Neighbourhood Change,Mobility and Incumbent Processes: Exploring Income Developments of In-migrants,Out-migrants and Non-migrants of Neighbourhoods

Abstract

Most studies on neighbourhood change attribute a key role to mobility in social upgrading and downgrading patterns, while incumbent processes—i.e. changes in the socioeconomic status of non-migrants—are often ignored. This paper explores the relationship between mobility and neighbourhood change by examining income developments of in-migrants, out-migrants and non-migrants of neighbourhoods in three Dutch cities. The paper demonstrates that in- and out-migration are not the only processes at work and that changes in the socioeconomic status of non-migrants are of importance too. Within upgrading neighbourhoods, incumbent processes of non-migrants seem to be an important driver of upgrading. Furthermore, although in-migrants have relatively low incomes when moving in, they experience strong income gains in the years after in-migrating. The contribution of out-migrants to upgrading is mixed. In line with previous studies, migration reinforces downgrading processes, although at the same time, incumbent processes of non-migrants impede downgrading.

Introduction

The study of neighbourhood change has a long history, dating at least from the 1920s when it was an important topic for the early-20th-century Chicago School. Ever since, scholars have investigated neighbourhood change. Gentrification in particular has received much attention, in its classic sense defined as a process of replacement of the working classes by higher-income households (Glass, 1964). Neighbourhoods undergoing downward development—i.e. decline in a neighbourhood’s housing stock and/or socioeconomic status—have also been intensively studied. What most studies have in common, is the fundamental role that is attributed to mobility in upgrading and downgrading patterns (Millard-Ball, 2002). Migration is even one of the defining characteristics of gentrification: it is widely agreed that gentrification is driven by the in-migration of higher-income households into working-class neighbourhoods, leading to replacement of working-class households (for example, Glass, 1964; Smith, 1996). Furthermore, downgrading is often related to the in-migration of low-income and the out-migration of high-income households. The basic assumption in many studies is that income changes lead people to resettle themselves to keep their socioeconomic status in line with their neighbourhood’s socioeconomic status (also see Bailey, 2012). Incumbent processes of neighbourhood change are often ignored—i.e. changes in the socioeconomic status of sitting residents of neighbourhoods. Although some authors acknowledge that neighbourhood change is achieved through both mobility and sitting residents (for example, Grigsby et al., 1987; McKinnish et al., 2010; Bailey, 2012), few have actually focused on incumbent processes and we know little about changes in the characteristics of sitting residents in relation to those of in- and out-migrants. This gap is largely due to a lack of data: most studies are unable to present data that distinguish between in-, out- and non-migrants, although some exceptions can be made (for example, van Criekingen and Decroley, 2003; McKinnish et al., 2010; Bailey, 2012). This paper explores the relationship between mobility and neighbourhood change and examines income developments of in-migrants, out-migrants and non-migrants of neighbourhoods. It demonstrates that although in- and out-migration are important in generating neighbourhood change, changes in the socioeconomic status of non-migrants are of importance too.

Understanding socioeconomic neighbourhood change caused by in-, out- and non-migrants is essential for policy-makers in formulating policies concerning future neighbourhood development. First, many deprived neighbourhoods in western Europe are targeted for renewal, since policy-makers believe that in-migration of low-income people causes decline. Through renovation and demolition of social housing and construction of owner-occupied housing, policy-makers aim to attract high-income households and decrease in-migration of low-income households (Uitermark et al., 2007). However, we know little about the actual socioeconomic characteristics of in-, out- and non-migrants. Secondly, high mobility is often seen as a trigger of neighbourhood decline (for example, Andersson and Bråmå, 2004; van Ham and Clark, 2009). This is associated with a lack of social cohesion, anonymity and criminality. High mobility could also be understood as a sign of a well-functioning neighbourhood serving as an ‘escalator’. On the other hand, low mobility may mean that households are at the top of their housing career, but it may also reflect a situation of not being able to move (Musterd and van Kempen, 2007). Knowledge about the relationship between mobility and neighbourhood change provides insight into the functions of neighbourhoods for its residents.

This paper aims to contribute to the literature as follows. Since it is often assumed that neighbourhood change goes hand-in-hand with high mobility, first, this relationship is examined. Secondly, income developments of in-migrants, out-migrants and non-migrants of neighbourhoods are analysed and related to neighbourhood upgrading and downgrading. The following research questions are addressed

What is the relationship between processes of social upgrading and downgrading and residential mobility at the neighbourhood level?

What are the income developments of in-migrating, out-migrating and non-migrating households at the neighbourhood level and how are these developments related to social upgrading and downgrading of neighbourhoods?

This study is part of a larger research project, which explores the relationship between changing positions of neighbourhoods and households in three Dutch cities: Amsterdam, The Hague and Tilburg. Neighbourhood change and mobility are embedded in and influenced by housing market conditions, such as tenure structures and governmental regulations and interventions (Clark and Dieleman, 1996). As the Dutch housing market is highly regulated, this may affect neighbourhood change and mobility patterns. Therefore, the paper discusses the Dutch institutional context and its possible impact on the observed patterns of neighbourhood change and mobility.

Social upgrading and downgrading are measured in terms of the relative growth in mean neighbourhood income: neighbourhoods are defined as upgrading when the growth of the neighbourhood income is above the growth of the average city level. When the growth of the neighbourhood income is below that of the average city level, neighbourhoods are defined as downgrading.

The paper first evaluates literature focusing on the relationship between neighbourhood change, mobility and incumbent processes. This is followed by a discussion of the Dutch institutional context and a methodology section. The results are presented in two parts to address both research questions. Finally, a conclusion section reflects on the results and highlights points for further research.

Theoretical Embedding

Neighbourhood Change and Mobility

The structural analysis of neighbourhood change started with the Chicago School. Particularly well known are the invasion/succession theory of Burgess (1925/1974) and Hoyt’s (1939) filtering theory. In both theories, mobility is the key to neighbourhood change: chain reactions in mobility take place from low-quality neighbourhoods (city-centre) to newer, higher-quality neighbourhoods. Until the 1960s, most studies assumed that the general trend in neighbourhood change was downward. In the 1960s, Glass (1964) observed a revaluation of inner-city neighbourhoods by young households and, ever since, gentrification has attracted widespread attention. Although there is still disagreement over its causes and consequences, within gentrification studies also, the key role attributed to mobility is central: a ‘new’ group moves into inner-city neighbourhoods, thereby replacing ‘old’ groups of (lower-income) residents and changing the socioeconomic and physical environment. Numerous studies argued that these new groups are highly educated, possess high levels of social and cultural capital and are young single/dual-person households (for example, Butler and Robson, 2001; Lees, 2008). Moreover, in-migrants are assumed to have higher incomes, although some argue that initial gentrifiers may have modest incomes, which increase as the gentrification process advances (Clay, 1979; McKinnish et al., 2010). Regarding out-migrants, there is an extensive literature focusing on negative effects such as displacement. These studies criticise gentrification for causing displacement of low-income residents (for example, Atkinson, 2000; Lees, 2008). Although Atkinson (2000) acknowledges that it is difficult to examine which out-migrants are displaced or left voluntarily, he states that displacement is apparent in London. Others challenge displacement by arguing that out-migrants often leave voluntarily, for instance to climb the housing ladder (Freeman, 2005; McKinnish et al., 2010). What is remarkable within gentrification literature, is a lack of data which systematically examine characteristics of in-, out- and non-migrants. Often aggregated changes of neighbourhoods are examined, which makes insight into the processes by which gentrification occurs difficult (Lupton and Power, 2004). Some exceptions can be made, as is shown later.

Downgrading has also been examined extensively. Many downgrading studies, especially in the Netherlands, focus on declining post-war neighbourhoods which are targeted for renewal, as it is often believed that in-migration of low-income residents causes decline. Through selective demolition of social housing and construction of owner-occupied housing, policy-makers aim to attract higher-income households. Many studies evaluate renewal processes and focus on how they changed the neighbourhood’s socioeconomic composition (for example, Murie et al., 2003; Uitermark et al., 2007). Others deal with the mobility of households who are forced to move through renewal and how this affects the neighbourhoods they move into (for example, Bolt et al., 2009). A central element, however, is the focus on in- and (forced) out-migration, while few studies focus on characteristics of non-migrants.

Andersson and Bråmå (2004) observed in Swedish distressed neighbourhoods that in-migrants have higher benefit dependency rates and lower incomes than out-migrants, while non-migrants are in between. Thereby, they argue, selective mobility reproduces distressed neighbourhoods. Bailey and Livingston (2008) also found that in- and out-migration increased spatial segregation in England and Scotland, but observed that its impact is weak. This is in line with Shuttleworth et al. (2012), who found no evidence for increases in segregation through mobility in Northern Ireland. However, both studies did not examine the impacts of socioeconomic changes of non-migrants on segregation. Bailey (2012) observed little association between status changes and the direction of mobility. Furthermore, net effects of selective migration were weak and did not offset effects of status changes of non-migrants and other processes (see later).

Incumbent Neighbourhood Change

Clay (1979) and Galster (1987) were among the first authors paying attention to incumbent processes. They refer to incumbent processes as a physical process, in which homeowners revitalise their dwellings. Of course these investments are related to socioeconomic neighbourhood change, as physical investments are likely to result in longer stays in the neighbourhood. Clay states that incumbent upgrading is the most significant part of neighbourhood revitalisation, as it reflects neighbourhood confidence. Galster argues that the collective upkeep behaviour of homeowners is an important trigger for neighbourhood change. He found that expectations of a neighbourhood’s future and length of tenure are important determinants of upkeep behaviour. Home investments will change if these expectations change, leading to incumbent upgrading or downgrading.

Grigsby et al. (1987) acknowledge that socioeconomic changes of non-migrants may occur, but argue that neighbourhood change is usually associated with mobility. They measure neighbourhood change by looking at aggregate changes in neighbourhood incomes and do not distinguish between changes achieved through mobility and non-mobility. McKinnish et al. (2010) analysed characteristics of in-, out- and non-migrants of gentrifying neighbourhoods and demonstrate that non-migrants have higher incomes than in-migrants, thereby suggesting incumbent upgrading. Van Criekingen and Decroley (2003) demonstrate the importance of incumbent upgrading in Brussels and Montreal; like Galster, they refer to incumbent upgrading as a physical process of housing stock reinvestments. Millard-Ball (2002) acknowledges the existence of incumbent upgrading, but states that “this route is likely to be slow and of relatively minor importance” (p. 833). Freeman (2005) also acknowledges incumbent upgrading of non-migrants, as gentrifying neighbourhoods may offer possibilities for social mobility. However, little is known about non-migrants.

As noted earlier, Andersson and Bråmå (2004) observed that the socioeconomic status of non-migrants is in between that of in- and out-migrants. However, they did not discuss their absolute contribution to neighbourhood change. Bailey (2012) demonstrates that, although selective migration reinforced spatial segregation to a limited extent, net segregation patterns declined. He attributes this to changes in socioeconomic status of non-migrants and processes of ageing and cohort replacement, which offset mobility effects.

Residential Mobility

Although it is demonstrated that income changes are among the main triggers to move (Clark and Dieleman, 1996), it is also argued that mobility is closely related to age and life-course. Changes in life-course are characterised by different sizes and compositions of households, which may create the need for more space (Rossi, 1955; Mulder, 1993; Clark and Dieleman, 1996). As the life-course generally shows strong dynamics in early stages, young people have higher mobility rates than older people (Bailey and Livingston, 2007). Also, young households usually start low in the hierarchy. Consequently, neighbourhoods with accessible dwellings and young populations are likely to show high mobility rates. In later stages, households usually climb the housing ladder to less accessible dwellings. At the final stage, households generally stay a long time, as the chances of obtaining better dwellings in relation to what they can afford are low (Hoogvliet, 1992). It is often assumed that people resettle themselves to keep their socioeconomic status in line with their neighbourhood’s socioeconomic status. Bailey and Livingston (2007), however, challenge this by arguing that a significant part of mobility occurs between neighbourhoods with similar statuses.

Together, these factors indicate that income changes do not necessarily lead to mobility and incumbent change might also be of importance in generating neighbourhood change.

The Dutch Institutional Context

Neighbourhood change and mobility are embedded in and influenced by housing market conditions, such as tenure structures and governmental regulations and interventions (Clark and Dieleman, 1996). The Dutch housing market is highly regulated: opportunities of households are strongly influenced by the government (Musterd and van Kempen, 2007). The share of homeownership is low compared with other countries (55 per cent), while respectively 31 and 13 per cent are socially and privately rented. Urban neighbourhoods have mixed tenure structures, although there are variations. Also, the government developed instruments (subsidies, taxes) leading to a redistribution of incomes. The Netherlands are characterised by a relatively equal society, which is reflected in a comparatively low gini coefficient (0.29)¹ (van den Brakel, 2009). The redistribution of incomes and mixed tenure structures result in relatively small differences in neighbourhood incomes. This may affect upgrading and downgrading, as market mechanisms are reduced.

In general, there are waiting-lists for obtaining social rented dwellings and allocation rules are strict. This affects mobility rates, as households may have a wish to move, but might not be able to. Once social rented dwellings are obtained, households cannot be forced to move after income gains. For households no longer complying with allocation rules, moving is associated with strong increases in housing costs. Consequently, households might stay when their dwelling is no longer suitable (Kullberg, 1997), leading to incumbent upgrading.

During the 1980s and 1990s, housing policies changed to a more market-oriented system. Subsidised construction of social housing was abolished and responsibility for adequate housing decentralised to municipalities. This resulted in decreases of newly constructed social housing. Furthermore, the government’s policy of privatisation and deregulation of the housing system meant that part of the social housing stock was sold. So, the owner-occupied housing stock is increasing, although urban neighbourhoods still have mixed tenure structures. The government intervenes in the owner-occupied sector too, through tax breaks, subsidies and guarantees on mortgages, to promote homeownership. Furthermore, conditions for obtaining mortgages have become less restrictive. At the same time, developments in society triggered homeownership, such as increases in incomes and the number of dual-earner households. Increases in homeownership might impede mobility, as homeowners are less likely to move than renters (van Ham and Clark, 2009).

Finally, in targeted deprived neighbourhoods, mobility and neighbourhood change are triggered by renewal processes. The government aims to improve the quality of deprived neighbourhoods through differentiating the housing stock (Murie et al., 2003). Through demolition of social housing, (low-income) households are forced to move. Simultaneously, newly constructed owner-occupied dwellings attract higher-income households. So, renewal policies affect both targeted neighbourhoods and those receiving forced migrants.

Study Area

This paper focuses on Amsterdam, The Hague and Tilburg, which are characterised by different housing market contexts. Amsterdam had 790,044 inhabitants in 2012 and is characterised by a commercial and service industry. The housing stock consists of a large social housing stock (47 per cent) and small owner-occupied stock (24 per cent). The market is characterised by high demand, which is reflected in long waiting-lists for social housing and expensive owner-occupied and private rented dwellings. The Hague had 502,802 inhabitants in 2012 and is the governmental centre of the Netherlands. The share of homeownership is 43 per cent, while respectively 33 and 22 per cent are socially and privately rented. Tilburg is smaller with 207,579 inhabitants in 2012 and characterised by its industrial past. Tilburg lacks a major historical centre, as it emerged from a conglomeration of smaller cities. The owner-occupied stock is comparatively large (54 per cent), while respectively 34 and 11 per cent are socially and privately rented.

Data and Methodology

Neighbourhood change is often measured through examining income data (for example, Grigsby et al., 1987; Bourne, 1993). This paper uses ‘standardised net income per household’ as an indicator, which corrects for differences in size and composition of households of a neighbourhood. Incomes are converted to that of a household of one person, which makes comparison between neighbourhoods possible. Standardised net income per household is calculated through dividing the total net household income (from work, benefits and pensions) by an equivalence factor.² The formula for the equivalence factor, E, is: E = {A+(0.8*C)} 0.5, where A is the number of adults and C the number of children in a household (Statistics Netherlands, 2011).

Data come from the Social Statistical Database (SSD) of Statistics Netherlands, which includes data on the entire registered population for the period 1999 to 2008. The dataset contains data on residential trajectories, individual and household characteristics and income. This paper uses data based on all individuals in Amsterdam, The Hague and Tilburg, except for institutional households and people who were born or died between 1999 and 2008. Data at the neighbourhood level are obtained through aggregating individual data. Income data were checked for extreme outliers before aggregating.³ Furthermore, this paper refers to neighbourhood real estate values, which come from Kadaster, a governmental organisation registering real estate transactions. To ensure reliability, a minimum of 30 transactions were required for each neighbourhood and two two-year periods are compared: 1999/2000 and 2008/2009. This paper refers to these periods as 1999 and 2008.

This paper measures neighbourhood change relative to city-wide change. Between 1999 and 2008, city-wide standardised net household incomes increased significantly: from 14,302 to 19,294 euro in Amsterdam, from 14,888 to 19,566 euro in The Hague and from 14,530 to 18,651 euro in Tilburg. Simultaneously, income levels of all neighbourhoods increased. If neighbourhood change was measured absolutely, nearly every neighbourhood would be classified as upgrading. In some neighbourhoods, incomes increased significantly faster than in others, which shows the hierarchical position of neighbourhoods within a city. This paper is interested in changes in this hierarchy and therefore measures neighbourhood change in comparison with other neighbourhoods. Of course, measuring neighbourhood change relative to city-wide change has shortcomings. It can be argued that absolute improvements make differences to people’s lives, instead of the position of neighbourhoods relative to others (Lupton and Power, 2004). Also, it might be counter-intuitive to define neighbourhoods as downgrading when they showed a rise in income.

Social upgrading and downgrading are calculated as follows: social grading = (mean neighbourhood income in year x/mean neighbourhood income in year y)/(mean city income in year x/mean city income in year y). Neighbourhoods are defined as upgrading when the growth of the neighbourhood income is more than half a standard deviation above the growth of the mean city income between 1999 and 2008. Neighbourhoods are defined as downgrading when the neighbourhood income development is more than half a standard deviation below the city development. When the neighbourhood income development is between half a standard deviation below or above the city level, neighbourhoods are considered as ‘keeping in pace’ with average city development.

First, the relationship between social grading and mobility is examined. Residential mobility reflects the percentage of individuals who moved to another neighbourhood between 1999 and 2008. Neighbourhoods constructed after 1999 are excluded. Neighbourhoods are classified as ‘high mobility’ when the percentage of individuals who moved is half a standard deviation above the city-wide percentage of individuals who moved. Neighbourhoods are classified as ‘low mobility’ when the percentage of moved individuals is half a standard deviation below the city-wide percentage. When the percentage of moved individuals is half a standard deviation above and below the city-wide level, neighbourhoods are classified as ‘average mobility’.

Secondly, income developments of in-migrants, out-migrants and non-migrants are examined, and related to the neighbourhood income development. For each year, incomes of all in-migrants, out-migrants and non-migrants are calculated and aggregated to the neighbourhood level. Furthermore, income developments of in-migrants after in-migration and of out-migrants before out-migration have been studied, in order to obtain insight in their contribution to neighbourhood change.

This study follows the national classification of neighbourhoods. Neighbourhood boundaries are determined by municipalities and are generally socially or physically homogeneous areas, often set off by streets, railroads or waterways (Statistics Netherlands, 2010). As Amsterdam is the largest and most densely populated city, the average neighbourhood population is highest: 11,781 inhabitants (2011). Tilburg and The Hague have an average neighbourhood population of respectively 6729 and 6008. However, the neighbourhoods vary in size: from a population of 150 to two neighbourhoods of over 20,000 inhabitants.

Neighbourhood Change and Mobility

The theoretical framework showed that neighbourhood change is often linked to mobility: upgrading and downgrading are related to in- and out-migration of lower- and higher-income groups. Therefore, it is expected that both upgrading and downgrading go hand-in-hand with high mobility rates. However, in the neighbourhoods of this study, there are downgrading neighbourhoods with high and low mobility, as well as upgrading neighbourhoods with high and low mobility. This is reflected in Figure 1, which demonstrates the relationship between social upgrading and downgrading and mobility patterns in the three cities. Figure 1 distinguishes nine categories: downgrading neighbourhoods with low, average and high mobility; average grading neighbourhoods with low, average and high mobility; and upgrading neighbourhoods with low, average and high mobility. In total, 54 per cent of the households moved between 1999 and 2008. Nevertheless, this distribution is irregular at the neighbourhood level: the proportion of households that moved ranges from 28 to 82 per cent.

Figure 1.

Social upgrading and downgrading and residential mobility patterns in Amsterdam, The Hague and Tilburg, 1999–2008.

Since this paper is interested in the relationship between neighbourhood change and mobility, the remainder of this paper focuses on neighbourhoods experiencing upgrading or downgrading. Since Amsterdam, The Hague and Tilburg show similar patterns for the relationship between mobility and neighbourhood change, it was decided to present the results for the three cities combined. Table 1 shows characteristics of downgrading neighbourhoods with low, average and high mobility (categories A to C) and upgrading neighbourhoods with low, average and high mobility (categories D to F). These categories are discussed next.

Table 1.

Relationship between residential mobility and upgrading and downgrading of neighbourhoods, by category, 1999–2008

	Number of neighbour-hoods	Percentage of neighbour-hoods	Social grading^a	Percentage of moved households(1999–2008)	Income level in 1999 (in euro)	S.D. of income level in 1999	Income level in 2008 (in euro)	S.D. of income level in 2008	Real estate value in 1999 (in euro)	S.D. of real estate value in 1999	Real estate value in 2008 (in euro)	S.D. of real estate value in 2008	Percentage owner-occupied housing	Percentage social rented housing	Percentage private rented housing
A. Downgrading and low mobility	16	11.9	−7	47.0	14,456	2017	18,043	2769	137,924	33,473	199,704	56,374	23.8	68.2	8.0
B. Downgrading and average mobility	20	14.9	−6	54.7	13,526	1652	16,959	1992	103,993	37,757	183,610	65,962	25.5	64.4	10.1
C. Downgrading and high mobility	23	17.2	−8	63.5	13,207	1277	15,933	1417	76,025	31,750	139,631	43,619	37.1	43.4	19.5
D. Upgrading and low mobility	24	17.9	6	44.7	17,907	3325	24,965	4932	199,711	118,341	394,651	263,404	48.7	30.5	20.8
E. Upgrading and average mobility	20	14.9	6	53.8	15,774	2784	22,267	3743	140,039	47,973	272,645	116,931	32.6	51.5	15.9
F. Upgrading and high mobility	31	23.2	7	60.6	13,555	1279	19,249	1996	121,186	51,426	246,010	74,866	29.2	41.0	29.8
Total	134	100

Social grading refers to the growth percentage of the neighbourhood income level in relation to the growth percentage of the city-wide income level between 1999 and 2008 (see Data and Methods).

Note: Each neighbourhood is categorised into one of the six categories, on the basis of their social grading level (downgrading or upgrading) and mobility rate (low, average or high).

Source: Statistics Netherlands and Kadaster, author’s adaptation.

Downgrading Neighbourhoods

Table 1 demonstrates that there are only marginal differences in the level of downgrading between low, average and high mobility neighbourhoods: between low and high mobility downgrading neighbourhoods, the difference is only 1 per cent. Downgrading neighbourhoods with low and average mobility (A, B) differ from those with high mobility (C). First, category A and B neighbourhoods show higher incomes and real estate values. Most of these neighbourhoods are post-war constructed and are characterised by homogeneous multi-family dwellings and ethnically diverse households. Especially A and B neighbourhoods in Amsterdam and The Hague have especially high proportions of social housing. Waiting-lists for social housing could impede mobility rates, as households relying on social housing may not be able to move. A and B neighbourhoods in Tilburg show higher levels of owner-occupied (single-family) housing, which is relatively accessible to lower-income householdss. Low mobility could be understood as a sign of well-functioning neighbourhoods. Category C neighbourhoods have lower incomes and real estate values. A number of these neighbourhood ares undergoing or have recently undergone restructuring. Households are forced to move when their dwelling is renovated or demolished, which could explain the high mobility rates. Other C neighbourhoods are characterised by accessible housing: real estate values are comparatively low and, especially in Amsterdam and The Hague, there is a large private rented sector. This could explain high mobility, as private renters often show higher mobility rates (Bailey and Livingston, 2007).

Upgrading Neighbourhoods

Within upgrading neighbourhoods, the difference in upgrading levels between low and high mobility neighbourhoods is also only 1 per cent (Table 1). Category F neighbourhoods had relatively low incomes and real estate values in 1999. Many are pre-war constructed and former working-class neighbourhoods. The housing stock is characterised by small apartments and the majority are privately and socially rented. These neighbourhoods recently started to gentrify, which is reflected in strong increases in real estate values. Many in-migrants are in the early stages of their life-course. After a few years they are likely to move, which could explain the high mobility rates. This is in line with van Criekingen (2009), who showed that living in gentrifying neighbourhoods is a transitional stage in the housing career of many young households.

Category D (and E to a lesser extent) neighbourhoods are generally neighbourhoods at the top of the hierarchy. Some have had an affluent population for many decades. Others were among the first gentrifying neighbourhoods, but gentrification is now at an advanced stage. The majority of D and E neighbourhoods are centrally located and pre-war constructed, characterised by owner-occupied dwellings and (native Dutch) families with children. The presence of these ‘settled’ households could explain the low mobility, as they are in further life-course stages. Also, these neighbourhoods may be inaccessible to households with modest incomes, through high prices and relatively few rented dwellings.

Income Developments of In-migrants, Out-migrants and Non-migrants

The previous section showed that differences in the level of social grading between low, average and high mobility neighbourhoods are relatively small. This poses the question of what the income developments of in-migrants, out-migrants and non-migrants are and how these relate to patterns of neighbourhood change.

In-migrants

Table 2 presents income levels of households who moved into a neighbourhood, aggregated per category, for each year between 2000 and 2008. These income levels are relative to their neighbourhood’s income level (which is 100). For instance, the table shows that in 2000 incomes of in-migrants in category A were 9 per cent below their neighbourhood’s income, while in 2008, income levels of in-migrants in this category were 31 per cent below the neighbourhood income. To be clear, the composition of in-migrants varies each year: for instance, in-migrants in 2000 were different households from in-migrants in 2001, since in-migrants in 2000 became non- or out-migrants in 2001. Table 2 shows two remarkable patterns: incomes of in-migrants show similar patterns both in upgrading and downgrading neighbourhoods and there are no substantive differences between low, average and high mobility neighbourhoods. In all six categories, average income levels of in-migrants in a year are below the neighbourhood level. This pattern applies to all years between 2000 and 2008.⁴ There are differences, however. In general, the gap between incomes of in-migrants and average neighbourhood incomes is wider in downgrading than in upgrading neighbourhoods. In downgrading neighbourhoods, in-migrants reinforce downgrading processes, as their incomes are much below the neighbourhood income. This is in line with assumptions in the literature. In upgrading neighbourhoods, however, in-migrants seem to impede rather than reinforce upgrading, as their incomes are a downward force on their neighbourhood’s income.

Table 2.

Relative average income of in-migrating households, per category, 2000–2008 (neighbourhood income level is 100)

	2000	2001	2002	2003	2004	2005	2006	2007	2008
A. Downgrading and low residential mobility	91	94	91	72	72	70	72	72	69
B. Downgrading and average residential mobility	90	91	88	71	71	72	73	72	72
C. Downgrading and high residential mobility	91	92	88	72	72	69	71	71	69
D. Upgrading and low residential mobility	95	97	94	74	74	73	76	76	78
E. Upgrading and average residential mobility	96	98	96	76	78	80	81	82	83
F. Upgrading and high residential mobility	96	99	95	76	76	76	79	80	82

Notes: The table presents the average income level of in-migrants in a neighbourhood, aggregated per category, between 2000 and 2008. The income level is relative to the neighbourhood income (which is 100). The formula is: ((average income level of in-migrants of a neighbourhood in category X in year Y/average neighbourhoods’ income in category X in year Y)*100); X means respectively categories A to F, Y means the years 2000 to 2008.

Source: Statistics Netherlands, author’s adaptation.

These observations pose the question as to how far incomes of in-migrants increase once they have moved in. Therefore, for each category, the income development of households who moved into a neighbourhood in 2000, 2001 and 2002 has been studied, but only with respect to those households still living in the neighbourhood in 2008. Table 3 shows the income development of households who moved into a neighbourhood in 2000, aggregated per category. The table demonstrates both the absolute income development of in-migrants and their income development relative to the neighbourhood development (which is 100). For instance, the table shows that between 2000 and 2008, in-migrants in category A experienced an absolute increase in income of 32 per cent. Also, the table shows that, incomes of these in-migrants increased 15 per cent faster than the neighbourhood income. Table 3 makes clear that in all categories, in-migrants experienced a significant absolute increase in income between 2000 and 2008. Moreover, their incomes increased faster than their neighbourhood’s income. So, while Table 2 showed that, in both upgrading and downgrading neighbourhoods, incomes of in-migrants were below the neighbourhood level at the moment they moved in, in the years after moving in, in-migrants experience (strong) upgrading. After a few years, their income reaches the neighbourhood income level. Households who in-migrated in 2001 and 2002 show similar patterns (results not shown here). One explanation could be that households anticipate having higher incomes in the near future when in-migrating.

Table 3.

Income development of households who moved into a neighbourhood in 2000, aggregated per category

		2000	2001	2002	2003	2004	2005	2006	2007	2008
A. Downgrading and low mobility	Absolute income development^a	100	114	119	119	119	119	126	131	132
	Relative income development^b	100	106	107	109	110	111	113	113	115
B. Downgrading and average mobility	Absolute income development^a	100	117	120	123	122	121	128	137	138
	Relative income development^b	100	108	108	113	113	113	115	118	120
C. Downgrading and high mobility	Absolute income development^a	100	117	121	120	120	121	127	138	136
	Relative income development^b	100	108	108	112	112	116	117	121	125
D. Upgrading and low mobility	Absolute income development^a	100	114	118	117	119	120	126	131	129
	Relative income development^b	100	104	104	103	104	105	105	103	104
E. Upgrading and average mobility	Absolute income development^a	100	116	123	125	127	129	140	148	144
	Relative income development^b	100	107	108	111	112	113	115	115	116
F. Upgrading and high mobility	Absolute income development^a	100	121	127	131	134	135	145	152	148
	Relative income development^b	100	110	110	115	116	117	119	117	120

This shows the absolute indexed income development of in-migrants in 2000. The formula is: ((income of in-migrants in category X in year Y)/(income of in-migrants in category X in 2000))*100; X means the categories A to F, Y means the years 2000 to 2008.

This income development of in-migrants is relative to the neighbourhood income development (which is 100), and aggregated per category. The formula is: ((income of in-migrants in category X in year Y/income of in-migrants in category X in 2000)/(neighbourhood income in category X in year Y/neighbourhood income in category X in 2000)); X means the categories A to F, Y means the years 2000 to 2008.

Source: Statistics Netherlands, author’s adaptation.

There are differences between the categories. In upgrading neighbourhoods with average and high mobility, in-migrants experienced the greatest absolute increase and after a few years, their incomes are (much) above the neighbourhood level. In other words, in-migrants in upgrading neighbourhoods are important less for their income when in-migrating, but primarily for their future income development. Hereby, they contribute significantly to the upgrading of their neighbourhood. In-migrants in category F experienced the strongest incumbent upgrading (absolute and relative). Many F neighbourhoods recently started to gentrify and many young households in-migrated. For instance, in 2000, about half of all in-migrants were below 30 years old.⁵ These households are in early life-course stages, which are usually characterised by strong upward dynamics (Hoogvliet, 1992). This could explain the observed incumbent upgrading. However, in-migrants in category D and E neighbourhoods also experienced significant income gains, while many in-migrants are in further life-course stages. For instance, in these categories, more than 60 per cent of all in-migrants are above 30 years old. So, although it is often assumed that later stages of one’s life-course are characterised by weaker upward dynamics (or stabilisation) in terms of income (Hoogvliet, 1992), these in-migrants still experience significant income gains.

Within downgrading neighbourhoods also, in-migrants experience incumbent upgrading, despite their low incomes when moving in. So, although in-migrants reinforce downgrading when moving in, they impede downgrading in later years, as their incomes increase significantly. As mentioned, many C neighbourhoods undergo restructuring whereby the government aims to attract higher-income households. Therefore, it is remarkable that these in-migrants show relatively low incomes. Nevertheless, this is in line with Wittebrood and van Dijk (2007), who showed that socioeconomic changes of restructuring neighbourhoods have not been substantial, despite restructuring policies. However, these in-migrants also show significant income gains in later years. In A and B neighbourhoods, characteristics of in-migrants are mixed: the proportion of older and younger in-migrants is similar and many non-native households and families with children have moved in. This could be explained by the presence of large proportions of social housing, on which many non-native and low-income families rely.

Out-migrants

This section examines the income developments of out-migrants, in order to obtain insight in their contribution to neighbourhood change. Table 4 presents income levels of households who moved out of a neighbourhood, aggregated per category, for each year between 1999 and 2007. These incomes are relative to the neighbourhood income (100). For example, the table shows that in 1999, incomes of out-migrants in category F are 3 per cent above the neighbourhood level, while in 2001, incomes of these out-migrants are 4 per cent above the neighbourhood average. Table 4 demonstrates that incomes of out-migrants in both upgrading and downgrading neighbourhoods are not consistently higher than the neighbourhood income. It is clear, however, that their incomes are above those of in-migrants. Within downgrading neighbourhoods, in some years incomes of out-migrants are above the neighbourhood level. Here, out-migrants reinforce downgrading, which is in line with assumptions in the literature. Nonetheless, in other years, incomes of out-migrants in the same categories are around or even below the neighbourhood income.

Table 4.

Relative average income of out-migrating households, per category, 1999–2007 (neighbourhood income level is 100)

	1999	2000	2001	2002	2003	2004	2005	2006	2007
A. Downgrading and low residential mobility	100	104	105	101	93	89	88	88	86
B. Downgrading and average residential mobility	96	99	101	97	90	89	88	90	86
C. Downgrading and high residential mobility	98	100	102	97	91	90	87	91	89
D. Upgrading and low residential mobility	89	94	95	93	84	84	82	84	81
E. Upgrading and average residential mobility	103	105	104	102	96	91	94	94	89
F. Upgrading and high residential mobility	103	103	104	102	95	95	96	96	94

Notes: The table presents the average income level of out-migrants in a neighbourhood, aggregated per category, between 1999 and 2007. The income level is relative to the neighbourhood income (which is 100). The formula is: ((average income level of out-migrants of a neighbourhood in category X in year Y/average neighbourhoods’ income in category X in year Y)*100); X means respectively categories A to F, Y means the years 1999 to 2007.

Source: Statistics Netherlands, author’s adaptation.

Regarding upgrading neighbourhoods, the literature often assumes that upgrading is related to the out-migration of households with lower incomes. This seems to apply to most of the years in category D, E and F neighbourhoods: generally, incomes of out-migrants are below the neighbourhood income. Noteworthy, their incomes are still (much) higher than the incomes of out-migrants of downgrading neighbourhoods. In categories E and F, however, in a number of years the incomes of out-migrants are above the neighbourhood level. Hereby, out-migrants impede upgrading.

In conclusion, the contribution of out-migrants to neighbourhood change is complex. In some cases, out-migration reinforces neighbourhood change, but in other cases, out-migration impedes these processes. In order further to examine the contributions of out-migrants to neighbourhood change, their income development has been studied for the years before moving out. Income developments of out-migrants in 2005, 2006 and 2007 have been examined, but only with respect to those who already lived in the neighbourhood in 1999. Table 5 demonstrates income developments of out-migrants in 2007, aggregated per category. The table shows both the absolute income development of out-migrants and their income development relative to the neighbourhood development (which is 100). For instance, the table demonstrates that between 1999 and 2007, out-migrants in category A experienced an absolute increase in income of 41 per cent. The table also shows that incomes of these out-migrants increased 11 per cent faster than the neighbourhood income.

Table 5.

Income development of individual households who moved out of a neighbourhood in 2007, in the years before out-migration, aggregated per category

		1999	2000	2001	2002	2003	2004	2005	2006	2007
A. Downgrading and low mobility	Absolute income development^a	100	118	127	129	132	133	132	139	141
	Relative income development^b	100	108	108	107	111	113	113	115	111
B. Downgrading and average mobility	Absolute income development^a	100	107	123	121	122	121	122	127	130
	Relative income development^b	100	98	105	100	103	104	104	104	102
C. Downgrading and high mobility	Absolute income development^a	100	110	122	125	126	125	128	132	137
	Relative income development^b	100	99	101	101	105	105	109	108	107
D. Upgrading and low mobility	Absolute income development^a	100	110	121	125	128	127	132	134	134
	Relative income development^b	100	98	98	97	99	97	100	97	92
E. Upgrading and average mobility	Absolute income development^a	100	112	121	130	129	132	134	141	145
	Relative income development^b	100	99	98	101	102	103	102	102	100
F. Upgrading and high mobility	Absolute income development^a	100	112	126	129	130	132	134	143	148
	Relative income development^b	100	98	100	98	100	101	102	102	100

This shows the absolute indexed income development of out-migrants who moved out in 2007. The formula is: ((income of out-migrants in category X in year Y)/(income of out-migrants in category X in 1999))*100; X means the categories A to F, Y means the years 1999 to 2007.

The income development of out-migrants is relative to the neighbourhood income development (which is 100), and aggregated per category. The formula is: ((income of out-migrants in category X in year Y/income of out-migrants in category X in 1999)/(neighbourhood income in category X in year Y/neighbourhood income in category X in 1999)); X means the categories A to F, Y means the years 1999 to 2007.

Source: Statistics Netherlands, author’s adaptation.

The table shows that out-migrants in both upgrading and downgrading neighbourhoods experienced strong income gains in the years before leaving. When looking at their income development relative to the neighbourhood development, there are differences between the categories. In downgrading neighbourhoods, incomes of out-migrants increased (much) faster than their neighbourhood’s income. So, they impede downgrading in the years before leaving. Out-migration could be related to disequilibrium between current and potential housing consumption: the dwelling and/or neighbourhood might not match the household anymore. Out-migration could also be related to dissatisfaction with the neighbourhood. Many downgrading neighbourhoods are low in hierarchy and the strong income gains might be an opportunity to climb the housing ladder. Interestingly, incomes of out-migrants in C neighbourhoods increased strongly too, where mobility was partly triggered by institutional interventions.

Although out-migrants in upgrading neighbourhoods experienced significant absolute income gains too, their increase in income was generally around their neighbourhood’s increase in income. Also, their absolute income levels are mainly around or below the neighbourhood’s income in all categories.⁶ In other words, these households neither reinforce or impede upgrading significantly.

Non-migrants

Table 6 demonstrates income levels of non-migrants, aggregated per category. These are both households who did not move and households who moved within neighbourhoods. These incomes are relative to the neighbourhood income (which is 100). To be clear, the composition of non-migrants varies each year: for instance, non-migrants in 2001 may have been in-migrants in 2000 or have become out-migrants in 2002. Unfortunately, the average stay of households who in-migrated before 1999 is unknown, so distinctions between long-term and short-term households cannot be made. Nevertheless, Table 6 demonstrates that incomes of non-migrants in all years are above the neighbourhood level, both in upgrading and downgrading neighbourhoods. The table demonstrates that within downgrading neighbourhoods, non-migrants impede downgrading, as their incomes are generally an upward force to their neighbourhood’s income. Within upgrading neighbourhoods, however, non-migrants reinforce upgrading, as their incomes are above the neighbourhood level. Non-migrants seem to be in advanced life-course stages: the share of families with children and retired households is high. A situation of not-moving may mean that households are satisfied with their neighbourhood, but households may also be unable to move. However, the data only allow formulating hypotheses about non-migrants.

Table 6.

Relative average income of non-migrants, per category, 2000–2007 (neighbourhood income level is 100)

	2000	2001	2002	2003	2004	2005	2006	2007
A. Downgrading and low residential mobility	102	101	102	104	104	105	105	105
B. Downgrading and average residential mobility	102	102	102	105	105	105	105	106
C. Downgrading and high residential mobility	102	102	103	107	107	109	107	107
D. Upgrading and low residential mobility	102	102	102	104	105	105	104	103
E. Upgrading and average residential mobility	101	101	101	104	104	104	103	103
F. Upgrading and high residential mobility	101	101	101	105	104	105	104	104

Notes: The table presents the average income level of non-migrants in a neighbourhood, aggregated per category, between 2000 and 2007. The income level is relative to the neighbourhood income (which is 100). The formula is: ((average income level of non-migrants of a neighbourhood in category in X year Y/average neighbourhoods income in category X in year Y)*100); X means respectively categories A to F, Y means the years 2000 to 2007.

Source: Statistics Netherlands, author’s adaptation.

Summary and Conclusions

This paper focused on the relationship between neighbourhood upgrading and downgrading and mobility. The paper examined income developments of in-migrants, out-migrants and non-migrants and related these to neighbourhood income developments. Most studies on neighbourhood change take for granted that mobility is the main contributor to neighbourhood change, while incumbent processes are often ignored. This paper has shown that in-migration and out-migration are not the only processes at work, and that changes in the socioeconomic status of non-migrants are of importance too, either in reinforcing or impeding neighbourhood change. There are differences between upgrading and downgrading neighbourhoods, however.

First, it is often assumed that high mobility triggers neighbourhood decline. This study has shown that, within downgrading neighbourhoods, there are only marginal differences in the level of downgrading between low and high mobility neighbourhoods. Nevertheless, in-, out- and non-migrants show generally the same patterns in downgrading neighbourhoods with low, average and high mobility. In line with previous studies, in-migrants reinforce downgrading, as their incomes are generally below the neighbourhood income. In other words, their incomes are a downward force on their neighbourhood’s income. The contribution of out-migrants to downgrading is more complex: in some cases they reinforce downgrading, when their incomes are above the neighbourhood level. This is in line with the findings of Andersson and Bråmå (2004). However, in other cases out-migrants impede the process. This paper has demonstrated that in-migrants experience strong incumbent upgrading after in-migration. This raises the question of whether in-migration of low-income households is problematic when their income increases significantly in later years. In addition, out-migrants show strong incumbent upgrading before moving out and incomes of non-migrants are generally above the neighbourhood income. This demonstrates that non-migrants of downgrading neighbourhoods impede downgrading. This corresponds with Bailey (2012), who demonstrated that (among others) status changes of sitting households reduced segregation patterns.

Within upgrading neighbourhoods also there are only marginal differences in the level of upgrading between low and high mobility neighbourhoods. In-migrants show generally the same patterns: initially, they impede upgrading, as their incomes are below the neighbourhood level when in-migrating. However, they experience strong upgrading in the years after moving in. So, in-migrants in upgrading neighbourhoods are important less for their income when moving in, but mainly for their future income development. The contribution of out-migrants to upgrading is more complex. In most of the years, out-migrants reinforce upgrading, as their incomes are below the neighbourhood level. However, in other years, incomes of out-migrants are above the neighbourhood level, whereby they impede upgrading. Incumbent processes seem to be the main driver behind upgrading, as incomes of non-migrants are systematically above the neighbourhood level and in-migrants experience strong incumbent upgrading in the years after in-migrating. So, although incumbent upgrading is often overlooked in the literature, this paper demonstrates the importance of looking at changes in the incomes of non-migrants. This is in line with Bailey and Livingston (2008), Bailey (2012) and Shuttleworth et al. (2012): they demonstrate that mobility is less important than assumed in patterns of neighbourhood change.

One explanation for the observed incumbent upgrading in upgrading and downgrading neighbourhoods may be that households are less willing to move after income changes than assumed (Bailey, 2012). Households may be satisfied in their neighbourhood through processes of attachment. Moreover, a significant part of the Dutch housing stock is social housing, characterised by long waiting-lists. So, households relying on social housing may wish to move, but might not be able to. After obtaining social rented dwellings, households cannot be forced to move after income gains. This means that, for households no longer complying with allocation rules, moving is associated with strong increases in housing costs. This may lead households to stay in their dwelling. Another explanation could be that differences between neighbourhoods are relatively small as a result of the highly regulated housing market. This might impede the urge to climb the neighbourhood hierarchy after income gains.

Whereas the observed patterns and trends merit further and more detailed research, this paper’s aim was to explore general patterns of the contribution of in-, out- and non-migration to neighbourhood change. This paper raises further questions and gives directions for future research. First, the paper challenges other researchers to pay more attention to the contributions of in-migration, out-migration and especially non-migration to neighbourhood change. Secondly, the paper has shown that mobility and non-mobility patterns differ per household and neighbourhood and indicate that mobility is closely related to age and life-course. Findings suggest that neighbourhoods with high mobility accomplish a function for households starting their housing career, while in low mobility neighbourhoods, more households in later stages of their housing career reside. This paper urges for more insight into stages of housing careers of in-, out- and non-migrants, as it would provide a better understanding of the social dynamics of neighbourhoods and allow formulating expectations concerning their future development.

Footnotes

Acknowledgements

The author thanks three anonymous referees, the UvA-NICIS-Corpovenista research board, Sako Musterd, Fenne Pinkster, Wouter van Gent and Willem Boterman for their help, suggestions and criticisms. The author also thanks Statistics Netherlands and Kadaster for providing data.

Funding

This paper is funded by NICIS Institute, Platform Corpovenista and the municipalities of Amsterdam, The Hague and Tilburg.

Notes

References

Andersson

Bråmå

(2004) Selective migration in Swedish distressed neighbourhoods: can area-based urban policies counteract segregation processes?, Housing Studies, 19, pp. 517–539.

Atkinson

(2000) Measuring gentrification and displacement in Greater London, Urban Studies, 37, pp. 149–165.

Bailey

(2012) How spatial segregation changes over time: sorting out the sorting process, Environment and Planning A, 44, pp. 705–722.

Bailey

Livingston

(2007) Population Turnover and Area Deprivation. Bristol: Policy Press.

Bailey

Livingston

(2008) Selective migration and neighbourhood deprivation: evidence from 2001 census migration data for England and Scotland, Urban Studies, 45, pp. 943–961.

Bolt

van Kempen

Weesep

J. van

(2009) After urban restructuring: relocations and segregation in Dutch cities, Tijdschrift voor Economische en Sociale Geografie, 100, pp. 502–518.

Bourne

L. S.

(1993) The myth and reality of gentrification: a commentary on emerging urban forms, Urban Studies, 30, pp. 183–189.

Brakel

M. van den

(2009) Inkomensongelijkheid relatief gering in ons land. CBS Webmagazine.

Burgess

E. W.

(1925/1974) The growth of the city: an introduction to a research project, in: Park

R. E.

Burgess

E. W.

McKenzie

R. D.

(Eds) The City, ch. 2. Chicago, IL: University of Chicago Press.

10.

Butler

Robson

(2001) Social capital, gentrification and neighbourhood change in London: a comparison of three South London neighbourhoods, Urban Studies, 38, pp. 2145–2162.

11.

Clark

W. A. V.

Dieleman

F. M.

(1996) Households and housing: choice and outcomes in the housing market. Center for Urban Policy Research, New Brunswick, QC.

12.

Clay

(1979) Neighborhood Renewal: Middle Class Resettlement and Internal Upgrading in American Neighborhoods. Lexington, MA: Lexington Books.

13.

Criekingen

M. van

(2009) Moving in/out of Brussels’ historical core in the early 2000s: migration and the effects of gentrification, Urban Studies, 46, pp. 825–848.

14.

Criekingen

M. van

Decroly

J. M.

(2003) Revisiting the diversity of gentrification: neighbourhood renewal processes in Brussels and Montreal, Urban Studies, 40, pp. 2451–2468.

15.

Fielding

A. J.

(1992) Migration and social mobility: south east England as an escalator region, Regional Studies, 26, pp. 1–15.

16.

Freeman

(2005) Displacement or succession? Residential mobility in gentrifying neighborhoods, Urban Affairs Review, 40, pp. 463–491.

17.

Galster

(1987) Homeowners and Neighborhood Reinvestment. Durham, NC: Duke University Press.

18.

Glass

(1964) Introduction: aspects of change, in: for Center Urban Studies (Ed.) London: Aspects of Change, pp. xii–xli. London: MacGibbon and Kee.

19.

Grigsby

Baratz

Galster

Maclennan

(1987) The dynamics of neighbourhood change and decline, Progress in Planning, 28, pp. 1–76.

20.

Ham

M. van

Clark

W. A. V.

(2009) Neighbourhood mobility in context: household moves and changing neighbourhoods in the Netherlands, Environment and Planning A, 41, pp. 1442–1459.

21.

Hoogvliet

(1992) Wijken in beweging: Bevolkingsdynamiek en wooncarrières in vroeg-20ste-eeuwse woongebieden. Stedelijke Netwerken, Utrecht.

22.

Hoyt

(1939) The structure and growth of residential neighborhoods in American cities. Federal Housing Administration, Washington, DC.

23.

Kullberg

(1997) From waiting lists to adverts: the allocation of social rental dwellings in the Netherlands, Housing Studies, 12, pp. 393–403.

24.

Lees

(2008) Gentrification and social mixing: towards an inclusive urban renaissance?, Urban Studies, 45, pp. 2449–2470.

25.

Lupton

Power

(2004) What we know about neighbourhood change: a literature review. CASE Report No. 27, Centre for Analysis of Social Exclusion, London School of Economics.

26.

McKinnish

Walsh

White

T. K.

(2010) Who gentrifies low-income neighborhoods?, Journal of Urban Economics, 67, pp. 180–193.

27.

Millard-Ball

(2002) Gentrification in a residential mobility framework: social change, tenure change and chains of moves in Stockholm, Housing Studies, 17, pp. 833–856.

28.

Mulder

C. H.

(1993) Migration Dynamics: A Life-course Approach. Amsterdam: Thesis Publishers.

29.

Murie

Knorr-Siedow

Kempen

R. van

(2003) Large housing estates in Europe: general developments and theoretical backgrounds. Faculty of Geosciences, Utrecht University.

30.

Musterd

Kempen

R. van

(2007) Trapped or on the springboard? Housing careers in large housing estates in European cities, Journal of Urban Affairs, 29, pp. 311–329.

31.

OECD (Organisation for Economic Co-operation and Development) (2009) What are equivalence scales? (www.oecd.org/els/socialpoliciesanddata/35411111.pdf; accessed 10 October 2012).

32.

Rossi

P. H.

(1955) Why Families Move: A Study in the Social Psychology of Urban Residential Mobility. Glencoe, IL: Free Press.

33.

Shuttleworth

James

Gould

(2012) Does internal migration in Northern Ireland increase religious and social segregation? Perspectives from the Northern Ireland longitudinal study (NILS) 2001–2007, Population, Space and Place, 19, pp. 72–86.

34.

Smith

(1996) The New Urban Frontier: Gentrification and the Revanchist City. London: Routledge.

35.

Statistics Netherlands (2010) Begrippen (www.cbs.nl/nl-NL/menu/methoden/begrippen; accessed 10 October 2012).

36.

Uitermark

Duyvendak

J. W.

Kleinhans

(2007) Gentrification as a governmental strategy: social control and social cohesion in Hoogvliet, Rotterdam, Environment and Planning A, 39, pp. 125–141.

37.

Wittebrood

Dijk

T. van

(2007) Aandacht voor de Wijk. Den Haag: SCP.