Governing urban diversity through myths of national sameness – a comparative analysis of Denmark and Sweden

Purpose – The purpose of this paper is to explore problematisations of urban diversity in urban and integration policies in Denmark and Sweden; the paper aims to show how such policies express social imaginaries about the self and the other and underlying assumptions of sameness that legitimise diverging ways of managing urban diversity and (re)organising the city. Design/methodology/approach – Inspired by anthropology of policy and post-structural approaches to policy analysis, the authors approach urban and integration policies as cultural texts that are central to the organisation of cities and societies. With a comparative approach, the authors explore how visions of diversity take shape and develop over time in Swedish and Danish policies on urban development and integration. Findings – Swedish policy constructs productiveness as crucial to the imagined national sameness, whereas Danish policy constructs cultural sameness as fundamental to the national self-image. By constructing the figure of “the unproductive”/“the non-Western” as the other, diverging from an imagined sameness, policies for organising the city through removing and “improving” urban diverse others are legitimised. Originality/value – The authors add to previous research by focussing on the construction of the self as crucial in processes of othering and by highlighting how both nationalistic and colour-blind policy discourses construct myths of national sameness, which legitimise the governing of urban diversity. The authors highlight and de-naturalise assumptions and categorisations by showing how problem representations differ over time and between two neighbouring countries.


Introduction
Through the last 30 years, we have seen increasing concerns about ethnic and socio-economic "segregation" and "ghettoization" in media and political debate in Scandinavian countries.
Such concerns illustrate the ways that migration, integration, urban and housing issues are related to policy fields (Gressg ard and Jensen, 2016). This connection is often made a question of managing diversity through urban planning, which represents a universal challenge to the metropolis due to an association between diversity and spatial disorder (Sandercock, 2003).
Visions of urban diversity suggest forms of inclusion and exclusion of otherness in the nation state and reflect national self-images. The construction of the nation as an "imagined community" (Anderson, 1983) plays a central role in the construction of "us" and "them", of who belongs and who does not belong. A white supremacy is often an unconscious yet dominant mode of self-perception associated with national space as a prerequisite for classifying others as undesired (Hage, 2012).
While previous research on urban policy has focussed on the othering of people and places in the urban periphery, selfing, i.e. the construction of a national self, inherent in processes of othering, is still in need of further exploration. By addressing this gap in the literature, we explore and compare Danish and Swedish urban and integration policies with an ethnographic approach and show how imagined national sameness is (re)produced through policy problematisations of urban diversity. Being similar welfare states with different migration and integration policies make the neighbouring countries Denmark and Sweden an interesting case for comparison. Discourses on urban diversity have evolved somewhat differently in Sweden and in Denmark. Comparative research has shown that despite a shared political concern for segregation, strategies differ significantly with regard to the framing of segregation, its causal explanations and proposed policies (Staver et al., 2019).
The purpose of this article is to explore policies' roles in organising cities: how Swedish and Danish urban and integration policies produce problem representations, social imaginaries and categories about self and other. The article starts with an overview of the research field and the methodological and analytical framework, which is followed by a contextualisation of Sweden and Denmark. This is followed by an empirical discussion, where we analyse the problem representations, categorisations and myths concerning urban diversityfirst in Danish policies and then in Swedish policies. Finally, we compare the two countries and discuss our main findings.
Research field and methodological and analytical framework While policies and media representations express concerns about "vulnerable housing areas", "segregation" and "ghettoization" in Sweden and Denmark, research has stressed how suburban housing areas and their residents are constructed as problems and as threats to a coherent city in policy and media discourse (e.g. Ristilammi, 1994;Backvall, 2019;Mukhtar-Landgren, 2016). Discursive shifts in focus on social, ethnic, religious and criminal otherness have been observed (Frandsen and Hansen, 2020;Sernhede et al., 2019). The othering of people and places echoes a colonial power hierarchy, where the others are constructed as outside the societal dynamics and are made the target of interventions to be "improved" to fit an imagined national norm (Dahlstedt et al., 2018;Gr€ unenberg and Freiesleben, 2016).
Research has highlighted how "race" is a social construction that structures local and global power relations and how racialisation in urban contexts refers to discrimination and differentiation of the housing market according to ascribed racial categories (Molina, 2008) and to the use of racial discourses in defining and legitimising social and spatial changes (Mele, 2019). Racialisation thus constitutes a strategy to colonize urban space and legitimise intervention in the name of diversity and social mix (Mele, 2019;Kipfer and Petrunia, 2009). However, processes of othering also imply selfing (Baumann and Gringrich, 2004). Discourses on the migrant other involve a national self as discourses on the other serve as a spatialised antagonistic identity necessary for the construction of an idea about a homogeneous nation (Simonsen, 2016). In the construction of an "us", a normative preference for sameness outlines difference as the undesirable opposite to sameness (Gullestad, 1989;Essed and Goldberg, 2002).
In order to explore social imaginaries about the self and the other, we approach policy as an ethnographic field and policy documents as cultural texts that construct the problem they address and create the categories that they govern, and thus are central to the organization of cities and societies (Bacchi, 2009;Shore and Wright, 2009;Wedel et al., 2005). The article positions itself within an anthropology of policy, which considers public documents as ethnographic subjects that reveal cultural practices and assumptions within political cultures (Sullivan, 2012). Exploring policy documents as ethnographic material is one way to redirect the ethnographers' analytical and empirical gaze away from a traditional focus on the marginalized Other (Wedel et al., 2005, p. 33). Inspired by Hadj Abdou (2019), we consider "immigrant integration" and anti-segregation policies as governance techniques that say more about the self-image of the nation than about those who are targets of integration measures. Hence, discourses on segregation and ghettoization play a crucial role in the governing of urban diversity and in structuring the city by constructing a myth of an "integrated society" as a peaceful whole, which some people and places are considered to be outside of.
An ethnographic approach to policy explores how policy both "says" and "does". This approach overlaps with discourse analysis in emphasising how policy constructs the problem it addresses, yet it further scrutinises the cultural underpinnings of policy, its mobilising metaphors and its underlying ideologies and uses. Such approaches explore how policies express taken-for-granted assumptions and legitimise certain solutions when identifying problems and classifying target groups (Wedel et al., 2005, p. 34). An anthropology of policy approaches policies as mythsthat is, as a charter of assumptions, values, and meanings (Wedel et al., 2005, p. 35) and as cultural texts and tools for classification (Shore and Wright, 2009). This classification involves categorisations of subjects and social technologies that construct problems and their solutions according to moral assumptions and definitions about "the good life" and "a better society" (J€ ohncke et al., 2004). In this article, this approach involves analysing myths or social imaginaries implied in policynamely, ways of imagining social existence, relations to others, dominant expectations, and their underlying normative notions and images (Taylor, 2004, p. 106). Furthermore, this methodological and analytical framework enables us to explore assumptions about the self and the other at stake in policy discourses and constructions of certain grammars of identity and alterity involved in social imaginaries (Baumann and Gingrich, 2004).
Consequently, the chosen framework enables us to investigate the development of policy discourses and how policy constructs categories of subjects and places through "problem representations" (Bacchi, 2009). By combining anthropology of policy with post-colonial concepts, including othering and racialisation as discussed above, this article develops and applies an analytical framework through which similarities and differences between two national discourses can be understood in relation to ideas about the self and the other.

Contextualising Sweden and Denmark
Despite their shared position in Scandinavia, Sweden and Denmark have followed contrasting paths to modernity (Knudsen and Rothstein, 1994). Sweden is comparatively larger and has asserted itself as a progressive and modern country that officially embraces multi-culturalism and colour-blindness (T€ orngren et al., 2018), whereas Denmark is a small country that after decolonization and military defeats became an increasingly closed, nationalistic society (Olwig and Paerreg ard, 2011;Linde-Laursen, 2007). Nevertheless, civic Governing urban diversity in Denmark and Sweden culture in both countries is infused with a prevalent notion of equality as sameness (Gullestad, 1989) that is reflected in the two countries' status as social-democratic universal welfare states. This is also reflected in both countries' efforts to develop public housing as part of their national welfare schemes. Public housing, as a means to reach the goal of "good housing for everyone", was established between 1930 and 1940s. In Denmark, non-profit housing (almene boliger) consists of housing for rent provided by non-profit housing associations (i.e. semi-autonomous bodies, economically subsidised and legally regulated by the state). This form of housing is often referred to as both "public" and "social" (Andersen et al., 2013), as the municipality has the right to allocate 25% of the housing to people with social needs. In Sweden, public housing consists of municipal housing companies (allm€ annyttiga bostadsbolag) that should offer good and affordable housing regardless of residents' resources and backgrounds. Nevertheless, since 2011, Swedish public housing has had to meet the demands of business profit and function on business-like conditions (Grander, 2018). During the 1960s and 1970s, both countries enlarged the public housing sector through large programmes for housing construction. However, as financial circumstances favoured property ownership, residents who had savings moved out of the public housing estates. With increased refugee and family immigration in the 1980s, the share of refugees and foreign born in public housing increased. Since the 1990s, urban and integration policies have expressed a growing concern with the development. Contemporary debates about segregation in Sweden and ghettoization in Denmark mainly revolve around these housing areas dominated by public housing and located in the suburbs of the larger cities. These areas have been subjected to various area-based projects and strategies for mixing tenure types and resident composition in both Sweden and Denmark.
While there are some similarities in housing policies between the two countries, their integration policies apparently diverge. The Swedish integration policy that was adopted in 1975 represented a break with previous assumptions about assimilation and introduced multi-culturalism as an official ideology. With the integration policy in place from 1997, Sweden has turned towards a more universalistic policy focussing on equal rights. However, a self-understanding of diversity as a foundation of Swedish society remains (Jørgensen, 2006). In contrast, Denmark has had an assimilatory approach to immigrants since the 1980s (Hamburger, 1990), which has been intensified with integration policies from 1999, when the Government issued an "integration law" (Integrationsloven). "Integration" became a highly problematised issue in public debates; it was made an objective in Danish policy, meaning assimilation into Danish cultural norms and values.
These contrasts seem to resonate with major overall contrasts between Swedish and Danish state formations. The constitution of the Swedish state as a corporate project with strong public bureaucracies in the administration of benefits contrasts with the ongoing historical struggle against a strong state by farmers and the urban bourgeois in Denmark (Knudsen and Rothstein, 1994). Furthermore, whereas modernity has been described as overdeveloped in Sweden, Denmark appears to have a much more ambiguous position towards modernity and globalisation, including global migration (Linde-Laursen, 2007).
This article's empirical focus is on policy discourses about public management of urban diversity since the 1990s. The empirical material consists of selected policy documents. We first constructed an overview of the policy development within relevant policy fields. From this overview, we made a qualitative selection of central policy documents for our analysis, in which we located patterns of continuity and changefirst, within each country and then between the two countries. For the empirical analysis of Sweden, we mainly relied on three documents: (1) Development and equitya policy for the metropole in the 21st century (Prop. 1997/98:165); (2) Empowerment against exclusion -The governmental strategy for integration (Skrivelse, 2008/09:24) and (3) The long-term governmental strategy to decrease and counter segregation (Regeringskansliet, 2018). For the empirical analysis of Denmark, we mainly focussed on the following four documents: (1) Status report of the Urban Council (Byudvalget, 1995); (2) The government's strategy against ghettoization (Regeringen, 2004); (3) The ghetto back to society (Regeringen, 2010) and (4) One Denmark without parallel societies (Regeringen, 2018b). In the following, we turn to an empirical analysisfirst of Denmark and then of Swedenbefore we enter the comparative analysis.

Governing urban diversity in Denmark
This section explores how problem representations concerning public housing in Denmark have changed during 1994-2018. As we will show, the main shift consists of a re-scaling of problem representations, from focussing on general social problems to specifically targeting migrants from non-Western countries. In this shift, the "non-Western immigrant" is increasingly constructed as the other, opposed to a Danish self.
Problem representations of urban diversity in Danish policy: from "social vulnerability" to "non-Western" "ghetto" Public housing was problematised in public debate for the first time during the 1990s and led to the establishment of the Urban Council (Byudvalget), which was coordinated between the Urban and Housing Affairs and the Ministry of Interior Affairs. The status report from the Urban Council (Byudvalget, 1995) was the first national effort to combat problems in public housing. The report primarily focusses on "vulnerable housing areas" (belastede boligomr ader) or "socially vulnerable housing areas" (socialt belastede boligomr ader) and defines their problems as "coherent and complex problems: higher moving frequencies, high rents, unemployed and people on social assistance, erosion, problems of alcohol and drug abuse, violence, vandalism" (Byudvalget, 1995, p. 2). The categories of people mentioned are "weak groups," "people with social problems," "unemployed," "people on social assistance," "people with drug problems," and "people with "psychological disorders". According to the report, the problems are "first and foremost of a social nature" (Byudvalget, 1995, p. 6). However, the Urban Council does not emphasise any explicit connection between migrants and social problems (see also Frandsen and Hansen, 2020, p. 16). The report's concerns about xenophobia and racism, and the formation of "white schools" reflect that problem representations involve a Danish "us". At this point in time, the referencing to these areas as ghettos is contested: "In Denmark, in the past ten years, a series of housing areas with big problems have arisen, primarily social (problems). In the public debate, they aresomewhat misleadinglycalled ghettos" (Byudvalget, 1995, p. 1).
Following the report, the Urban Council initiated economical, physical and social efforts (including mixing strategies) in "vulnerable housing areas". In the late 1990s, under a social democratic government, migration and integration became important topics on the political agenda. In 2001, under a new liberal government, parts of the former Ministry of Urban and Housing Affairs were fused with the new Ministry of Integration, signalling the connectivity of migration and urban issues. The prime minister's New Year speech of 2004 proclaimed that ghettoization had to be stopped: "Ghettoization leads to violence and crime and confrontation . . . we neither can nor will accept this in Denmark". In effect, a policy against ghettoization was launched in 2004. The Government's strategy against ghettoization (Regeringen, 2004) signals a new legitimacy of the term "ghetto". In the strategy, mixing is motivated to "create a more balanced resident composition" through attracting "resource strong" residents. Further, the strategy contains plans to sell public housing as areas of investment. A crucial initiative of the strategy is to create an indicator system in the shape of a "problem barometer"later known as the "ghetto list"based on (1) adults on social assistance, (2) lack of educational training and (3) skewed moving patterns.

Governing urban diversity in Denmark and Sweden
The policy document The ghetto back to society: A showdown with parallel society in Denmark (Regeringen, 2010) marks a radical reformulation of "ghetto problems". The overall concern is now "cohesion" in terms of "values that hold Denmark together" and ghettos where "Danish values no longer are a backbone" (Regeringen, 2010, p. 5). In this policy, "non-Western immigrants" are seen as the primary problem, which seems to have shifted place with unemployment. This change in problem representation is reflected in the revised definition of the ghetto, where immigrants and their descendants represent the first concern on the list: (1) the amount of immigrants and descendants from non-Western countries exceeds 50%; (2) the amount of residents aged 18-64 years without employment or educational skills exceeds 40%; (3) the amount of persons convicted for violations of criminal laws concerning weapons and narcotics per 10,000 residents exceeds 270 persons (Regeringen, 2010, p. 37).
In 2011, a new social democratic government was elected, accompanied by the appearance of a new Ministry of Urban, Housing and Rural Districts as separate from the Ministry of Integration. Although this government continued the hard line on integration policy, the focus was mainly on poverty, general vulnerability in society, "negative social heritage" and "un-integrated" migrants in, for example, public housing. While the new government abandoned the term ghetto, it apparently followed the approach of the previous government in terms of governance in form and content.
In 2015, the year of the so-called refugee crisis, the liberal right won the national election. In 2016, a new problematisation of "parallel societies", "isolation" and "extremist milieus" in public housing areas appears (National handlingsplan, 2016). In 2017, the prime minister defined public housing areas as "holes on the map of Denmark" (huller i Danmarkskortet). In his 2018 New Year speech, the prime minister once again brought the subject of the ghetto and parallel societies to the political agenda. In March 2018, a governmental paper was launched: One Denmark without parallel societies: No ghettos in 2030 (Regeringen, 2018b). The main problempublic housing where a minimum of 25% of the residents have a "non-Western background"is mentioned right from the start. With this document, the Government launches a new term for public housing that has been on the ghetto list for five years or more: "the hard ghetto". For areas categorised as "hard ghettos", a mix of tenure types through demolition or privatisation is touted as a "solution", thus displacing residents.
In brief, the problem representations in 1994-2018 shift from a concern for public housing developing into vulnerable neighbourhoods and related white flight to a concern for ghettoization. The initial concern involves the relationship between Danes and migrants and issues of xenophobia/racism. Over time, the problem definition shifts to focus on "the ghetto"/ "the hard ghetto"/"parallel societies", where particular "problematic" migrant groups are blamed for this development. The us-them relationship develops into a dichotomic themagainst-us relationship. Thus, the problem is re-scaled and narrowed to certain problematic racial "identities and mentalities," with the category of "non-Western" as a new legal problem definition that racialises problems in public housing.
Constructing problem spaces: "the ghetto" Since 2004, when the Danish Government launched its first strategy against ghettoization, the Government has created an image of "the ghetto" as an isolated, enclaved parallel society. The ghetto is described as "physically isolated from the surrounding society [. . .] can develop into ethnic enclaves and parallel societies without vital economic, social, and cultural contact with the remaining society" (Regeringen, 2004, p. 12). This image is reflected in repeated descriptions of ghettos as "holes on the Danish map" that constitute spaces of otherness. Furthermore, the ghetto is imagined as a place cut off from the "surrounding" Danish society, thereby constituting a potential threat to Danish society, as a dystopia for a scary, unwanted and dangerous place (Stender, 2018). This image is produced in the prime minister's New Year speech in 2010, where he spoke about "balances" that he was afraid would "tip" in ghettos, indicating this would come at the cost of perceived Danishness.
The discourse on the ghetto as a threat to Danishness also appears in the way issues of norms and values enter as a new concern in the Government's first strategy against ghettoization: "The residential areas are platforms for general integration into society and for an increased knowledge about the norms and values that rule here" (Regeringen, 2004, p. 11). In One Denmark without parallel societies: No ghettos in 2030 (Regeringen, 2018b), the ghetto is seen as a contrast and threat to Danish society, which is based on "democratic values like freedom and legal security. Equity and free mindedness. Tolerance and equality" (Regeringen, 2018b, p. 4). A general concern in this document is "cohesion", which is related to the insistence on one Denmark. The document states, "the government will insist that all citizens embrace the possibilities and the values and norms that the Danish society is based on. That ensures one cohesive Denmark without parallel societies" (Regeringen, 2018b, p. 9). Consequently, the problematisation of "the ghetto" involves a Danish "us"; the ghetto is constructed as an antagonistic identity that arouses a discourse on Danish identity, in which the ghetto serves to uphold a Danish "we" that allow for Danish identity to appear as a coherent identity (Simonsen, 2016, p. 2).
Constructing the other: the "non-Western immigrant" The category of the "non-Western immigrant" is gradually established since 2010 as an independent variable in the definition of ghettos, positioning "race" as one of the main focus points in strategies against ghettoization. With the ongoing definitions of the "ghetto list", the category of the "non-Western" becomes a new legal problem category with real consequences. This racialisation becomes more specified, for example, in 2018, when certain nationalities (primarily "Lebanese") are pointed out as problems in "parallel societies" (Regeringen, 2018a, p. 13). Furthermore, in 2017, a political concern about public housing with a majority of residents from "non-Western countries" led to a parliamentary resolution that distinguished between "Danes" and "immigrants and descendants from non-Western countries". This distinction introduces a dichotomy based on a grammar of orientalisation pointing out oppositions between "us" and "them" (Baumann and Gingrich, 2004).
A recurrent concern is the anticipation that "non-Western immigrants" do not participate in the overall society but "lump together in ghetto areas without contact to the surrounding society" (Regeringen, 2018b, p. 5). The spatial trope of making people get out of their homes in the ghetto and into a Danish milieu implies that migrants are passive and non-participatory in society. This movement of getting out (of the ghetto) and into (majority) society is also requested in overall integration efforts to make migrants leave the presumed isolation in their homes to integrate into society.
The myth about "one Denmark" Policy works similarly to mythas a charter conveying assumptions, values and meanings (Wedel et al., 2005, p. 35) that constitute social imaginaries about social existence and relationships (Taylor, 2004). The myth about "One Denmark without parallel societies", as the 2018 strategy against ghettoization is entitled (Regeringen, 2018b), emphasises cultural homogeneity to which diversity is a threat. The concept of integration, which plays an increasingly important part in the problematisation of public housing, is embedded in the specific Danish social imaginary and represents a specific construction of social problems and their solutions and an asymmetrical relationship between majorities and minorities (Rytter, 2018). The political discourse on integration envisions a process where migrants are supposed to become absorbed into Danish society by suppressing their "original" culture and Governing urban diversity in Denmark and Sweden taking on Danish norms and valuesan act of assimilation and disappearance (J€ ohncke, 2011). Thus, the rhetoric on ghettos as holes that are not part of a Danish public space reflects the national assimilation politics, where migrants are supposed to assimilate or to disappear. Consequently, national policy associates diversity with disorder, conceiving diversity as an obstacle to national and social cohesion. Constructions of Danish norms and values position "non-Western immigrants" in ghettos as culturally and morally "deviant" (J€ ohncke, 2011, p. 53; Gr€ unenberg and Freiesleben, 2016). The proposed mixing strategies do not involve concerns over inter-ethnic coexistence; rather, they mainly involve the persistence of Danishness, and thus an unequivocal request that non-Western immigrants mix with majority Danes by taking on their norms and values. Notions of mixing as a tool to promote social cohesion imply that migrants need to mix and cohere with majority Danes. Hence, the proposed mixing strategies reflect notions about right forms of mixing according to which migrants must cross what is perceived as cultural boundaries in order to dissolve into a homogenous Danishness (J€ ohncke, 2011, p. 54). This invokes images of national territorial power over space, and thus a process of colonising space and legitimising interventions (Kipfer and Petrunia, 2009).

Governing urban diversity in Sweden
In this section, we explore the problem representations and classifications of people and places in anti-segregation policy in Sweden over the past 30 years. We highlight how policy developes from focussing on "immigrant-dense areas" in the early 1990s to emphasise "socioeconomic challenges" and "segregation". The problematisation of unemployment has been continuous and has increasingly acted as classificatory device in defining the other as opposed to a Swedish self.
Problem representations of urban diversity in Swedish policyfrom "immigrant density" to "socio-economic challenges" Ethnic segregation entered the policy agenda in the early 1990s, when the Metropolitan Inquiry emphasised an increased coincidence between ethnic and socio-economic segregation (SOU 1990:36). In 1995, a social democratic government (established in 1994, after three years with a right-wing government) initiated the first government-financed, area-based projects under the name "Special interventions in immigrant-dense housing areas" (S€ arskilda insatser i invandrart€ ata bostadsomr aden, also known as Blommansatsningen). The main objective was to increase employment rates among immigrants in "immigrant-dense areas" with high unemployment rates and high dependence on social assistance (Prop. 1994/95:100, p. 2). Thus, immigrant density is central to the problem representation during the early 1990s. In 1997, integration, housing and metropolitan policies take a first step away from their focus on immigrant density to a focus on individualised "vulnerability" and unemployment. The integration policy of 1997 defines integration as concerning everyone and being closely related to mutual respect and participation in societal development (Prop. 1997/98:16, p. 23). Further, it describes Sweden as a multicultural society in which everyone should have "the same rights and possibilities regardless of ethnic and cultural background" (Prop. 1997/98:16, p. 21). Hence, multi-culturalism is combined with colour-blind ideology through a focus on equal treatment and participation (T€ orngren et al., 2018).
The metropolitan policy of 1998 describes the goal as twofold: to create opportunities for long-term sustainable growth and to break the social and ethnic segregation and promote equal living conditions (Prop. 1997/98:165, p. 30). The policy problematises "socially vulnerable housing areas" (socialt utsatta bostadsomr aden) located in the suburbs, highlighting an increasing divide between those with employment and those without and between domestic-and foreign-born (Prop. 1997/98:165, p. 31). Unemployment and (ethnic) segregation is described as threats to the individual, to society, to the metropolis, and to the municipality's economy, stressing how "unemployment can lead to social exclusion and marginalization and create social unrest and conflicts in society, especially if following ethnic lines" (Prop. 1997/98:165, p. 51).
The 1998 metropolitan policy highlights categories of "people having difficulties to gain employment", including "immigrants, youths, single parents, people on long-term sick leave, long-term unemployed, immigrants with insufficient Swedish skills, and foreign-born women" (Prop. 1997/98:165, p. 85). Further, it problematises how unemployed immigrants "miss out" on opportunities for integration and how many of them end up in the same housing areas. Accordingly, the problem is represented to be segregation and lack of integration caused by unemployment.
From 2006, when a conservative-liberal government replaced the social democratic government, the focus changes from vulnerability and segregation to "social exclusion" (utanf€ orskap). The metropolitan policy was cancelled, and urban development became part of the general integration policy. In the integration strategy of 2008, suburbs are described as "areas of social exclusion" (utanf€ orskapsomr aden) (Skrivelse, 2008/09:24, p. 25). The strategy stresses how employment rates are lower among foreign born in these areas and lowest among foreign-born women (Skrivelse, 2008/09:24, p. 18), proclaiming the solution to be general proposals to strengthen the labour market and make everyone "employable" (anst€ allningsbar) (Skrivelse, 2008/09:24, p. 37). The strategy argues that so called integration problems are not actual problems of integration but will be solved through employment (Skrivelse, 2008/09:24, p. 43).
In 2014, the social democrats regained governmental power, supported by the Green Party (Milj€ opartiet). An anti-segregation reform programme was launched in 2016 and was followed by "The long-term governmental strategy to decrease and counter segregation" in 2018 (Regeringskansliet, 2018). The latter strategy aims to improve the situation in the suburbs and to counteract structural issues behind segregation. The strategy problematises societal divides leading to "a fragmented society where injustice grows and growth declines" (Regeringskansliet, 2018, p. 2). Its vision is to decrease divides and create "a safe Sweden that sticks together" (Regeringskansliet, 2018, p. 8). The strategy defines five interrelated areas for interventions: housing, education, labour market, democracy/civil society, and criminality (Regeringskansliet, 2018, p. 22). Moreover, it considers counteracting discrimination on the labour and housing market as one component in decreasing segregation.
Making use of a new vocabulary, "areas with socio-economic challenges", the 2018 strategy describes these as areas characterized by high unemployment and long-term unemployment, low participation on the labour market, high percentage of people on social assistance, insufficient school results, and low democratic participation (Regeringskansliet, 2018, p. 11). "Immigrant density" is no longer mentioned amongst the area characteristics, and ethnic segregation is considered a result of socio-economic segregation. Nevertheless, the strategy addresses groups within the previously used category of "immigrants". It problematises the concentration of "asylum seekers and newly arrived" in "areas with socio-economic challenges" for contributing to overcrowding and delayed integration and for burdening the welfare system of the municipalities (Regeringskansliet, 2018, p. 22). It also problematises low employment rates amongst "women with non-western immigrant background" and lack of participation in sports activities amongst "girls with foreign background".
Constructing problem spaces: "the suburb" By approaching policies as cultural texts, we see how policies construct suburban places as problem spaces through shifting categorisations. While "immigrant density" is no longer Governing urban diversity in Denmark and Sweden emphasised, the same suburbs are continuously reproduced as problem spacesas "outside" of and diverging from an imagined Swedish normality (Dahlstedt et al., 2018). With the vocabulary of social exclusion, "the problem" is narrowed down to only being about "areas of social exclusion" instead of seeing segregation as a problem of urban divides and polarisation in cities, between the wealthy and the poor (Karlsson, 2016). This problematisation indicates a shift from understanding integration as a two-way (or multidirectional) process to focusing on inclusion (etablering) as a one-way process. The 2008 strategy describes immigrants as an "untapped resource" in societal economy, and its vision is to "break the social exclusion" (bryta utanf€ orskapet) through a strengthened "workfare approach" (arbetslinjen) for everyone and, in particular, for newly arrived refugees (Skrivelse, 2008/09:24, p. 37). Thus, the problem is made the responsibility of those living in "areas of social exclusion", who have to include themselves by becoming part of the workforce. The focus on inclusion through employment mirrors the Swedish development from welfare to workfare state and the assumption of the labour market being colour blind and rational (see e.g. T€ orngren et al., 2018).
While still problematising "the suburb", the 2018 strategy's focus on societal divides and socio-economic segregation implies a re-scaling of the problem space as also concerning the city and the nation as a whole. Mixing tenure types, primarily through densification, and "even distribution" of the newly arrived between the municipalities of Sweden are described as ways to change the resident composition to achieve reduced unemployment and societal divides (Regeringskansliet, 2018, p. 51). Thus, governing urban diversity includes strategies for social mixing as part of the organisation of cities.
Constructing the other: "the unproductive" While the categorisations of problem spaces have been shifted and re-scaled, the problematisation of unemployment has been continuous. The 2018 strategy states that the most important factor to decrease socio-economic segregation and societal divides is for people to have a job and provide for themselves (Regeringskansliet, 2018, p. 7). This echoes the goals of the 1998 policy, which described how "one of the most important goals of the metropolitan policy is to increase the employment rate in vulnerable housing areas" (Prop. 1997/98:165, p. 82), and the goals of the 2008 policy, which described employment as the key factor to "break social exclusion" and ethnic segregation (Skrivelse, 2008/09:24). The emphasis on employment and economic growth in metropolitan and integration policies from the 1990s reflects a general neo-liberal turn (Schierup and Alund, 2010), and unemployment becomes what defines the other. Hence, policies produce and problematise the figure of "the unproductive".
The othering of "the unproductive" mirrors a process of selfing and reveals an ideal of the "citizen-as-worker" and the "active citizen", contributing to the welfare state and participating in the workforce, democracy, sport and leisure activities. This social imaginary about the self and the other produces a racialised and gendered problem representation, framing certain groups of migrants as problemsespecially "non-Western women"by problematising their "falling behind" on the labour market. Hence, policies reveal a social imaginary in which urban diversity has to be productive and profitable to be recognised as a constitutive part of Sweden, which reflects how diversity is made a matter of economic growth in the neo-liberal city (Hadj Abdou, 2019).
The myth of "a safe Sweden that sticks together" By exploring categorisations and their underlying assumptions, we see how policies reflect an association between migrants and suburbs as people and places disturbing the national selfimage. Policies express an assumption that "they" have to be "improved" or removed in order to become part of an imagined "us"for their own sake and for the city, nation and society to "stick together". Thus, policy documents (re)produce a myth of "a safe Sweden that sticks together" (Regeringskansliet, 2018, p. 8)a myth based on a "citizen-as-worker" ideal, in which productiveness is crucial for the national self-image of being an equal, colour-blind and integrated society. Constructing migrants as an "untapped resource", Swedish anti-segregation policy reflects a grammar of encompassment (Baumann and Gingrich, 2004). In this hierarchical act of Selfing, the "difference" of the other is subordinated to the "sameness", and the other is co-opted into an "us" through labour market integration.
While research has emphasised structural mechanisms behind ethnic segregation, such as racial and ethnic discrimination on the housing-and labour market (Molina, 2008), this explanation is silenced in policy, not by being ignored but by being given a periphery role in the problematisation (as one reason among others) while emphasising the responsibility of the individual in becoming employable. Hence, the Swedish self-image of productiveness, equality and colour-blindness disguises inequalities and the process of othering and hinders identifying structural racism as the problem and the basis for marginalisation (T€ orngren et al., 2018).
Concluding discussion: comparing Danish and Swedish approaches to urban diversity We shall now turn to the similarities and differences between the problem representations of "segregation" in Sweden and "ghettoization" in Denmark and how urban diversity is managed in the two contexts.
In line with previous research (e.g. Backvall, 2019;Dahlstedt et al., 2018;Molina, 2008), our analysis shows how suburban places in Sweden and Denmark are constructed as problems through varying vocabulary (e.g. "ghettos", "vulnerable housing areas", "immigrant-dense housing areas", "areas of social exclusion" and "areas with socio-economic challenges"). With an ethnographic approach to policies as cultural texts, acting as classificatory devises, we see how problem spaces are constructed as diverging from a Swedish/Danish national average, thus legitimising governing urban diversity through interventions targeting these places.
However, our analysis shows that problem representations concerning urban diversity develop in contrasting ways in Sweden and Denmark, with implications for housing policy and the organisation of cities. In Sweden, problem representations develop from a focus on "immigrant dense areas" to a focus on "socio-economic segregation". Hence, the problem space is scaled up: it shifts from being a metropolitan problem located in the suburbs to being a problem concerning the whole of Sweden and societal divides. In Denmark, we see the opposite development: problem representations change from a concern with general social conditions and unemployment related to "vulnerable social housing" to a problematisation of "non-Western immigrants" and "ghettos". Accordingly, the problem space narrows down in Denmark to "ghetto areas" inhabited by residents racialised as "non-Westerns".
Social mixing is represented as (a part of) the solution in both Sweden and Denmark, which can be seen as an implication of the construction of problem spaces; furthermore, it illustrates the ways problems are defined by their solutions and the morally defined "good life" and "better society" implied (J€ ohncke et al., 2004). Meanwhile, the methods and expected effects of this strategy differ. In Denmark, mixing is expected to lead to increased integration and social cohesion through a strengthening of "Danish values", thereby controlling and removing urban diversity. In Sweden, the expected outcome of mixing (e.g. through densification) is decreased unemployment. Mixing is expected to realise the labour market potential amongst diverse others, to be beneficial for society and to reduce the income gaps.
Thus, the vision of urban diversity differs drastically between the two countries. In the Swedish context, the problematisation of urban diversity creates the figure of "the Governing urban diversity in Denmark and Sweden unproductive" diverse other, with a vision of cultural diversity as a potential resource in societal economy. On the other hand, in Denmark, the figure of the "non-Western" is associated with visions of cultural diversity as a threat that needs to disappear through assimilation to Danish norms and values. The diverging ways of managing urban diversity reflect different grammars of identity and alterity (Baumann and Gingrich, 2004) related to different state formations and paths to modernity. In Sweden, alterity is from the outset seen as a pre-condition, combining multiculturalism and colour-blind ideology. Swedish corporatism requires that all parts contribute to a common good, which implies a state-centred approach that emphasises the welfare state as crucial in creating an integrated and equal society (Borevi, 2017). With a grammar of encompassment, it is necessary to include the other into the labour market and the welfare/ workfare state. Thus, alterity is encompassed and hierarchised in the corporate project of the Swedish state. In Denmark, a grammar of orientalisation explicitly points out oppositions between Danes and "non-Western" migrants. With a society-centred understanding of cultural homogeneity as a pre-condition for social cohesion in the welfare state (Borevi, 2017), migrants constitute parts that are constantly problematised as an expense of the whole. Through social imaginaries of problematic integration, the other is deemed as unwanted, requiring exclusion or complete assimilation. However, the concept of integration as de facto assimilation indicates an idea of incorporating others perceived to be at a lower stage into a national self at a higher stage, and thus a grammar of encompassment. Therefore, the two countries' relationships to Otherness mayin different waysinvolve encompassment as a hierarchical act of appropriating selected forms of Otherness (Baumann and Gingrich, 2004, p. 25) into a national Self.
The discourses on ghettoization and segregation reflect discourses about national identity, as these discourses serves as negations that allow national identity to appear as a fixed and full identity (Simonsen, 2016). Through exploring policies as cultural texts, we see how the diverging problematisations reveal different social imaginaries. While both societies reproduce themselves through preference for sameness and outline difference as the undesirable opposite to sameness (Essed and Goldberg, 2002), what defines the self and the other differs, along with what type of sameness that is required to be recognised as part of the national self. While the Danish construction of the self rests on an idea of social cohesion as dependent on shared cultural identity, the Swedish construction of the self is based on an idea of equality and integration as dependent on productiveness.
Nevertheless, with a postcolonial lens, we see how selfing/othering is hierarchical according to global power relations. Categorisations play a central role in defining the self as superior and legitimising the governing of the diverse others. Through the use of binaries -Swedish/foreign born, Danish/non-Western, employed/unemployed, integrated/excluded, educated/non-educated or man/womanpolicies construct and ascribe meanings to categories. The first in these binaries is obscured as the silent norm of imaginary sameness, whereas the latter is problematised and has to be "improved" (i.e. made employable) or "removed" (through mixing/schemes for dispersal)in Sweden, for societal divides to heal, and in Denmark, for social cohesion to be re-established. Racialisation thus form a constitutive process in the governing of urban diversity and the (re)organisation of cities in both Denmark and Sweden.
By exploring the assumptions of the self and the other in policy categorisations, our ethnographic approach to policy has enabled us to show how urban and integration policy constructs the figure of "the unproductive"/"the non-Western" and a myth of national sameness based on productiveness and cultural homogeneity, respectively. The national policies of both Sweden and Denmark contribute to a process of othering of people and places who are defined as different, thus legitimising measures to "improve" or "remove" diverse others from urban space as a way to include/integrate "them" into a national "us". Accordingly, this article contributes by highlighting how both nationalistic and colour-blind policy discourses reproduce myths of national sameness that legitimise the governing of urban diversity. Through myths of sameness, existing inequalities and different interests within the "imagined community" or "integrated society" are concealed (see also Hadj Abdou, 2019). Racial, gendered and class-based inequalities in society are thus normalised and obscured in the (re)production of imagined national sameness.
Policy is often treated as an unproblematic given without reference to sociocultural contexts. Ethnographic approaches de-mask this framing of policy and introduce a reflexive perspective on policy as both an idea and a process (Wedel et al., 2005, p. 43). By showing how problem representations develop over time and differ between two countries, this article has highlighted and de-naturalised assumptions, categories and ideas behind the organising of diversity in the city. Hence, with this article, we stress the possibility to think differently about urban diversity.