Knocking on the Doors of Integration: Swedish Integration Policy and the Production of a National Space

This article contributes to theoretical discussions on how immigrant integration is produced as a part of redefining national policy to local everyday practice, and what this tells us about the society in which the policy is formed. Integration is from this perspective a way to understand imagined social communities, how they are produced, who is considered to belong, and who is not. Document analysis and interviews with immigrants, local politicians, and officials in small- and medium-sized Swedish towns give insights into both what taken-for-granted assumptions the integration policy builds on and reproduces, and what consequences the integration policy have for the persons the policy is intended for. It is shown that the Swedish integration policy is itself a part of the production of a Swedish space as a container with closed doors, where immigrants are not given equal access and possibilities. This is created by putting sameness and difference at the core of the integration policy and by describing integration as an act of entering the Swedish space.


Introduction
Immigrant integration has received increased attention in societal debate and academia, with a political discourse suggesting Bthe failure of integration^as a core challenge in Europe and beyond (e.g., Goodman 2010). Besides anti-immigrant parties and discourses gaining increased scope in many European countries (e.g., Hagelund 2003;van Heerden et al. 2014), what came to be called Bthe refugee crisis^in autumn 2015 put a lot of attention on immigration and integration policy. Sweden was among the countries that received most asylum seekers in 2015-163,000 persons, which is more than ever before (Eurostat 2017;Swedish Migration Agency 2016). However, Sweden has for long been one of the largest receivers of immigrants in the Western world and has since the late 1960s had policies for immigrants and integration. The system for the establishment of those receiving residence permit is today centralized to the national level (see, for example, Lidén et al. 2015;Qvist 2016), but local and regional authorities, as well as NGOs, are still working with issues of integration on a local level. What this integration is supposed to look like, and what measures will lead to that goal, is however not obvious. It does not seem to be obvious what integration really implies, and whom it includes, and integration has been described as a chaotic and contested concept (see Ager and Strang 2008), while traditions, concepts of community, citizenship, nationhood, etc. have developed different Bphilosophies of integration^in different countries (Kaczmarczyk et al. 2015). Even though definitions of integration can be found in policies at the international, national, and local levels, clear criteria for operationalization or measurement for what an integrated society looks like is missing (e.g., Favell 2014; Grip 2010). As a policy, integration is often described as a Btwo-way process^(see, e.g., European Commission 2005), but studies on the practice of integration show that it commonly ends up being a one-sided process of adaptation by the immigrants (e.g., Li 2003;Philips 2010;Schinkel 2013;Schmauch 2011;Schrover and Schinkel 2013). Studies of the practice of integration show that policy documents are often concentrated around key issues like employment, housing, education, and health (Ager and Strang 2008), where the local municipality is the provider of resources, while it is up to the immigrant to use the resources to become integrated. However, in Ager and Strang's (2008) study, immigrants themselves identified belonging-including social bonds, feeling at home and cultural knowledge-as the Bultimate mark^of integration. The place, people in the neighborhood, and local practice therefore become as important as the local practical resources provided.
In this study, spatial dimensions of integration policies are discussed and problematized, with the aim to understand how this dimension shapes integration policy and the local worlds in which immigrants live and act (cf. Smith 2005). The article therefore contributes to theoretical discussions on how integration is produced as a part of redefining national policy to local practice, and what this tells us about the society in which the policy is formed. Integration is from this perspective a way to understand imagined social communities, how they are produced, who is considered to belong, and who is not. Conceptions of immigrants, 1 as well as meanings of integration, are in this study considered neither eternal nor static, but as depending on daily reproduction and renegotiation (cf. Smith 2005). It is also the daily reproduction that gives them their Bnatural, self-evident, taken-for-granted^order (Haldrup et al. 2006, p. 175) and that forms the image of the receiving society. Through daily meetings, conceptions about others take on new forms and meanings, which express themselves in everyday physical meetings, which also produce local spaces and local spatial practices. In these everyday practices, it is also clear that notions of gender play an 1 The concept Bimmigrants^in this study used in a broad sense. I make no distinctions between reasons for leaving the country of origin, or their legal status in Sweden; refugees, asylum seekers, workforce migrants, etc. are all included in the concept. important role. Stereotyped perceptions about immigrant women and men-in relation to perceptions about Swedish gender equality-are in this article shown to be important factors in how local integration policies are formed.
In the following section, integration as concept, policy, and process is discussed, ending up in describing Swedish integration policy and history. After that, empirical data, method, and theoretical departing points are presented, introducing a modified version of Lefebvre's (1991) spatial triad. The three following sections discuss the empirical material of this study, from themes relating to the concepts of Representations of Space, Spatial Practice, Representational Space, Lived Experience, Templates, and Situations. The article ends up in a discussion on the paradoxical space of integration, that is constructed on a discourse of similarity at the same time as assumptions and constructions of difference are a fundamental point of departure for the policy objectives. In this way, the study contributes to the theoretical discussion on the importance of space, through, for example, the use of spatial metaphors, in the everyday practice of integration policy.

The Idea of Integration
Even if integration is a contested concept (Ager and Strang 2008), irrespective of how it is defined, it builds on a division between Binsiders^and Boutsiders,^between those in need of integration and those not. Schrover and Schinkel (2013) argue that behind the idea of integration in the first place lies an understanding of the nation as a container, not only with definable territorial borders to other countries but also with social borders to define who is inside or outside society. This thought is also in line with, e.g., Anderson's (2006) argument that modern nation states are based on the idea of an imagined community. In this lies an implicit exercise of power where the national space is represented with only certain groups of people as belonging and Bconstituted in particular times and places through relations of power already existent in society ( Sharp 1996, p. 103). The construction of national identity is not an equal process, but builds on contrasting oneself against Bothers^(e.g., Benhabib 1997;Sharp 1996;Winter 2011Winter , 2015. In this lies a construction of the self that Bmakes itself invisible precisely by asserting its normalcy, its transparency,^which is in contrast with Bthe marking of others on which its transparency depends^ (Frankenberg 1999, p. 6). The ones counting as insiders are imagined to have an evident community, to which those outside society is supposed to enter, in the political goal of integration.
From this discussion on national identity, insiders, and outsiders, it is not hard to understand that studies on integration show that in practice, integration often becomes a one-sided process aiming for the immigrant to become similar to the resident population (e.g., Li 2003;Schinkel 2013;Schrover and Schinkel 2013). Many states have, in line with this reasoning, shifted away from earlier integration models towards civic integration, which builds on the idea that successful integration rests as much on the immigrant's individual commitment to acquire language, cultural knowledge, and values (often connected to requirements to gain citizenship), as on economic integration (Goodman 2010). It is likely that this shift has affected the discourses of integration, even in those countries (like Sweden) where no formal decisions have been made to introduce, for example, a language test as a condition for citizenship. The ideas of civic integration also strengthen the boundaries between Bus^and Bthem^where a lack of integration can be attributed to Bthem^outside and is not about Bus^inside, as something outside society distanced from something depending on Bus.^The mutual change is reduced to goodwill to solve the problem by transforming the other and pulling them inside (Schinkel 2013).

Swedish Integration Policy, Practice, and Rhetoric
Sweden has over the last century changed from a country of emigration to a country of immigration and has since the late 1960s had policies for immigrants and integration. Before 1968 assimilation as an idea predominated. Swedish integration policy has by Dahlström (2004) been described in four policy models: the universal policy model of the late 1960s, the multicultural model of the 1970s, the selective policy model of the 1980s, and back to the universal policy in the late 1990s. The universal policy model is characterized by universal welfare services intended for the whole population, which since 1968 includes immigrants. In order not to undermine the universal character of the system, everybody needed to be included and Borevi (2014) describes this model furthermore as important for the promotion of the overall national identity of post-war Sweden.
In the 1970s, multicultural goals were introduced. This kind of policy recognizes ethnic and cultural identities and communities, and actively facilitates them through different measures (Borevi 2014;Dahlström 2004;Kaczmarczyk et al. 2015). The introduction of multicultural goals did not mean that the universal policy model was rejected, rather the welfare state universalism was combined with the promotion of cultural diversity (Borevi 2014). From the mid-1970s, equality has been the overriding policy objective in Sweden, and Kaczmarczyk et al. (2015) write that Sweden was a pioneering state in Europe in adopting a Bmulticultural policy^as early as the middle of the 1970s. Many European states followed and approached towards acceptance of cultural and ethnic diversity (Kaczmarczyk et al. 2015;Westerveen and Adam 2018).
The selective policy model in the 1980s is a period of critique towards the multicultural model (Dahlström 2004)-yet not the universal policy model that still made the base for the integration policy. One of the main critiques was that immigrants had been given too much right to choose between Btheir own culture^and Bthe Swedish culture( as if these were two easily definable things). Gender equality was one issue addressed in the debate, with the argument that immigrants with a patriarchal attitude towards women might choose not to adapt to Swedish gender equality norms (Dahlström 2004). The discussions resulted in a governmental bill that withdrew the support for immigrants' maintenance of collective cultural identities, and instead stated that measures should target individuals for their integration into society (Borevi 2014;Government Bill 1985/86, p. 98). Just as many other European countries adopted a multicultural policy approach, this Bdownscaling^of multiculturalism was not a sole Swedish approach either. Similar processes and debates also happened in other countries where multiculturalism had been embraced (Borevi 2014;Westerveen and Adam 2018). Kaczmarczyk et al. (2015) write that the belief that a multicultural approach is positive for integration of immigrants was diminished in the last year's debate in Europe and has reverted to a more socioeconomic conception of integration, with focus on the adaptation of immigrants and their establishment to the labor market (Favell 2014). Labor market establishment has always been more or less central in Swedish integration policy but was reinforced in 2010, when an Bactivation reform^was introduced. In line with the overall European trend, the activation reform had as the main purpose to facilitate labor market establishment and meant transference of power from the municipalities to the state (Brännström et al. 2018;Qvist 2016). Most recently, on a European level, mainstreaming strategy for integration policy has emerged, with a return to a sort of universalism where specific measures for immigrants are replaced with universal ones. An increased Bcolor-blindization^to speak with Westerveen and Adam (2018, p. 3). This can be seen in the changes in the Bactivation reform^that was introduced in 2018, making the legal framework for immigrants establishment more similar to the conditions that apply for other jobseekers, and increased responsibility on the migrant for the establishment (Government Bill 2017, p. 584). In Sweden, the universal policy model is yet still the official approach, and the government states that to provide equal rights, obligations, and possibilities, irrespective of ethnic or cultural background, are the main purpose of the integration policy. To achieve this, universal measures in, for example, the labor market, education, health, and housing, are promoted (Government Offices of Sweden 2017).
The four different policy models describing Swedish integration policy are, however, above all about rhetoric rather than practice, and Dahlström (2004) writes that these rhetorical changes have been made in order to achieve policy legitimacy in relation to dominant trends in public debate. In practice, no new programmes have been established since 1977, and the ones initiated have remained (except support for newspapers and journals, which was cut in 1986) (Dahlström 2004). The Bactivation reform^introduced in 2010 did not either replace the existing governance structure, even though managerial ideas and rhetoric dominated the reform (Qvist 2016). Since the late 1990s, there have been ongoing discussions both in academia and in politics, pointing at the gap between ideology and practice in the integration policy (e.g., Arora-Jonsson 2017; Dahlström 2004;Riksrevisionen 2005). In 1997, the Swedish government stated in a governmental bill that the policy had been unsuccessful and that it had, contrary to its goal, reinforced a division of the population into Bus^and Bthem ( Government Bill 1997/98:16 1997. Several academic studies also show that conceptions of Swedishness and otherness, sameness and difference, permeate Swedish integration politics, debates, and rhetoric (e.g., Arora-Jonsson 2017; Boréus 2006;Mattsson 2001). The activation reform has not implied any change, and Bucken- Knapp et al. (2018) show that immigrants are still stuck in poorly working introduction and language learning programmes.
One way to understand the gap between ideology/rhetoric and practice is to focus on the people working to implement the policy in their everyday work with the task of integration, the Bstreet-level bureaucrats^ (Lipsky 1980). The policy is not only made in decision-making arenas, but also by individual workers in daily meetings with clients. Lipsky (1980) therefore argues that to understand public policy, one must also look into the everyday practice of the street-level workers. Hagelund (2009), in her study on the implementation of integration policy in Norway, shows that Bdiversity workers^often find themselves in situations where the policy do not provide a solution and that they balance between equality and diversity agendas. The implementation of policy, or translation from policy to practice, therefore-and because street-level bureaucrats see themselves as professionals (Lipsky 1980)-rely on the bureaucrats strategies and lived experiences of what is good practice. Even though this article does not focus on the practical output of policy in form of measures, and therefore not on the Bstreetlevel-bureaucrats^everyday practice in their work, the theory brings an important perspective on policy. The policy is in this study understood not solely as political goals and formulations, but also as the everyday practice of policy, produced by both officials and local politicians.

Data and Analytical Approach
The empirical data of this article mainly consists of interviews gathered as part of the research project BResources or Problems? -Participation of Immigrant Women in Organized Activities^2 2004-2006. In the ongoing research project BThe Geography of Integration,^3 I have returned to issues of integration and have found that similar processes are still permeating the local integration policy and practice. Concerning the research questions of spatial dimensions of integration in this study, the integration reform in 2010 has not resulted in any substantial changes and it is the stability in the everyday practices of integration that made me write this text. The Bold^data is not old or irrelevant and is therefore given new life in this text. The former study focused integration policy and practice, the latter homing processes and belonging among immigrants. Both the spatial dimensions and everyday lived experiences appear as important, and that is the focus of this study. Comparisons of Swedish integration policy before and after 2010 have commendable been done by Brännström et.al. (2018) and are not the purpose of this study, nor pointing out the stability in practice. That the spatial dimension of the production of integration policy and lived experiences are similar 10 years later, however, strengthen the arguments and results of the study.
In both studies, interviews with immigrants have been the main resource, to understand the everydayness of integration policy and to work Bfrom what people are experiencing to bring the beyond-their-experience into the scope of ordinary knowledge^ (Smith 2005, p. 221). The interviewed immigrant persons, in a total of 30 persons, 4 were collected through initial contacts with various immigrant associations and through a snowball method. They have origins in different parts of the world, different motives for migrating to Sweden (refugees, migration for work, love, studies, etc.), and they had lived in Sweden between 1 and 40 years at the time of interview. The interviewed in the later study all came to Sweden as refugees and have lived in Sweden for less than 5 years. The interviews were conducted in Swedish or English. To counterbalance hierarchies (researcher/informant, Swedish/immigrant, language, etc.), and also to gain Bmultiple and transpersonal understandings^on my topic (Bedford and Burgess 2001, p. 123), group interviews have been the chosen method when possible. Most often, 2-3 persons have participated in the interview, and individual interviews were also conducted. To understand the produced representations of Sweden, integration and immigrants, the study includes interviews with four politicians and seven officials, 5 selected based on their responsibility and knowledge about the integration policy in their municipality, and also local documents on integration policy and practice. The data collection has been conducted in four medium-sized (between 100 and 125,000 inhabitants in the municipality) and one small city (around 10,000 inhabitants), with the argument that research on these topics so far has tended to focus on larger cities (e.g., Arora-Jonsson 2017; Schmauch and Girtli Nygren 2014).
To be able to catch the everyday production of integration, an analytical tool based on Henri Lefebvre's spatial triad has been worked out; BThe complexity of spatialization^(see Fig. 1).
The bearing thought here is that through our actions, we are constantly interacting with space in that what we do takes place in space and is given meaning in/through space, at the same time, as we influence and recreate the space of which we are part. Representations of Space, according to Lefebvre (1991), relates to what is conceived and is space linked to knowledge, the ordering of production relations, symbols, and codes. I interpret Bthe Swedish space,^and who is included and who is not, as produced through representations of space. Spatial Practice concerns perceived space and include production and reproduction, people and the geographies of society, ongoing actions and perceptions, as well as fixed installations of space such as buildings, urban structures, and the division of places for specific purposes (Lefebvre 1991). Representational Space is lived space and the inhabitants'/users' space. This space passively perceived through associated images and symbols and is thus a dominated space, since representations, or conceptions, is not a thing outside people's doing, but are rather their doings and activities (cf. Smith 2005). At the same, representational space is space with a clear emotional core, including passion, action, dynamics, and a constant now (Lefebvre 1991). Here, we find everyday life, with the conceived Swedish space and stereotype conceptions, as a background for immigrants to navigate in and make resistance to.
Lefebvre's triad, in large, lacks feminist thoughts (e.g., Kipfer et al. 2012;Simonsen 2005;Shields 2004). However, I agree with, e.g., Simonsen (2005) that it still has much to contribute to geographical theory on gender and everyday life. Therefore, and to elucidate the process and physical and individual action, the terms template, situation, and lived experience are in my analysis added to Lefebvre's triad.
In this context, Templates are simplified formulations that can be repeated and are intended to facilitate the creation of different types of patterns: a model or a pattern which routinely mass produces a simplified conception (Grenholm 2005), just like Haldrup et al.'s (2006) banal Orientalism, which is an Beveryday thinking […] naturally appearing daily in the words of politicians, in media coverage and in the way the dominant discourse is circulating in everyday narratives^ (2006, p. 176). It reflects the ordinary and is based on repetition in the actions of the individual. Templates are shaped by representations and spatial practice, and are added to mark the individual person in this process. The term Situation captures something extremely subjective and individual, where the body is a situation in itself at the same time as it is also in situations (Moi 1999). Each situation is an ongoing process and, as such, open to interpretation and choice of position in every new Bnow,^which places the focus on social inception. Even though the situation is always a unique occasion, it is restricted by previous experiences, the presence of others, and other surrounding conditions, and this creates the lived experience (Moi 1999;Nahnfeldt 2006). Lived experience is about interaction with one's surroundings and comprises the individual's collected experiences and subjectivities. Moi (1999, p. 63) writes that Blived experience is […] sedimented over time through my interactions with the world, and thus itself becomes a part of my situatedness.^All three of these terms that are added to Lefebvre's triad are linked to space through the everyday, which takes place in space-represented and practiced. The physical body, process, and space are bound together and cannot be separated, just as the various spatialities in Lefebvre's theory cannot be separated but are parts of one and the same system. Even though representations of space, representational space, and lived experience are in focus in this text, they cannot be understood without the context of the other parts of the model. The arrows in the figure symbolize a constant motion and that the different parts are never static or separated, nor are scale. The processes can be local as well as global, municipal, or national, and influence each other. To quote Smith (2005, p. 2), this is a way to find out Bhow people are putting our world together daily in the local places of our everyday lives and yet somehow constructing a dynamic complex of relations that coordinates our doings translocally.Î argue that to understand the production of integration and how it is experienced, this combination of perspectives is important. The produced Swedish space is filled with templates, situations, lived experiences, and representations, which will be shown by empirical examples.
In the analysis of the empirical material, the figure and its theoretical concepts have been the tool to sort and understand interviews and documents, with focus on the production of both integration policy and practice, and of space. The analysis was carried out through repeated readings of documents and transcribed interviews, and the results produced are therefore based on the material in total, even if a few quotes are chosen as examples of my analysis.

The Production of a Swedish Space
Conceptions and ideas on integration, i.e., who is supposed to be integrated into what are recurring themes in the empirical material of this study. In the local setting, constructions and representations of Sweden as a national space and a space of similarity are central to the process of integration, producing the definable social and territorial borders as Schrover and Schinkel (2013) describe. The local place where the practice of integration policy takes place is hardly mentioned when stating what immigrants should be integrated into. Instead, Sweden and the Swedish are in focus, shaped by thoughts about Sweden as an imagined community (cf. Anderson 2006). Sweden and Bthe Swedish^are described as something fixed, something obvious that does not need to be explained further. It is presented as something that everyone knows and understands the content of, just what characterizes representations: abstraction, dominance, hegemony, and concept (Lefebvre 1991). Examples are the use of terms like BThe Swedish culture,^BThe Swedish society,^BSwedish food,^or BThe Swedish attitude towards gender equality.^Through narratives like BThe Swedish culture,^a taken-for-granted inclusion of some, and exclusion of others takes place. Distinguishing between BSwedes^and Bimmigrants,^those who are integrated and those who are to be integrated, is inevitably a basis for the integration policies studied, and in the production of Swedish space (cf. Schrover and Schinkel 2013). In the process of achieving integration, Bthe others^become important places and individuals, since for the policy sphere to be necessary at all, there must be objects to be integrated. Immigrants are in the empirical material primarily described as different and are placed in a hidden contrast to what is supposedly Swedish. Simplified templates and representations of immigrants also simplify ideas on integration, and preconditions the relationship between officials and the immigrated people they meet (cf. Brännström et al. 2018). Examples of how the image of the immigrant is formed are shown in the following narratives: Most women who come here from all those countries that we are talking about -Iran, Iraq, Afghanistan, Somaliathey have no education, and for the most part they cannot read. I know that [teachers] say, like, they [the women] do not know that the world is round. It is on that level huh. If you show a globe, and you point out Sweden and Somalia, that is the level you must keep it at, eh. (Local councillor, municipality 2, 2005) When we meet the Iranian association and Imams, we ask them BIn what way do you work for women's emancipation and liberation?^But it's a deadlock, they don't fucking understand the question. (Local councillor, municipality 4, 2006) Modernity and modern beliefs become the division between Bthem^and Bus,^and the second quote is also an example of how thoughts about gender relations and gender equality are one of the markers-maybe the most important one-that will deem you modern or non-modern. But religion vs. secularity is also important in this division (Schinkel 2013), and these markers draw the spatial line between those inside and those outside.
The quotes are examples of the reproduction of the modern/non-modern dialectic, which is a part of the construction of templates of immigrants, as well as Swedishness. This kind of rhetoric lies in an overall discourse of positive self-presentation and negative other presentation, where Bour^good actions are emphasized and actions that do not fit in are ignored or described in a fairly abstract and undetailed way (Schrover and Schinkel 2013). And when people talk of gender equality for immigrants, this is rarely connected to the actual (in)equality that exists between men and women in Sweden, which further reinforces the image of Bthem^as unequal and Bus^as equal. It all comes down to a comparison between Btheir reality^and Bour ideals^ (Gressgård and Jacobsen 2003).
This production of a Swedish space is omnipresent in the discourse of integration in this study. At the same time, as the Swedish space dominates the material, I have tried to understand the significance of the local place. The local place is in the material not produced in the same way as the Swedish space, with an imagined community and distinct borders. The local place is instead produced through the lived and perceived and is often described as a multicultural place with a mix of people with different social and cultural backgrounds. It is also described as Bopen^to diversity, which in other studies have been shown to be of great importance for integration (e.g., Philips and Robinson 2015;Platts-Fowler and Robinson 2015). Integration-related problems, connected to the local place by municipal representatives, are foremost residential segregation and unemployment-tangible and practical issues to handle by the municipality. Immigrants often emphasize the significance of the small-/medium-sized town for integration that it is easier to find oneself a context, a situation and a place to belong to in a small town, compared to larger cities. The quote below is an example off this: Here I can go to the central parts of the town, and everybody knows me. Everybody knows who I am and I say hallo to them. We sort of know each other, and I wouldn't like to leave that. It's so nice. (Immigrant, municipality 5, 2017) The quote also touches the importance of being welcomed by civil society, stressed in many interviews with immigrants. The municipal measures of employment, housing, education, and health (Ager and Strang 2008) are not enough for feeling integrated. The local place, a personal contact, people in the neighborhood and local practice of individuals therefore become as important as the local resources provided, for actually creating feelings of home and belonging. Small things in the everyday lived space are what make people feel included in the community, but as we shall see, there are also practices that works excluding. Inclusion in one situation does therefore not mean included in all situations.

To Enter or to Be Let in…
The study shows that on the local arena, there is an unclear perception of what integration policies should achieve-who should be integrated and what the objectives of the policy are. The national integration policy has in the municipalities partly been redefined. Here, equal participation is not the only obvious task. The most recurrent themes, when it comes to questions the goals of integration, are in interviews and documents belonging to society and the right to one's identity and culture. Two different main features also transpire on what integration is about. One feature focuses on immigrants through talking about learning a language, getting a job, and adaptation, in line with the perception of the society as a container that, through integration, transforms outsiders to insiders (Schrover and Schinkel 2013). The other describes integration as a mutual process where the whole society has to be engaged-two different parts melting together and in that process forming a new society; the twoway process formulated by, e.g., the European Commission (2005). Often aware of the vagueness of the term and policy, but without clarifying what integration really is about, interviewed refer to integration as if it had a clear definition and what constitutes proper integration is often taken for granted and not articulated despite underlying differences (cf. Li 2003). This opens up the potential for misunderstandings.
Despite the fact that it is unclear how the national policy should be put into practice and what should be achieved, I have found some recurrent themes. In the interviews, the most common way to talk about integration is the concept to enter, which again leads us back to the spatial idea of society as a container with easily definable borders. It can be about entering the Swedish labor market, the Swedish language or the Swedish system, and society at large. By means of various metaphors with their point of departure in a notion of something spatial, integration is construed as a movement by something Boutside^of Sweden as a space to be entered into (see also Schrover & Schinkel, 2013, p. 1132 for a discussion about the history of using metaphors for describing migration). Being integrated, coming into the community, is compared to entering a room or a house. Society becomes a room you can enter if you have the right Bkeys^or if you are allowed in by one of the Bguardians of integration.Ŵ e still have measures that means that we treat immigrants in a different way, with special information, allowance, projects.
[…] However, when we have managed to let them over the threshold, these measures are not necessary anymore. (Local councillor, municipality 1, 2006) Being allowed to enter means being allowed to become a part of the Swedish community, a move from being non-integrated to becoming integrated and included. The doors are often described as opened by the society for immigrants, which picture integration as an active process of inclusion (which becomes paradoxical in thoughts about civic integration, the immigrants' own commitment to become integrated). Other examples are an official stating that immigrants need to Bhave the right keys^to be able to enter society or an immigrant woman who speaks of integration as open doors that are constantly being shut in front of her.
You can't find a door to enter. There is no inletthat's the fact.
[…] I always say that integration has 500,000 doors, and when I have opened one, and it is not even fully open, it is always being shut in our faces. (Immigrant, municipality 1, 2006) It is a powerful metaphor for the experience of resistance and exclusion in a desire to enter. The metaphor of closed doors also describes the yet-to-be-achieved integration as a result of shutting out. The spatial metaphors create concrete images of a complicated process and turn integration into something tangible. These are key elements of how integration is produced. The Beveryday thinking^and Beveryday narratives^ (Haldrup et al. 2006, p. 176) produce integration as a process of entering, and what is acquired from the objects that should be let in. Representations of Space are the base for the construction of Sweden as a space of similarity and who has access to the space, and who has not.
These metaphors lead again to the division between people who belong and people who do not, the ones inside (Swedes) and the ones outside (immigrants). The perspective that integration is about entering or being let in is even more interesting in relation to the recurring theme in the interviews with immigrants in this study; the need to be oneself and yet be allowed to feel that one belongs to society and to be accepted as different (or not be treated as different). This need is often stated by the informants, but is also said to be lacking in society at large, and described as a choice of either adapting or Bretaining one's culture.^Being oneself and still being accepted and allowed into the community is not seen as a possibility. Therefore, many also adapt in various ways. Yet, some informants state that they have given up since they will never be accepted as BSwedes^anyway, no matter how much they try. Others keep trying by, for example, bleaching their hair to fit in, and many are quite upset by the division of BSwedes^and Bimmigrants^and that not even their children born in Sweden can be accepted as Swedes. The following quote from a woman, who has lived in Sweden for 40 years, is an example of this.

It [seems] it doesn't matter how long I have lived in Sweden.
[…] And they call my daughter an immigrant. What the fuck, immigrant?! She was born in the maternity ward in this municipality, how can she be called an immigrant? (Immigrant, municipality 2, 2005) Lived experience suggests, however, that adapting, whether it is a matter of practice or physical change, does not open the doors to the space of similarity.
Sometimes you get tiered and feel that you don't belong to this country. These feelings will never disappear, so you have to stay strong. (Immigrant, municipality 5, 2017) In other words, the fact that a person has immigrated at some point means that, whatever they do or look like, they will be stamped as different in accordance with the principle Bonce an immigrant, always an immigrant..

and to Get Out
In relation to the recurring theme of entering, there is another common theme which relates only to immigrant women. Integration is not only described as entering, but also as getting out into the community. This fills the concept of integration with the reverse logic of relying on the space that is linked with alienation: the notional isolated home. Examples of this are stating that many immigrant women need Bto get out to meet some people because they have no one to talk to^(Immigrant, municipality 2, 2005), or the municipal initiated project for Somalian women so that they could Bleave their home and come to a meeting place^ (Local councillor, municipality 4, 2006). Here, home and community are two opposed spaces. The division home/community reproduces the classic division between public and private space (e.g., Rose 1993) where the private is linked to women and the public to men. This is also the case with immigrants, where women are described as tied to their homes and in need of help and support to be able to get out. In this divided social landscape, various immigrant/women associations become a means of getting out of the house and Ba second home,^a space somewhere between home and the community. By those active in an association, it is often described as a home, or a place where they feel at home. It is also described by those active as a place where one can be oneself, in contrast to the situation in the rest of society. The association is therefore a Representational Space, dominated by politicians' and officials' expectations of poor, oppressed women finding a way to get out of their homes, but at the same time a lived space where women get together to escape lived experiences of racism and discrimination in the rest of society (see Grip 2012 for a discussion on immigrant women associations as a reinforcing collective action).
To be a part of an association is, however, not equal with entering society. Several of the interviewed politicians and officials express the idea that those immigrant women active in associations are quite Bweak^persons who need support to be integrated into society and that the activities in the associations are more segregating than integrative. One women active in an international women's association describes the relationship with the municipality: They think that we are a burden to the society. They don't want us to be able, unfortunately. Authorities, the municipality, all of them, they believe that … they don't want us to stand on our own feet. (Immigrant, municipality 4, 2005).
But at the same time, to be active in an association is viewed by interviewed politicians and officials as a step to enter society, to reach integration, both by the commitment as such (being a part of an association is seen as something Swedish, a Swedish tradition) and by the act of leaving their homes-which is only mentioned as a goal for women. None of the politicians or officials, however, mentions the importance of the association for creating social bonds, providing a voice and cultural and social activities (cf. Ager and Strang, 2008;Beiser 1993), mentioned by many of the immigrant interviewees.

Conclusion: Spaces of Similarity-Practices of Difference
In this article, I have explored how integration is produced and lived, through templates of immigrants and Swedishness. The Representations of (the Swedish) Space produce a conception of Sweden as a container with an imagined community among those belonging-the insiders. The ones not belongingthe outsiders-are produced with their difference as the main characteristic. In this production, gender and gender relations are important markers of difference, reinforcing the idealized BSwedish^gender equality.
Spatial Practice of integration departs from the lived experience of living in the local place, intertwined with experiences of other places. The size of the city is here stated to be important-expressed by immigrants, politicians, and officials-with closeness to both people and places, which is said to benefit integration. However, this closeness does not mean that the templates are challenged, at least not in the relation between the municipal organization and immigrants. Everyday situations of The Representational Space shows that immigrants are well aware-and constantly reminded-of the templates they are squeezed into and the study shows that immigrants are expected to change themselves to fit in, be similar and become integrated. The one that is too dissimilar will, in metaphorical terms, not be Blet over the threshold,^and the Bkeys^that are mentioned in the interviews can be understood as a kind of BSwedish code,^therefore, keys to the similarity. At the same time, lived space and everyday experiences show that when integration policy is not practiced, the borders do not seem to be that important. In everyday encounters and situations in the local place, people can be just people. The following quote is another example: … where integration has actually happened, it is in sports associations, because they see no difference. Say like a soccer team, they don't care if a person is white or black.
[…] When we have made follow-ups and reports, what have they done? They have done nothing; they have treated everybody the same. That makes me a bit frustrated, because we have put so much focus in the integration policy. (Local councillor, municipality 4, 2006) The purpose of the policy of integration is equal opportunities, rights, and responsibilities. My conclusion from this study is, however, that integration policy is, more than anything else, a practice of difference rather than a space of similarity/equality, which is the purpose of the policy (cf. Hagelund 2009). A division into Bwe^and Bthem^and BSwedes^and Bimmigrants^predominate even among those who oppose this categorization, and become central parts of the production of integration. This points out the importance of highlighting and analyzing representations, characterized by abstraction, domination, hegemony, and concept (Lefebvre 1991). Integration policy is based on the existence of someone to integrate, that is, someone who is different from the majority of the population-the outsider. The field of integration policy therefore both enhances and repudiates the notion of difference and is formulated around a fundamental construction of difference (Tesfahuney and Grip 2007). I have termed this the (dis)similarity paradox. By, on the one hand, defining similarity as the purpose of the policy and, on the other, only identifying and dealing with differences-for the individual who is similar, there is no need for an integration policy-integration policy places itself at the center of this process, at the same time as the borders around national identity and space are actively upheld.