satisfaction if they were ‘stuck in place’ (Sharkey, 2013). To rephrase, neighbourhood satisfaction and especially neighbourhood happiness might in some part depend on the residents’ perceived possibility to move elsewhere if they want to.

The present study has certain limitations that could be addressed in future research. Data collection was performed in Oslo, Norway thus the study’s outcomes may be more relevant to the Nordic context. Future studies could explore drivers of intentions to stay longer in neighbourhoods of densifying cities in other spatial, social, cultural, and economic contexts. Although qualitative data provide some input on the directions of the identified relationships, future research could further investigate this with longitudinal data. Additional neighbourhood (e.g. neighbourhood satisfaction, neighbourhood attachment), dwelling (e.g. housing satisfaction), and sociodemographic (e.g. income, education) variables would have been useful to assess for their role in intentions to stay in a neighbourhood, but these were not included in our study to shorten the length of the surveys and increase the response rates. Future studies could include these additional variables in quantitative models. Neighbourhood design was assessed mostly in the qualitative part of the present study. Future research could also explore physical neighbourhood features with quantitative models. Our qualitative analysis focused on one neighbourhood that has been under densification. Qualitative comparisons between various neighbourhoods (newer and older) would make a useful contribution in future research. Finally, a shortcoming of the present analysis is that there is no comparison of the results with non-dense areas of the city so we cannot conclude whether drivers of intentions to stay longer in a neighbourhood are unique to dense areas or are common across the entire city. Such a comparison could be performed in future studies making a useful contribution to existing knowledge.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by The Research Council of Norway [grant no. 259888].