1 Introduction

Within Northern and Western European countries, there seems to be a growing consensus in policy circles that residential segregation of ethnic groups is undesirable. Ethnic segregation is believed to have a negative effect on the integration of minority ethnic groups and to augment the social problems in certain neighbourhoods (Musterd 2003). Despite the consensus on the adverse effects of residential segregation, European countries differ very much in their policies to combat segregation (Harrison et al. 2005a, b). These differences may be attributable to several factors such as the influence that authorities have on the housing market, different views about the main causes of residential segregation and the experience as a country of immigration (the UK, for instance, has been an immigration country much longer than Finland).

While concerns about ethnic segregation are of rather recent date in many European countries, most of them have a long history of combating socioeconomic segregation. In several, creating social mix is one of the principal aims of their housing and urban renewal policy (Veldboer et al. 2002). In Sweden, for instance, social mix has been a policy goal since 1974 (Andersson 2006). In the United Kingdom, creating socially mixed neighbourhoods was already an important policy goal in the 1950s, and since New Labour came into power, ‘balanced’ communities are being strived for again (Goodchild and Cole 2001). In France, social mix is stimulated by eliminating income requirements for social rented houses and by obliging relatively rich municipalities to build more social rented dwellings (Veldboer et al. 2002). In the Netherlands, social mix has been back on the agenda (like it was in the 1950s) since the second half of the 1990s. In line with changes in the Dutch political climate over the last few years, the concern with the concentration of poverty in certain neighbourhoods has given way to anxiety about ethnic concentrations (Bolt and Van Kempen 2008).

This special issue contains papers from five European countries: Finland, Germany, the Netherlands, Sweden, and the UK. Each of the papers is focused on the policy reactions in the respective countries to the increasing ethnic diversity of their cities. The authors discuss the extent to which ethnic concentration is considered a problem there and what explanations for segregation emerge from the policy discourses. Furthermore, they assess the effectiveness of housing policy measures aimed at desegregation. The section in this introduction where the contents of each separate paper are briefly described is preceded by an overview of the various types of desegregation policies in Europe and the US. The introduction concludes with an overall picture of the similarities and differences between policy discourses and practices in the five countries.

2 Desegregation policies

On the basis of the international literature (see e.g., Bolt and Van Kempen 2008; Goetz 2003; Gijsberts and Dagevos 2007) we may distinguish five types of housing policy measures aimed at desegregation: (1) scattered-site programmes, (2) rental subsidy/section 8/housing vouchers, (3) housing allocation procedures, (4) mobility programmes, and (5) housing diversification. Each type will be explained below.

2.1 Scattered-site programmes

Scattered-site programmes are aimed at dispersing public housing by building it in white, non-poor neighbourhoods. This is mainly an issue in the United States, where public housing is very much concentrated in the poorest areas. Only 10% of the public housing can be found outside concentration neighbourhoods of poor and black households (Goetz 2003). As scattered-site programmes are small in scale, the effect on the overall segregation is limited (Galster and Zobel 1998). Nevertheless, it is very difficult to implement these programmes, as municipalities are very reluctant to cooperate. They often fear that scattered-site public housing will have negative effects on the surrounding neighbourhoods, such as a decrease of neighbourhood cohesion and increase of out-mobility (white flight). However, an evaluation of projects in Yonkers (New York) showed that these negative effects did not occur (Briggs et al. 1999).

2.2 Rental subsidies

The idea behind the rental subsidy is to expand the housing options of households with a low income. In the United States it is only possible to receive a rental subsidy (housing voucher) if one lives in the private sector. In the Netherlands, a rental subsidy is also available in the social rented sector (depending on the rent level of the dwelling and the household income). Although rental subsidies were not invented to decrease segregation, they certainly suppress it. Each time the government intends to cut spending on rental subsidies, fierce debates are triggered about the possible negative consequences for residential segregation (Bolt et al. 2002). Unlike rental subsidies, housing vouchers are not linked to the amount of rent a household has to pay. That means that households are free to use their voucher for any kind of dwelling as long as they accept the financial responsibility if they choose to move to an expensive dwelling. Research in the US has shown that the housing voucher system not only increases the choice set of households but also leads to a decline in segregation (Goetz 2003).

2.3 Housing allocation policy

Housing allocation procedures can be put into operation in such a way that ethnic minorities will be more dispersed. Several European cities (like Rotterdam, Birmingham, Berlin and Frankfurt) have implemented quota systems which made it possible to ban the further settlement of ethnic minorities in neighbourhoods where they were already strongly represented (Bolt 2004). Rotterdam introduced the 5% regulation in 1972. In order to preserve the ‘balance’ of the population composition, minority ethnic groups were not allowed to move into neighbourhoods that had already reached a maximum quota of ethnic minorities. This quota was 5% (!), corresponding to the share of minorities in the city as a whole. In this way, Rotterdam tried to avoid tensions between minority groups and the native Dutch. Moreover, spatial dispersal was seen a means to stimulate the integration of ethnic minorities. Due to conflicts with the Constitution, the dispersal policy was abolished soon after it had started.

Recently, Rotterdam again proposed changing the allocation policy to deal with the problematic concentrations in certain neighbourhoods. As it is not legally possible to apply ethnicity as a criterion in allocation policy, the focus is now on low-income groups. In 2004 Rotterdam was allowed to start an experiment whereby low-income groups from outside Rotterdam were not allowed to move into certain deprived neighbourhoods. After the positive evaluation of this experiment, a new regulation (the ‘Rotterdam law’) that included many of the proposals of Rotterdam came into effect. The income criterion was replaced by dependency on welfare benefits. Those who receive unemployment or social welfare benefits are not allowed to settle in certain designated deprived neighbourhoods. As this regulation only applies to people from outside Rotterdam, it seems to have a very limited effect on the income and ethnic mix of the target neighbourhoods, despite the intensive debates about the ‘Rotterdam law’ in the Netherlands (Gijsberts and Dagevos 2007).

In places where dispersal through housing allocation has been enforced for a longer period, no evidence can be found for positive effects in the integration of minorities. On the contrary, dispersal policies have led to isolation from social networks (Arin 1991). Furthermore, ethnic minorities had to wait longer for the allocation of a dwelling compared to native households (Phillips and Karn 1991). Moreover, quota systems have not proven very effective in reducing ethnic segregation, which is partly due to the reluctance of native households to move into ethnic concentration neighbourhoods (Musterd et al. 1998; Münch this issue).

2.4 Mobility programmes

Mobility programmes combine rental subsidies with mobility counselling (Goetz 2003). Moreover, participants are often required to move to a neighbourhood with a low proportion of blacks (Gautreaux Program) or a low proportion of poor households (MTO: Moving to Opportunity Program). From a research perspective, the MTO Program is very interesting as it offers insight into the effect of the neighbourhood context on the social mobility households. Residents of poor neighbourhoods in five cities (Baltimore, Boston, Chicago, Los Angeles and New York) who volunteered to participate were randomly assigned to one of three groups:

  1. 1.

    An experimental group: Families in this group received vouchers that could only be used in low-poverty neighbourhoods and special counselling to assist them to find suitable accommodation.

  2. 2.

    A voucher comparison group: These participants received regular vouchers (or section 8 certificates) with no special geographic restrictions.

  3. 3.

    A control group: These persons did not receive vouchers or special counselling.

The expectation was that the experimental group would make more gains in terms of social mobility as a consequence of the move to a non-poor neighbourhood. Compared to the other two groups in the research, participants belonging to the experimental group indeed turned out to have a higher sense of self-efficacy (Rosenbaum et al. 2002). Moreover, they came to live in better dwellings and safer neighbourhoods, but there is less evidence for the positive effects on their labour market position and level of education (Orr et al. 2003). As mobility programmes are limited in size, the effect on segregation at the city level is rather small. A recent re-analysis of the MTO data revealed that the effect on segregation is actually negligible. In the long run, people in the experimental group are not more likely than people in the voucher control group to end up in a neighbourhood with a low proportion of poor or African Americans. In some cities, the experimental groups did even not significantly differ from the control group in this respect (Clark 2008). This raises the question whether investments in programmes like these can be justified. It would require many individual moves and a lot of money to effectively reduce the concentration of the poor and/or ethnic minorities (Goetz 2003). Johnson et al. (2002) argue that apart from the direct costs that a mobility programme brings about, there are also possible indirect costs for the neighbourhoods of destination and origin.

2.5 Housing diversification

Housing diversification policies are implemented in several European countries, including the Netherlands, Sweden, Denmark, France and the UK (Harrison et al. 2005a; Veldboer et al. 2002). Often the idea is that diversification leads to more social cohesion, more bridging ties between groups, more social mobility options for those with a weak socio-economic position, and a better (social) integration of ethnic minorities. In practice, it is demonstrably possible to get more social mix, but the social returns fall short of expectations. The reason is that there is not so much interaction (bridging ties) between social groups, even in mixed neighbourhoods (Joseph et al. 2007; Kleinhans 2004). That should be no surprise, as certain social categories (e.g., owner-occupiers, people with a high level of education) are not really focused on their neighbourhood for their social activities and interactions (Atkinson and Kintrea 2000; Guest and Wierzbicki 1999). Instead, people mainly interact with their direct neighbours, who tend to have the same social status. In keeping with that, the composition of the housing stock is usually very homogeneous at a very low spatial level, even in mixed neighbourhoods (Atkinson 2005). Finally, as Herbert Gans (1961, p. 176) already emphasized, physical measures can encourage or discourage social contact between neighbours, but a homogeneity of background or a similarity of interests or values is “necessary for this contact to develop into anything more than a polite exchange of greetings. Without such homogeneity, more intensive social relations are not likely to develop and excessive heterogeneity can lead to coolness between neighbours, regardless of their propinquity.”

In a review of the international literature, Gijsberts and Dagevos (2007) show that most of the studies on housing diversification deal with the effects on social cohesion and the quality of life in target neighbourhoods, but that little is known about the effects on income and ethnic segregation. On the basis of an analysis of data on urban restructuring in the Netherlands, Bolt et al. (2009) show that there is only a small effect of housing diversification on the level of ethnic and income segregation. The proportion of ethnic minorities in urban restructuring areas is increasing at a slower pace than in other areas, while lower-income households are decreasing in numbers at a faster pace. It should be mentioned, however, that this effect only occurs when a substantial number of social rented dwellings is replaced by owner-occupied dwellings. Moreover, forced relocation tends to reinforce existing relocation patterns, as forced movers do not seize the opportunity to move to less segregated neighbourhoods.

A general conclusion, which applies to desegregation policies of all types, is that they are not very effective in reducing segregation by income or ethnicity. At best, such measures lead to a limited decrease of residential segregation, but in many cases there is no clear effect all. Besides that, Gijsberts and Dagevos (2007) reveal a shortfall of evaluation research and point out that those evaluation studies that have been done can rarely withstand the test of methodological criticism. A notable exception is the MTO Program, which has been evaluated rigorously, notwithstanding the contradictions between the results of several evaluation studies (Clark 2008).

Of the five types of desegregation policies, only two are common in Western Europe. Scattered-site programmes are characteristic of the US, where they should be understood as a reaction to past discriminatory and segregationist practices in public housing. Rental subsidies, on the other hand, are common in some European countries, but they are not intended as a policy to combat segregation. Mobility programmes, like MTO, would be unthinkable in Western Europe, for it would imply that some poor neighbourhoods have been ‘given up’ and only the lucky few are given the opportunity to escape to richer neighbourhoods (Kearns 2002). Urban policies in Europe are more focused on changing areas in situ through area-based policies (Musterd 2003, p. 626). Consequently, housing diversification is the main desegregation measure in most of the countries foregrounded in this special issue (Musterd 2003, p. 626). In some cases, this is combined with housing allocation policy. Indeed, in several German cities, housing allocation on the basis of ethnicity is even the most substantial type of desegregation policy. The next section gives a short description of each of the papers in this special issue.

3 The papers in this issue

British housing policy has a somewhat distinct history compared to that of other European countries. Britain has been committed to multiculturalism and has consequently aimed to respond to cultural diversity through its housing policy whilst widening minority ethnic housing choices (Harrison et al. 2005b). Next to that, the role of discrimination in producing ethnic concentrations has long been acknowledged in the UK, which has resulted in more extensive anti-discrimination legislation (Harrison et al. 2005a). However, recent shifts in the debate on ethnicity and housing in the UK has led to a new focus on assimilation and social cohesion at the cost of combating racial inequality (Robinson 2005). Against this background, Van Ham and Manly note that there appears to be a contradiction in housing policies. On the one hand, there is a growing consensus that self-segregation tendencies have promoted ethnic polarization. The policy response to that trend is to create socially and ethnically mixed neighbourhoods. On the other hand, choice-based letting (CBL) is introduced to give people more choice in where and how they live. This would potentially strengthen the self-segregation tendencies. Van Ham and Manly show that CBL does indeed increase the level of segregation, which can be attributed to the strong tendency of ethnic minorities to move into ethnic concentration neighbourhoods. This finding is in contrast to previous research, which suggested a decline of segregation as a consequence of CBL.

Dhalmann and Vilkama analyse the policy responses to ethnic segregation in Helsinki. Unlike many other European countries, Finland has a relatively short history as a country of immigration. Policies on immigrant housing and ethnic segregation were initiated as a reaction to the sudden growth in immigration at the beginning of the 1990s. Prevention of socio-economic segregation has long been one of the main objectives of housing policy in Helsinki. Different tenure types have been mixed at the neighbourhood level since the 1970s, which has resulted in a ‘balanced’ city structure. Unique to the Finnish case is that differentiation of tenures has not only taken place at the level of the neighbourhood but was also implemented at the level of the street or apartment block. Socio-economic differences have been exceptionally low, but since the 1990s social and ethnic differentiation has started to increase. The spatial concentration of immigrants has occurred especially within the social housing sector, and more precisely in some ‘multi-problem areas’. The spatial dispersal of immigrant households has been strongly encouraged by the authorities, mainly through the procedures of council housing allocation. At the same time, the municipality of Helsinki seems to be increasingly aware of the limited influence it has on the ethnic makeup of neighbourhoods. The development of ethnic concentrations is mainly explained by the immigrants’ own preference to live close to one another. By emphasising the choice-inflicted concentration of immigrants, housing authorities have lost sight of the alternative causes of segregation. Somali interviewees point out that living in the vicinity of relatives and fellow countrymen, although important, is not by any means the main factor influencing their choice of a location. On the contrary, the Somalis feel they have been channelled into the least attractive housing estates, places that have been refused by others.

Like in Finland, the self-segregation of immigrants is seen as the main cause of ethnic segregation in the German policy discourse. As Münch explains, there is a strong belief that a ‘balanced’ population composition is a necessary pre-condition for the integration of ethnic minorities. Germany differs from the other European countries in the sense that cities have the capacity to directly combat ethnic segregation. Ethnic concentrations are prevented by imposing moving-in bans or quotas on foreigners in certain quarters, despite recent anti-discrimination legislation. Desegregation policies are, however, not very effective in German cities, as the allocation policies of municipalities are targeted at the social rented sector, which is rapidly shrinking. Moreover, native Germans are reluctant to move into neighbourhoods with a large concentration of ethnic minorities.

Van Kempen and Bolt focus on urban restructuring policies in the Netherlands. Since the 1990s the ideal of social mix has made a comeback on the political agenda. The predominance of social rented dwellings in the housing stock of certain urban neighbourhoods is held responsible for the ‘one-sided’ population composition. Therefore, housing diversification is seen as an instrument to create a ‘balanced’ population structure. Over the past few years, public concerns about the increasing ethnic diversity of Dutch cities have sparked a pessimistic debate about crumbling social cohesion and disappointing ethnic minority integration. The neighbourhood is expected to play an essential role in curbing these negative trends. Consequently, social mix is defined more and more in ethnic terms, despite the fact that housing policy measures cannot be directly targeted at ethnic segregation. A document analysis reveals that there are contradictions between the apparent importance that the national government has placed on the pursuit of social cohesion through social mix and the indifference to this concern within plans at the local level. Furthermore, Van Kempen and Bolt show that the supposedly positive effect of social mix on cohesion, which is especially emphasised in national policy documents, falls short of the mark in light of empirical evidence.

On the basis of interviews with public actors, Holmqvist and Bergsten describe the foundations of the Swedish mixing policy. Despite increasing immigration numbers, Swedish housing policy still focuses first and foremost on counteracting socioeconomic segregation. Unique to the Swedish case is the goal to achieve social mix in all neighbourhoods. Thus, tenure mix is prescribed not only for distressed neighbourhoods but for all other neighbourhoods as well. In contrast to Dutch and British policies, housing diversification is implemented without demolition and only through new construction. Ethnic segregation is seen as a consequence of poor integration on the labour market. The logical response to the increasing ethnic segregation has therefore been to implement programmes that lower the unemployment rates among ethnic minorities.

4 Similarities and differences between countries

A comparison of the five countries covered in this special issue reveals that they differ in their identification of the main cause of ethnic segregation as well as in their housing policy responses. Finland and Germany see the residential choices and preferences of ethnic minorities as the driving force behind ethnic segregation. Consequently, their main policy response is to directly address residential choices through the procedures of council housing allocation. There are also clear differences between both countries, as German cities are more inclined to use ethnic criteria in the allocation of housing. In Helsinki, the consensus is that people should have the choice to move where they want.

In the UK policy discourse, self-segregation of ethnic minorities is also seen more and more as the main reason for increasing ethnic concentrations. The most substantial measure to counteract segregation, however, is tenure diversification, even though it affects ethnic segregation only indirectly. Housing diversification is also the main policy response to segregation in the Netherlands. The Dutch discourse identifies the ‘one-sided’ composition of the housing stock as the decisive factor in social and ethnic segregation. Housing diversification also plays a prominent role in Sweden, but without the large amount of demolition that has become common in the Netherlands and the UK. A lack of economic integration is held responsible for ethnic segregation. Consequently, the policy logic in Sweden is that ethnic concentrations can be best counteracted by socioeconomic policies like unemployment programmes.

Despite the differences among the countries in terms of housing policies, there are also quite a few similarities. First of all, they all see ethnic segregation in a very negative light. Possible advantages of ethnic concentration (Bolt and Van Kempen 2008) are almost completely ignored, with the exception of Finland. Even in countries that used to be known for their adherence to multiculturalism, like the UK and the Netherlands, minority ethnic residential segregation is seen as problematic, posing a substantial threat to the integration of minorities and the social cohesion of society. Secondly, the policy discourses on the reasons behind segregation are one-dimensional. Whereas spatial segregation is determined by a large number of factors, many of them are almost completely overlooked, particularly discrimination and self-segregation tendencies among the native majority (Bolt et al. 2008). Thirdly, all countries strive for a ‘balanced’ population structure in urban neighbourhoods, but it is never made clear what exactly constitutes an ideal mix. Only in Germany do the cities indicate specific maximum percentages of foreigners within the framework of the quota policies, but these figures lack any scientific justification. In general, there is a wide contradiction between housing policy discourses and scientific research. There is an unjustified pessimism about the effects of ethnic concentrations and there are unrealistic expectations about the positive social returns of mixing policies. Finally, there is a huge gap between the ambitious policy rhetoric and the limited possibilities to combat segregation. First of all, the demographic reality of increasing numbers of ethnic minorities makes it harder and harder to create ‘balanced’ neighbourhoods, with Helsinki as a possible exception. Besides that, housing policies have a limited effect on ethnic concentration, not only because they contradict each other (as the British paper has shown) but also, and more importantly, because they fail to address the main causes of segregation. In the words of Clark (2008, p. 533), “At the very least, we must be cognizant of the strong forces built into choice and selection, processes that daily make and remake our urban fabric. These forces are often more powerful than our limited ability to intervene.” In Finland, there seems to be an increasing awareness of the limited possibilities to counteract segregation. It is almost an embarrassment that many policy-makers in countries with a much longer immigration history still have to learn that lesson.