New housing development, selective mobility patterns and ethnic residential segregation

Transcription

1 Enhr Conference July, Toulouse New housing development, selective mobility patterns and ethnic residential segregation Sanne Boschman PBL Netherlands Environmental Assessment Agency, The Hague, The Netherlands Ronald van Kempen Faculty of Geosciences of Utrecht University, Utrecht, The Netherlands Gideon Bolt Faculty of Geosciences of Utrecht University, Utrecht, The Netherlands Frank van Dam PBL Netherlands Environmental Assessment Agency, The Hague, The Netherlands Abstract Urban restructuring programmes in the Netherlands try to achieve more mixed neighbourhoods in terms of ethnicity by building more expensive and owner-occupied dwellings in deprived concentration neighbourhoods. At the same time large-scale new housing estates are built at the edges of cities, with mostly expensive and owner-occupied dwellings. In this paper the focus is on the effect of building these dwellings on selective mobility patterns and ethnic residential segregation. Our results show that newly built dwellings in deprived concentration neighbourhoods attract relatively high shares of native Dutch from elsewhere. At the same time however they keep ethnic minorities within concentration neighbourhoods that otherwise might have left. Large scale new housing estates attract average shares of minority groups and therefore have a small effect on ethnic residential segregation. Policymakers striving for less segregated cities should take into account this mixed impact of new housing development on selective mobility patterns and segregation. Introduction Segregation refers to the unequal distribution of population groups across space. When a group lives concentrated in one area and is underrepresented in another area compared to the total population this group is segregated. In this article the focus is on the dynamics of segregation of ethnic minority groups in cities in the Netherlands. One of the goals of the Dutch policy of urban restructuring is to achieve more mixed neighbourhoods in terms of ethnicity. Neighbourhoods with large numbers of relatively inexpensive dwellings and concentrations of ethnic minority groups are restructured; inexpensive dwellings are demolished and new more expensive dwellings are built in its stead. It is however unclear to what extent this policy succeeds in achieving a residential mix of ethnic groups.

2 Mixité : an urban and housing issue? 2 Besides building for restructuring, large numbers of new dwellings are built to accommodate the increase in the number of households, mostly at large scale new housing estates on greenfield areas bordering cities. These new and mostly relatively expensive dwellings will also engender selective mobility patterns and thereby influence ethnic residential segregation. Large scale expensive housing development on greenfield locations may enlarge segregation and undo the results of the investments in restructuring and mixing neighbourhoods. The main question in this paper is: How does new housing development, both in restructuring neighbourhoods and in new housing estates affect ethnic selective mobility patterns and how does this affect ethnic residential segregation? In the next section we will give an overview of the theory. Firstly we will describe the theories of preferences and constraints to explain segregation. Secondly, we focus on theory on the relation between new housing development residential mobility and ethnic residential segregation. In the research design the data and methods are described and the three cities, Rotterdam, The Hague and Utrecht are introduced. Then we proceed by describing the new housing development in these three cities; differences in the amount of new housing development and the distribution over the city will have different influence on the mobility patterns and ethnic residential segregation. In the next paragraph the segregation index is calculated and the effects of the different neighbourhoods on the development of the ethnic residential segregation are described. The last step is to zoom in on the residential mobility patterns to and from neighbourhoods and there effects on ethnic residential segregation. We describe the ethnic selectivity in mobility patterns from neighbourhoods and to newly built dwellings and the effect of these patterns on ethnic residential segregation in the origin and destination neighbourhoods. Based on this we will be able to draw some conclusions about the relation between new housing development in cities, ethnic selective mobility patterns and ethnic residential segregation. Theory Basically, there are two main explanations for ethnic residential segregation. Firstly ethnic residential segregation can be explained by preferences and choices. Differences in housing preferences between ethnic groups or the wish to live among the own group, or not among other ethnic groups lead to ethnic residential segregation. Secondly constraints can limit the housing opportunities of certain ethnic groups or exclude them from living somewhere and thereby lead to ethnic residential segregation. Opportunities and constraints can be induced by personal characteristics, for instance when ethnic minorities have on average lower incomes and therefore less housing opportunities. Constraints can, however, also be directly related to ethnicity when discrimination limits the housing opportunities of ethnic minority groups. Ethnic differences in preferences and ethnic residential segregation Research into residential segregation started with the Chicago School (Park et al, 1925). They explain segregation as a pattern of different neighbourhoods in which different population groups live. As one (ethnic) group starts to enter a neighbourhood (invasion), individuals from the (ethnic) group that used to live there will leave (succession) and the new group slowly takes over the whole neighbourhood. Schelling (1971; see also Clark, 1991) explains this process this process of invasion and succession by small individual differences in preferences on the residential ethnic composition. Most people will tolerate some people from another ethnic group in their neighbourhood. However, as members of a new group enter a formerly mono-ethnic neighbourhood, some former inhabitants will leave the neighbourhood, then more members of the new group will enter and more former inhabitants will leave the neighbourhood, as they do not tolerate the new and higher shares of the new group in their neighbourhood. This process will continue until the new group has completely taken over the neighbourhood, therefore there can not be durable mixed neighbourhoods. Research in the Netherlands shows that also in the Netherlands in neighbourhoods with increasing concentrations of ethnic minorities, individuals more often want to leave their neighbourhood (Van Ham and Clark, 2009). Especially indigenous Dutch are more likely to move as the share of ethnic

3 3 Workshop 16: Minority Ethnic Groups and Housing minorities in their neighbourhood increases (Feijten and Van Ham, 2009) and they more often move to white neighbourhoods (Bolt et. al., 2008). This indicates that individual preferences and choices to live among the own ethnic group or not among other ethnic groups will partly explain ethnic residential segregation in the Netherlands. Besides the preferences of the indigenous majority to live among the own ethnic group, also preferences of minority ethnic groups can affect segregation. Members of minority ethnic groups can have preferences to live in concentrated neighbourhoods among the own ethnic group. These neighbourhoods provide a support network for new immigrants (Zorlu and Mulder, 2008), ethnic facilities and shops (Wilson and Portes, 1981), more social contacts with neighbours (Putnam, 2007) and inhabitants do not have to fear discrimination (Logan et al, 2002). Research in the Netherlands shows that ethnic minorities are more likely to move to black neighbourhoods than native Dutch (Zorlu and Latten, 2007). Thirdly, besides the preferences of the indigenous majority and of ethnic minority groups to live among the own ethnic group, segregation can be affected by differences in housing preferences between ethnic groups. According to the theories of Wolpert (1966) and Brown and Moore (1970) residential mobility can be explained by differences between the desired housing situation and the actual housing situation. Every household has different preferences regarding their desired housing situation. Dissatisfaction with the actual housing situation, (e.g. when it deviates from the desired housing situation) will lead to a decision to move. Personal characteristics like age, household situation but also ethnicity will influence the desired housing situation and therefore influence residential mobility patterns and ethnic residential segregation. Ethnic minority groups in the Netherlands spend relatively less than natives on housing (Kullberg and Nicolaas, 2009). Some minorities are especially attached to certain neighbourhoods and cities, because they want to live close to family members or specific (ethnic) facilities (Zorlu, 2009). Also minorities that do not want a mortgage for religions reasons are restricted in their housing opportunities (Kullberg et. al., 2009). Minority groups more often than the indigenous majority want to move within their own neighbourhood or city and they more often have a preference for newly built dwellings (Van Til, 2008). These specific preferences and choices of minority ethnic groups lead to differences in the housing situation and can therefore engender ethnic residential segregation. Besides these real differences in preferences between minority groups there are also compositional differences. As ethnic minorities in the Netherlands are in general younger and have lower incomes they will have different housing preferences than the indigenous majority. This will also induce ethnic residential segregation. As minority ethnic groups get assimilated in Dutch society they will have less need to live in the proximity of their own ethnic group and their housing preferences will differ less from the indigenous majority. This will decrease ethnic residential segregation (Bolt et al, 2008). Following the indigenous majority, in the ninety nineties suburbanisation started also among immigrants (De Groot, 2004). Especially Surinamese and to a much lesser extent Turks leave the cities (Bolt et al, 2008). Moroccans are still strongly attached to the concentrated cities and neighbourhoods (Zorlu, 2009). Non-western minorities are less likely than the indigenous majority to leave deprived neighbourhoods. This can, however partly be explained by their lower income. When income is taken into account, Surinamese are as likely as indigenous Dutch to leave deprived neighbourhoods, while Turks and Moroccans still more often stay in these neighbourhoods (Bolt en Van Kempen, 2003). Also research of Bolt and others (Bolt et. al, 2008) shows that Surinamese and Antilleans are less segregated and more likely than Turks and Moroccans to move to integrated neighbourhoods. Ethnic differences in opportunities and constraints and ethnic residential segregation As said above both preferences and constraints can induce ethnic residential segregation. Constraints can exclude ethnic minority groups to live in certain dwellings or neighbourhoods and thereby force them to concentrate elsewhere. Institutions on the housing market like real estate agents, lending institutions and social housing letting mechanisms can have discriminatory effects (Alba and Logan, 1991; South and Crowder, 1998), thereby inducing constraints on ethnic minorities. Lending

4 Mixité : an urban and housing issue? 4 institutions can have (possibly unintentionally) less trust in minority ethnic groups, who therefore have more problems in getting a mortgage (Roscigno et al., 2009; Aalbers, 2006). Different researchers (Dagevos, 2009; Kullberg, 2002) state that in the Netherlands since the start of the new letting systems in the social rented sector there is no room for ethnic discrimination in this sector. The letting system can, however, still disadvantage ethnic minorities as they have a worse starting position, less waiting time and in general less knowledge of the letting mechanism (Dagevos, 2009; Bolt, 2001). Besides that, discrimination can still induce constraints on the housing opportunities of ethnic minorities in the private rental sector and the owner occupied sector (Laan Bouma-Doff, 2006). The opportunities on the housing market are largely determined by household income. Ethnic minorities in the Netherlands have a lower income and a weaker housing market position (Bolt et. al., 2002), therefore they will be more dependent on the social housing market. The spatial distribution of social housing and more expensive rented and owner-occupied dwellings over the city will therefore also affect segregation. If social housing is concentrated in certain neighbourhoods this will limit the housing opportunities of low income households among which many ethnic minorities. Restructuring and building more expensive dwellings in concentration neighbourhoods will only further limit the housing opportunities of low income households. An extension of inexpensive social housing in new housing estates and other (high income) neighbourhoods will increase the housing opportunities of the poor (Bolt et. al., 2008). As said above, living in concentration neighbourhoods can have advantages for minorities. If, however, minorities with higher incomes leave the concentration neighbourhood and only the underprivileged minorities stay behind, the effect of ethnic residential segregation will be mainly negative. Also involuntary segregation, if (underprivileged) minorities are forced to live in concentration neighbourhoods because they have no housing opportunities elsewhere, will have mainly negative effects (Bolt et. al., 2002). Policymakers can influence ethnic residential segregation and its effects. Building more expensive dwellings in concentration neighbourhoods will decrease the housing opportunities of the underprivileged, but at the same time enlarge the opportunities of successful minorities to proceed their housing career within the neighbourhood (Bolt et. al., 2008). New housing development and selective mobility patterns New housing development of mostly expensive dwellings both on greenfield locations and in existing neighbourhoods will engender selective mobility patterns and affect socioeconomic segregation. New housing development can also enlarge the housing opportunities of minorities, and therefore their chances to move out of concentrated neighbourhoods (Dwyer, 2007). New housing development enlarges the chances that ethnic minorities move towards mixed or indigenous neighbourhoods (South and Crowder, 1998). Especially new housing estates offer housing opportunities for all individuals who want to improve their housing situation. Also successful minorities from underprivileged neighbourhoods can proceed their housing career in new housing estates because they do not have to fear discrimination from current inhabitants of the neighbourhood (South and Crowder, 1998). Minority ethnic groups will rather move to new housing estates than to existing indigenous neighbourhoods or suburbs, because they fear discrimination in neighbourhoods where they are the only minority and do not want to be far apart from the ethnic community and facilities (Permentier and Bolt, 2006; South and Crowder, 1998). Especially newly built dwellings within their own neighbourhood will be attractive to successful minority groups who want to improve their housing situation. Minority ethnic groups more often have a preference for newly built dwellings and more often want to move within their own neighbourhood to stay close to their family and ethnic facilities (Van Til, 2008). Large scale new housing development of relatively expensive dwellings on greenfield locations will increase socioeconomic segregation. It will attract high income households from existing neighbourhoods (Bolt et al., 2002) and the enlarged housing opportunities will make it more difficult to attract high income households to restructuring projects in deprived neighbourhoods (Van Kempen

5 5 Workshop 16: Minority Ethnic Groups and Housing and Priemus, 1999). The effects on ethnic residential segregation are different. Large scale housing development will create opportunities to move to better and less segregated neighbourhoods, especially for segregated minorities (Dwyer, 2007; South and Crowder, 1998). Restructuring in segregated neighbourhoods will probably not lead to a decrease in ethnic residential segregation. These new dwellings are especially attractive to successful ethnic minorities that thereby can improve their housing situation without having to leave behind the ethnic community and facilities (Van Til, 2008). Restructuring will therefore not lead to a decrease in ethnic residential segregation, but it will succeed in keeping successful minorities within deprived neighbourhoods. Research Design The focus in this research is on three large cities in the Netherlands: Rotterdam (610,000 inhabitants, The Hague (495,000 inhabitants) and Utrecht (311,000 inhabitants). For these three cities we describe the mobility patterns between and within neighbourhoods, their effect on the ethnic composition of the neighbourhoods and on ethnic residential segregation. More than ten per cent of the Dutch population is part of a non-western minority group, non-western minorities are however largely overrepresented in (large) cities. In Rotterdam the share of non-western minorities is 35%, in The Hague this is 33% and in Utrecht 22%. The research is based on data from the Dutch Social Statistical Database (SSD), a database that contains information on all inhabitants of the Netherlands (i.e. about 17 million people). The SSD contains information on personal characteristics (age, ethnicity, income, household situation, etc.) and residence address for each year over the years 1999 to Therefore individuals can be followed over time. The dataset can for example be used to trace if an individual has moved house and which neighbourhoods he moved to and from. In the latest Dutch neighbourhood policy (2007) the most deprived and problematic neighbourhoods were selected and were indicated as priority neighbourhoods (Aandachtswijken). These are generally neighbourhoods with a high share of non-western minorities, a low average income, high welfare dependency and an accumulation of social and liveability problems. The policy especially focuses on improving these priority neighbourhoods (VROM-WWI, 2007). In this research, priority neighbourhoods are therefore distinguished as a separate category. Besides priority neighbourhoods we distinguish new housing estates 1 and other neighbourhoods. In all three neighbourhood types there are both newly built dwellings and other (old) dwellings. In this research we have characterised all dwellings built since 1999 as newly built dwellings. In order to analyse mobility patterns we focus on the place of residence of individuals in 1999 and in Individuals that have the exact same address in 2005 as in 1999 are classified as non-movers. Other individuals are classified as movers to and from priority neighbourhoods, new housing estates, other neighbourhoods and/or as movers to newly built dwellings based on their address in 1999 and their address in Housing development in Rotterdam, The Hague and Utrecht The selected cities, Rotterdam, The Hague and Utrecht, differ in the amount and the character of new housing development. In The Hague and Utrecht large scale new housing estates have been developed on greenfield locations. In Rotterdam relatively few dwellings have been built in such new housing estates. Most newly built dwellings in Rotterdam are built in existing neighbourhoods. Extensive restructuring programmes in priority neighbourhoods and other neighbourhoods has resulted in demolition and new housing development in these neighbourhoods. To a lesser extent also in The Hague new houses have been built in restructured old neighbourhoods, while in Utrecht restructuring has not started on a large scale yet. New housing development, both in new housing estates and within 1 New housing estates are defined as neighbourhoods where more than 80% of the dwellings is built between 1999 and 2004 and/or where more than 1000 buildings are built in this period.

6 Mixité : an urban and housing issue? 6 existing neighbourhoods will engender ethnic selective mobility patterns that will influence the neighbourhoods ethnic composition and segregation on city level. Because of the different character of new housing development in the three cities, we also expect different mobility patterns and different developments in segregation. Figure 1. Number of newly built dwellings, by location (dwellings built ) Priority neighbourhoods Other neighbourhoods New housing estates Rotterdam The Hague Utrecht Ethnic residential segregation All cities are to some degree segregated, that is minority groups have different patterns of spatial concentration than natives. There are, however, differences between ethnic groups and between cities in the degree to which they are segregated. Ethnic residential segregation in cities in the Netherlands is lower than in the United States or in British cities, but higher than in many other European cities (Musterd and Van Kempen, 2009). Ethnic residential segregation can be measured using the segregation index. The segregation index is a measure of how even minority groups are spread over the neighbourhoods of a city. It can be interpreted as the share of a minority group that has to move to another neighbourhood in order to achieve an even spread over all neighbourhoods. The segregation index varies between zero (even spread of minorities over the neighbourhoods) and 100 (all minorities clustered in neighbourhoods with no members of the majority group). The segregation index is calculated as follows: In this formula x i is the number of a certain (minority) group in neighbourhood i, X is the total number of the group, y i is the number of all other people in neighbourhood i, and Y is the total number of all other people. When the share of minorities that live in neighbourhood i (x i /X) is equal to the share of all other households that live in neighbourhood i (y i /Y), this neighbourhood does not contribute to the segregation index. Any neighbourhood where minorities are over- or underrepresented compared to other households contributes to the segregation index (Duncan and Duncan, 1955). Ethnic residential segregation in Dutch cities is moderate to low (Boschman, forthcoming). The Hague is the most segregated city with a segregation index in 2005 of 47 per cent. Figure 2 shows the

7 7 Workshop 16: Minority Ethnic Groups and Housing segregation indices of non-western minorities, Turks and Moroccans and Surinamese and Antilleans in 1999 and 2005 in Rotterdam, The Hague and Utrecht. Figure 2. Segregation indices non-western minorities in 1999 and Rotterdam The Hague Utrecht Rotterdam The Hague Utrecht Rotterdam The Hague Utrecht Non-Western Minorities Moroccans and Turks Surinamese and Antilleans Segregation of non-western minorities in Rotterdam has decreased, while The Hague and Utrecht have experienced an increase in segregation. The segregation of Surinamese and Antilleans is lower than the segregation of Turks and Moroccans and has decreased further in all three cities. Segregation of Turks and Moroccans has only decreased in Rotterdam, in The Hague and Utrecht there has been a slight increase in the segregation of this group. Differences in spatial concentration between minority groups can be partly explained by the immigration history of these groups. Surinamese and Antilleans migrated to the Netherlands because of the colonial relation between the countries. The decolonisation of Surinam in 1975 caused a large peak in their immigration. Turks and Moroccans were recruited as guest workers in the ninety sixties and ninety seventies. Later on more people from these countries followed for family reunification and formation. In recent years, asylum migration brought new and other (non-western) minority groups to the Netherlands (Musterd et. al., 1998). Our main aim is to find out if and how new housing development affects ethnic residential segregation. Figure 3 shows the contribution to the development in the segregation index of the different neighbourhood types. The contribution of a neighbourhood to the segregation index depends on the under- or overrepresentation of the minority group in these neighbourhoods and on the number of inhabitants in the neighbourhood. An increase in the under- or overrepresentation or in the relative number of inhabitants in a neighbourhood type will lead to an increase in its contribution to the segregation index. The segregation index of Turks and Moroccans in Rotterdam, for example, decreased from 48,2 per cent in 1999 to 43,6 per cent in 2005, a decrease of 4,6 per cent point, as is shown in figure 2. Figure 3 shows the effects of the population development in the different neighbourhood types on this development in the segregation index. The effect of priority neighbourhood on the segregation index strongly decreased (-4,3 per cent point), also the effect of other neighbourhoods decreased (-0,7 per cent point) and there was only a small increase in the contribution of new housing estates (+0,3 per cent point). Overall Turks and Moroccans have become more evenly spread over the neighbourhoods in Rotterdam.

8 Mixité : an urban and housing issue? 8 The decreasing influence of priority neighbourhoods on the segregation index is explained by the decreasing (relative) population size in these neighbourhoods due to demolition and restructuring, the overrepresentation of Turks and Moroccans in these neighbourhoods did not decrease. In other neighbourhoods the population size did increase but the underrepresentation of Turks and Moroccans decreased and therefore the effect of these neighbourhoods to the segregation index. As there are few dwellings built in new housing estates in Rotterdam, these neighbourhoods only have small influence on the segregation index. In the same way the effect of the population development in the different neighbourhood types in The Hague and Utrecht can be read from figure 3. In Utrecht and The Hague there has been large scale new housing development in new housing estates. Turks and Moroccans are underrepresented in these new housing estates, in The Hague even more than in Utrecht. Therefore the effect of new housing estates on the segregation index has increased. The overrepresentation of Turks and Moroccans in priority neighbourhoods has increased in both The Hague and Utrecht, but in The Hague the relative number of inhabitants has decreased. Therefore in Utrecht the effect of priority neighbourhoods to the segregation index has increased while in The Hague the effect of these neighbourhoods has remained the same. The contribution of other neighbourhoods to the segregation index decreased both in Utrecht and in The Hague, due to the decrease in the (relative) number of inhabitants of these neighbourhoods. Concluding, in Rotterdam the segregation of Turks and Moroccans has decreased, this was mainly caused by the decreasing population size in priority neighbourhoods due to restructuring. In Utrecht and The Hague, especially new housing estates with an underrepresentation of Turks and Moroccans caused an increase in segregation. Segregation of Surinamese and Antilleans has decreased in all three cities. The share of Surinamese and Antilleans in priority neighbourhoods in all three cities has decreased. This has led to a large decrease in the segregation index. The effect of new housing estates will always increase; they had (almost) no inhabitants and therefore no effect on the segregation index in The share of Surinamese and Antilleans in new housing estates in The Hague, however, is almost equal to the city level average, therefore the effect of these neighbourhoods on the segregation index is minimal. In Utrecht, new housing estates indeed have a strong effect on increase of the segregation index, because Surinamese and Antilleans are overrepresented (!) in these neighbourhoods. As large new housing estates are built in The Hague and Utrecht, the share of inhabitants in other housing estates decreased, and therefore the contribution of these neighbourhoods to the segregation index. Overall Surinamese and Antilleans have become more evenly spread over the neighbourhoods in all three cities. Especially in Utrecht the overrepresentation of Surinamese and Antilleans in priority neighbourhoods has decreased which has led to a strong decrease in the segregation index. Figure 3. Contribution of neighbourhood types to the development of the segregation index

9 9 Workshop 16: Minority Ethnic Groups and Housing Surinamese and Antilleans succeed in moving out of segregated neighbourhoods to new housing estates. Especially in Utrecht, where the share of Surinamese and Antilleans in new housing estates is even higher than on average for the city. This has caused a decrease in the segregation of Surinamese and Antilleans. Turks and Moroccans succeed less then native Dutch, Surinamese and Antilleans in moving to new housing estates. Therefore the segregation of Turks and Moroccans has increased in Utrecht and The Hague where large scale new housing estates have been built. In Rotterdam, due to restructuring, inhabitants of priority neighbourhoods, among which many Turks and Moroccans are spread over other neighbourhoods, this has led to a decrease in segregation. Mobility patterns and ethnic residential segregation Above we described the development in the ethnic residential segregation in Rotterdam, The Hague and Utrecht. These developments in the segregation indices are due to population developments in different neighbourhood types. If priority neighbourhoods, the most ethnic concentrated neighbourhoods, succeed in attracting more native Dutch, or if the mainly native Dutch other neighbourhoods attract more minorities, ethnic residential segregation will decrease. New housing development will have major influence on these mobility patterns; every newly built dwelling will attract a household to the neighbourhood its built in, but at the same time also attract a household from its former neighbourhood, thereby influencing the population composition in both neighbourhoods. Mobility patterns from priority neighbourhoods Firstly we will focus on the mobility from priority neighbourhoods. The latest Dutch neighbourhood policy especially focuses on priority neighbourhoods. One of the goals in this policy is to decrease the ethnic concentration in these neighbourhoods. To achieve an ethnic mix in priority neighbourhoods, besides the ethnic composition of the inflow, also the ethnic composition of the outflow is important. Inhabitants of priority neighbourhoods are a very mobile group. About half of them moved to another dwelling in the period between 1999 and A quarter (Rotterdam) to a third (The Hague and Utrecht) of the inhabitants of priority neighbourhoods in 1999 lived in another neighbourhood type by In The Hague and Utrecht 6 per cent of the former inhabitants of priority neighbourhoods had moved to new housing estates by 2005, in Rotterdam this is only 3 per cent. Because of this high population turnover, the population composition in priority neighbourhoods can be heavily affected by selective mobility patterns. Figure 5 shows that the mobility patterns from priority neighbourhoods are selective in terms of ethnicity. Among the group that has left priority neighbourhoods, non-western minorities are underrepresented. The share of native Dutch is especially high among the group that moves to other neighbourhoods or out of the city. Among the group moving from priority neighbourhoods to new housing estates, non-western minorities are only slightly underrepresented in The Hague and Rotterdam and overrepresented in Utrecht. Non-western minorities are overrepresented among the group that in 2005 still lives in priority neighbourhoods in all three cities. Native Dutch more often succeed in leaving priority neighbourhoods, while ethnic minorities are left behind, therefore the concentration of ethnic minorities in priority neighbourhoods and thereby ethnic residential segregation has increased.

10 Mixité : an urban and housing issue? 10 Figure 4. Housing status in 2005 of inhabitants of priority neighbourhoods in % 90% 80% 70% 60% 50% 40% Did not move Moved within or between priority neighbourhoods Moved to new housing estates Moved elsewhere 30% 20% 10% 0% Rotterdam The Hague Utrecht Figure 5. Share of non-western minorities in mobility patterns from priority neighbourhoods. 70% 60% 50% 40% 30% 20% 10% 0% Rotterdam The Hague Utrecht Average percentage of non-western minorities in priority neighbourhoods in 1999 did not move moved within or between priority neighbourhoods moved to new housing estates moved elsewhere Ethnic differences in mobility patterns from priority neighbourhoods Above we showed that there are ethnic selective mobility patterns from priority neighbourhoods. In this paragraph the focus is on differences between the four largest minority ethnic groups in the Netherlands; Turks, Moroccans, Surinamese and Antilleans. Among people that move within or between priority neighbourhoods all four groups are overrepresented. Moroccans are most likely to move to existing dwellings in priority neighbourhoods, or to not move at all; they are probably unable to decisively improve their housing situation. Surinamese disproportionally often move to new housing estates or newly built dwellings in other neighbourhoods. New housing development in the city can attract especially Surinamese from priority neighbourhoods, and thereby reduces the segregation of Surinamese. New housing development in priority neighbourhoods gives successful minorities the opportunity to improve their housing situation without having to leave their neighbourhood with all its (ethnic) facilities and relations. This is probably especially important for Turks, as they do move to newly built dwellings, but mostly within priority neighbourhoods. New housing development in priority

11 11 Workshop 16: Minority Ethnic Groups and Housing neighbourhoods will keep Turks in these neighbourhoods that otherwise might have left to new dwellings elsewhere and therefore increase the residential segregation of Turks. Mobility to newly built dwellings As shown in figure 1, in Rotterdam and to a lesser extent in The Hague and Utrecht new dwellings have been built within existing neighbourhoods. These newly built dwelling have attracted selective mobility patterns in terms of ethnicity to these neighbourhoods. Figure 6 shows the share of nonwestern minorities within existing neighbourhoods and among movers to newly built dwellings in these neighbourhoods. In Utrecht newly built dwellings within priority neighbourhood have attracted a relatively low share of non-western minorities. This has an decreasing influence on the ethnic segregation. In The Hague and Rotterdam, however, on average the share of non-western minorities moving to newly built dwellings in priority neighbourhoods is higher then on average in these neighbourhoods. The share of non-western minorities is, however, especially high among people that move within priority neighbourhoods to newly built dwellings. Among people that move from elsewhere to newly built dwellings in priority neighbourhoods the share of non-western minorities is lower than on average in these neighbourhoods. Newly built dwellings in priority neighbourhoods succeed in attracting higher shares of native Dutch to these neighbourhoods. At the same time they, however, keep ethnic minorities within the neighbourhood that otherwise might have left to new housing estates or other neighbourhoods. Successful ethnic minorities who want to improve their housing situation might otherwise have left to better neighbourhoods. Building new dwellings in priority neighbourhoods keeps these successful ethnic minorities within the priority neighbourhood. Thereby ethnic residential segregation increases, but income segregation decreases (Boschman et. al., 2011) and these successful minorities will probably have a positive influence on the priority neighbourhoods. Newly built dwellings within other existing neighbourhoods in general attract lower shares of nonwestern minorities than on average in these neighbourhoods. This will lead to an even further decrease in the share of non-western minorities in these neighbourhoods and therefore to an increase in segregation. There are differences between Moroccans, Turks, Surinamese and Antilleans in the probability of moving to newly built dwellings. Moroccans and Antilleans move relatively little to newly built dwellings. Turks often move to newly built dwellings in priority neighbourhoods. Because the share of Turks in priority neighbourhoods is already high, building new dwellings in priority neighbourhoods leads to further segregation of Turks. The share of Surinamese moving to newly built dwellings is higher than on average in the destination neighbourhood for both priority neighbourhoods and other neighbourhoods. New dwellings in priority neighbourhoods will induce segregation of Surinamese, while new dwellings in other neighbourhoods can reduce segregation.

12 Mixité : an urban and housing issue? 12 Figure 6. Share of non-western minorities in existing neighbourhoods and moving to newly built dwellings in existing neighbourhoods 80% 70% 60% 50% 40% 30% 20% 10% 0% Rotterdam The Hague Utrecht Rotterdam The Hague Utrecht Priority neighbourhoods Other Neighbourhoods Average share of non-western minorities Share of non-western minorities among movers to newly built dwellings from priority neighbourhoods from elsewhere in the urban region from outside the urban region Conclusions In this paper we compared new housing development, mobility patterns and segregation in three large Dutch cities. In The Hague and Utrecht large scale new housing estates were built, while in Rotterdam most new dwellings are built in restructuring programmes within existing neighbourhoods. These different spatial patterns in new housing development have led to different selective mobility patterns and different development in segregation on city level. Ethnic residential segregation has decreased in Rotterdam, and increased in Utrecht and The Hague. The segregation of Surinamese and Antilleans decreased in all three cities, the segregation of Turks and Moroccans increased in Utrecht and The Hague and decreased in Rotterdam. These developments in ethnic residential segregation can be explained by population development in the different neighbourhoods in the cities. The population development is caused by selective mobility patterns, that are partly caused by restructuring and new housing development. Large scale new housing estates, as are built in The Hague and Utrecht, enlarges the opportunities for both native Dutch and minorities to move out of concentrated priority neighbourhoods (Dwyer, 2007). Native Dutch are, however, overrepresented among the group that succeeds in moving from priority neighbourhoods to new housing estates, while the ethnic minorities are left behind in the priority neighbourhoods. This explains the increase in ethnic residential segregation in the cities The Hague and Utrecht. There are, however, differences between ethnic groups; Surinamese and Antilleans much more than Turks and Moroccans move to new housing estates. The share of this group in new housing estates is almost equal to the city level average. Therefore the segregation of Surinamese and Antilleans decreased in Utrecht and The Hague. Turks and Moroccans are less inclined to leave the concentrated priority neighbourhood than native Dutch, Surinamese and Antilleans, they are more attached to the ethic relations and facilities in those neighbourhoods (Zorlu, 2009).

13 13 Workshop 16: Minority Ethnic Groups and Housing Concluding, new housing estates engender ethnic specific mobility patterns that result in an increase in the segregation of Turks and Moroccans and a decrease in the segregation of Surinamese and Antilleans. In Dutch urban restructuring programmes old dwellings are demolished and new dwellings are built within existing neighbourhood with the aim of creating a more mixed neighbourhood population in terms of ethnicity. Newly built dwellings in priority neighbourhoods succeed in attracting relatively high shares of indigenous Dutch from elsewhere to these neighbourhoods. At the same time, however, they keep mostly non-western minorities within these neighbourhoods, that otherwise might have left the neighbourhood to improve their housing situation. Therefore the effect of restructuring on ethnic residential segregation is mixed. Offering successful minorities housing opportunities within their neighbourhood will increase ethnic residential segregation; income segregation will, however, decrease and these successful minorities will most likely have a positive influence on the neighbourhood. Again there are differences between minority groups. Especially Turks are very likely to move within priority neighbourhoods to newly built dwellings. They are able to and want to improve their housing situation and have a preference for newly built dwellings, without leaving behind their ethnic network and facilities. Therefore newly built dwellings in concentrated priority neighbourhoods will keep especially Turks within these neighbourhoods. A group that otherwise might have left to other neighbourhoods for a next step in their housing career. Restructuring will therefore increase the residential segregation of Turks. Also Surinamese are inclined to move towards newly built dwellings, but they are more likely to move to newly built dwellings in new housing estates or other neighbourhoods. Surinamese are less then Turks attached to their own (ethnic) neighbourhood. Moroccans and Antilleans do not move or move within old dwellings in priority neighbourhoods, probably they do not have the means to decisively improve their housing situation. Concluding, the effect of restructuring and new housing development in existing priority neighbourhoods on ethnic residential segregation is mixed. Although it succeeds in attracting relatively high shares of indigenous Dutch from elsewhere, at the same time, it keeps non-western minorities within these neighbourhoods.

14 Mixité : an urban and housing issue? 14 References Aalbers, M.B. (2006) Who s afraid of red, yellow and green?: geographies of redlining and exclusion in the Netherlands and Italy. Amsterdam: AMIDSt. Alba, R.D. and Logan, J.R. (1991) Variations on two themes; racial and ethnic patterns in the attainment of suburban residence. Demography, 28, 3, pp Bolt, G. (2001) Wooncarrières van Turken en Marokkanen in Ruimtelijk perspectief. Utrecht: KNAW/ FRW. Bolt, G., Hooimeijer, P. and Kempen, R. van (2002) Ethnic segregation in the Netherlands; New patterns new policies? Tijdschrift voor Economische en Sociale Geografie, 93, 2, pp Bolt, G. and Kempen, R. van (2003) Escaping poverty neighbourhoods in the Netherlands. Housing, Theory and Society, 20, 4, pp Bolt, G., Kempen, R. van and Ham, M. van (2008) Minority ethnic groups in the Dutch Housing Market; Spatial segregation, relocation dynamics and housing policy. Urban Studies, 45, 7, pp Boschman, S. (forthcoming) Residential segregation and interethnic contact in the Netherlands. Urban Studies. Boschman, S., Bolt, G., Kempen, R. van and Dam, F. van (2011) The effect of new dwellings on selective migration patterns and residential segregation in the Netherlands. Paper presented at the Conference of the Urban Affairs Association, New Orleans, Brown, L.A. and Moore, E.G. (1970) The intra-urban migration process: a perspective. Geografiska Annaler, 52B, pp Clark, W. A. V. (1991) Residential preferences and Neighbourhood racial segregation; a test of the Schelling Segregation Model. Demography, 28, 1, pp Dagevos, J. (2009) Ruimtelijke concentratie van niet-westerse migranten. Den Haag: SCP. Dwyer, R. E. (2007) Expanding homes and Increasing Inequalities; U.S. Housing Development and the Residential Segregation of the Affluent. Social Problems, 54, 1, pp Feijten, P. and Ham, M. van (2009) Neighbourhood change, reason to leave? Urban Studies, 46, 10, pp Groot, C. de (2004) Zwarte vlucht; de suburbane locatiekeuze van klassieke allochtonen in Amsterdam. Den Haag: VROM. Ham, M. van and Clark, W.A.V. (2009) Neighbourhood mobility in context; household moves and changing neighbourhoods in the Netherlands. Environment and Planning A, 41, pp Kempen, R. van and Priemus, H. (1999) Undivided Cities in the Netherlands; Present Situation and Political Rhetoric. Housing Studies, 14, 5, pp Kullberg, J. (2002) Consumers responses to Choice-based letting Mechanisms. Housing Studies, 17 (4), Kullberg, J. and Nicolaas, H. (2009) Wonen en Wijken. In: Jaarrapport integratie Den Haag: SCP. Kullberg, J., Vervoort, M. and Dagevos, J. (2009) Goede buren kun je niet kopen. Den Haag: SCP. Laan Bouma-Doff, W. van der (2006) Soort zoekt soort. Rooilijn, 3, Logan, J. R., Alba, R. D. and Zwang, W. (2002) Immigrant enclaves and ethnic communities in New York and Los Angeles, American Sociological Review, 67, pp

15 15 Workshop 16: Minority Ethnic Groups and Housing Musterd, S. and Kempen, R. van (2009) Segregation and housing of minority ethnic groups in western European Cities. Tijdschrift voor Economische en Sociale Geografie, 100, 4, pp Musterd, S., Ostendorf, W. and Breebaart, M. (1998) Multi-ethnic metropolis; patterns and policies. Dordrecht: Kluwer Academic Publishers. Park, R.E., Burgess, E.W. and McKenzie, R.D. (1974) The City. Chicago: Chicago University Press. Permentier, M. and Bolt, G. (2006) Woonwensen van allochtonen. Utrecht: Nethur. Putnam, R. D. (2007) E Pluribus Unum. Scandinavian Political Studies, 30, 2, pp Roscigno, V.J., D.L. Karafin en G. Tester (2009) The complexities and processes of Racial Housing Discrimination. Social Problems, 56 (1) Schelling, T.C. (1971) Dynamic Models of Segregation. Journal of Mathematical Sociology, 1, pp South, S.J. and Crowder, K.D. (1998) Leaving the hood; residential mobility between black, white and integrated neighbourhoods. American Sociological Review, 63, pp Til, R. J. van (2008) Verhuisstromen en Verhuismotieven. Delft: ABF Research. VROM-WWI (2007) Actieplan Krachtwijken. Van aandachtswijk naar krachtwijk. Den Haag: VROM-WWI. Wilson, K. L. and Portes, A. (1981) Immigrant enclaves; an analysis of labour market experiences of Cubans in Miami. American Journal of Sociology, 86, pp Wolpert, J. (1966) Behavioural aspects of the decision to migrate. Papers in regional sciences, 22, pp Zorlu, A. (2009) Ethnic differences in spatial mobility; the impact of family ties. Population, Place and Space, 15, pp Zorlu, A. and Latten, J. (2007) Ethnic sorting in the Netherlands. IZA discussion paper. Zorlu, A. and Mulder, C. H. (2008) Initial and subsequent location choices of immigrants to the Netherlands. Regional Studies, 42, 2, pp