Who’s Homeless and Whose Homeless?

What does the persistent construction of ‘the homeless’ and the revitalised term ‘our homeless’ include, imply, and exclude in Swedish political debate? And how is it politically and morally related to other houseless groups in the country? These questions are approached through an analysis of minutes from the Swedish Parliament 2015–2019. Inspired by Simmel’s (1908/1965) definition of ‘the poor’ as those who get (or would get) public assistance as poor, I claim that in Swedish political discourse, ‘(our) homeless’ comprise only those to whom the society acknowledges a responsibility to give shelter, thereby excluding the tens of thousands of people without homes that are temporarily accommodated by other authorities, private providers or individuals—or not at all. Although official definitions are housing-related, migrants without homes tend to be defined outside the ‘homeless’ concept, as well as from themunicipalities’ responsibilities. I will argue that the reasons for this are institutional: regulations and their interpretation, coupled with traditions to care for only ‘our’ people which, in turn, are fortified by current nationalist sentiments.


Introduction
Many years ago, when I went through old records of poor relief in the 19th century in a little Swedish town I was surprised to see the consistency of the list of recipients. The same widows, disabled men and orphans received alms by the church by Midsummer and Christmas every year. These people were called 'our poor.' When poor relief in 1848 was regulated as an obligation of the secular Poor Board, there was initially some discussion on whether the 'regular poor' would continue to get poor relief, even though they sometimes did not qualify according to the new legislation. Simmel (1908Simmel ( /1965 referred to this kind of obligation as 'moral induction,' i.e., the duty to continue to give alms to someone one has habitually given to previously. Another common distinction in the 19th century concerned local connectedness. At that time, the Poor Board was obliged to provide for the local poor, but not for the 'non-residential' ones, who could be expelled or transferred to their hometowns. This practice was common in all European cities as a means to delimit the municipal costs for the poor and the number of beggars, as well as a measure to keep contagious diseases out, ever since the 16th century (Geremek, 1994). It is today reflected in the European refugee policies, where a residence permit is a minimum requirement for access to housing and shelter.
These considerations tell us something about the weight of institutionalised conceptions of poverty, poor people and obligations to assist them, and I will argue that similar tendencies prevail regarding homelessness today: institutions, comprising both traditions and legalrational definitions, determine homeless people's rights and entitlements and, by extension, who will be seen as homeless.
While the balance and dominance of individual versus structural causes of homelessness have been debated continuously in political fora, in planning for inter-ventions and prevention, as well as in research (see, e.g., Bullen, 2015;Fitzpatrick, 2005;Neale, 1997;Somerville, 2013), the institutional causes and perspectives are less highlighted. These include what kind of housing situations, citizenship, national and local belonging, etc., that qualify a person for being counted, regarded and treated as 'homeless' in a specific local and historical context. This article aims to investigate the meanings and political use of the term 'homeless,' and in particular 'our homeless.' I will argue that even though official definitions of 'homeless' are rational and based on objective housing situations, the selection of people who are actually counted, assisted and included as homeless in the political debate is related to the traditional view of the poor, as well as to current nationalist sentiments.
In his classic essay "The Poor," Georg Simmel (1908Simmel ( /1965) reflects on the definition and position of the poor and concludes that "what makes one poor is not the lack of means. The poor person, sociologically speaking, is the individual who receives assistance because of this lack of means" (p. 140). As such, the poor are related to society in a way similar to the stranger: he is both outside (confronting it), and inside it (a member of the collective). Being confronted and an object of actions by a community, however, is also a kind of relationship, and hence "a particular kind of being inside" (Simmel, 1908(Simmel, /1965. Simmel further highlights that the obligation to provide for the poor does not imply a corresponding right to poor relief. Instead, the Poor Board is accountable to the tax-payers; regulation and rule-bound services are developed to satisfy this constituency, rather than the wanting poor. Accordingly, public assistance is aimed at preventing trouble and unrest, rather than at the poor as individuals. He asks: "Where do the poor belong?" Of course, they may be members of a family, an occupation or a church. "But if they are no more than poor, where do they belong?" (p. 127).
The problem of different definitions of homelessness has attracted substantial research interest over the years (Benjaminsen, Busch-Geertsema, Filipovic Hrast, & Pleace, 2014;Busch-Geertsema, Culhane, & Fitzpatrick, 2016;Jacobs, Kemeny, & Manzi, 2004;Sahlin, 1992) but here, I will primarily focus on their political functions and implications. Here, I will take Simmel's definition and delimitation of the poor as a point of departure when discussing the definition of 'homeless,' as well as the expression 'our homeless' in political debates. What does it mean, when and why is it used, and who is included in or excluded from this term? While this wording might be specific for Sweden, the practical distinctions between homeless groups are probably not (see Baptista, Benjaminsen, & Pleace, 2015).
My empirical data are primarily excerpts from the Swedish Parliament (Sveriges Riksdag, n.d.-a, n.d.-b, n.d.-c) September 2015-December 2019. Parliament debates and committee reports are important data in political and social science research and frequently subjected to discourse analysis (Davidsson, 2010;Kronick & Rousseau, 2015;Verkuyten & Nooitgedagt, 2019). Rightwing and racist discourses, in particular, have been studied in several countries since the 1980s (Fairclough, 2001(Fairclough, , 2003Potter & Wetherell, 1987;van Dijk, 1998). In this study, I make use of a mixture of discourse analytical tools and pay special attention to implicit moral accounts and judgements.
Data and methods used in this article will be further presented in the following section. Who is regarded or referred to as homeless by concerned public authorities in Sweden will be presented next, in Section 3, "Who's Homeless," which includes brief information on other houseless groups of people, who are named and counted differently and targeted by other policies and authorities. The section "Whose homeless?" is about belonging: who-if any one-is ascribed responsibility for accommodating people without homes. The fifth section, "If not Homeless-What?" deals with how houseless groups that are not called homeless are characterised and positioned in the political debate. Apart from the summary, the concluding discussion returns to Simmel's claim that definitions are institutionally determined by obligations, while at the same time reflecting and affecting our thinking and feeling about vulnerable groups and their entitlements.

Data and Method
The Swedish Parliament has 349 members, distributed across eight political parties in proportion to the results of the election that takes place in September every fourth year; the latest were in 2014 and 2018. The Social-Democrats (S) and the Green Party (MP) have formed a minority government since 2014. However, after the 2018 election, another S-MP Government was accepted by the Parliament only after an agreement with the Liberals (L) and the Center Party (C) in January 2019. The new government promised then, inter alia, to abstain from any cooperation with the Left Party (V) or the Sweden Democrats (SD).
In the 2018 election, the extreme-right party SD grew substantially (from 49 to 62 mandates) and has a pivotal role in the assembly. After the last election, it has occasionally scored as the most popular party in opinion polls. However, because of its aggressive xenophobic or 'migration-critical' policy, this party has-as yet-no close allies in the Parliament, although it has approached the two conservative parties: the Moderates (M) and the Christian-Democrats (KD).
The parliament year runs from September to June. For this research, I have studied various kinds of Parliament documents from September 2015 to December 2019, covering 4.5 parliament years and parts of two mandate periods. Most materials for this study were gathered from the website www.riksdagen.se, which contains verbatim transcriptions of all debates, discussions and decisions in the Parliament, as well as motions (i.e., written suggestions to the Parliament by its members), written questions and interpellations to ministers in the Government (and their replies), and reports and opinions from the parliament committees. It is possible to search for specific words in certain kinds of documents for a specific period.
The word 'homeless*' (in Swedish: hemlös*) was found in 324 documents during this time, whereof 70 were chapters in committee reports and 22 propositions from the Government. Of the rest, 80 were minutes from the Parliament, 112 motions and about 30 interpellations and questions to ministers and their replies. I opened all of them and excluded a few as irrelevant, for instance, if 'homeless/ness' referred to consequences of war or natural catastrophes in other countries, runaway cats or notes on postponed debates. I copied the rest, saved them in word files and read through them several times before subjecting them to a more systematic, qualitative analyse. Since I was especially interested in how '(our) homeless' was applied, defined and delimited implicitly or explicitly, I singled out several posts for closer analysis. The aim of qualitative analysis, especially in an explorative study like this one, is not to determine the frequency of a phenomenon or the validity of a claim, but rather to explore varieties and nuances of its meaning. To get an impression of the contexts in which the expression 'our homeless' was used elsewhere, I also searched for the term on Google and in news media archives online.
On riksdagen.se, minutes from Parliament debates are identified with date and number and divided into sections reflecting the topics of the discussion. Each speech, in turn, has a heading with the number of the post, or entry, and the speaker's name and party belonging. The first post of the day gets number 1, and the following ones are numbered continuously, regardless of the topic. References to the quotes presented in this article comprise the number of the post, name and party of the speaker, and the date of the discussion.
As recommended by, for instance, Jørgensen and Phillips (2000), I have formed my package of discourse analytical tools, adjusted to this kind of material and research questions. As I am interested in political distinctions and categorisations of homeless people, I have looked for contrast structures, a concept coined by Dorothy Smith (1978) in a famous article where she showed how the deviant or strange might be constructed through contrasting them to the 'normal' ones in a narrative. In political debates, speakers tend to start by presenting 'facts' to motivate the urge, plea or call for action that ends their posts. Modality-that is, to what extent the speaker/writer seems to agree with and be certain of the factuality of a statement, or presents it as a possibility or an opinion, respectively-is an analytical instrument elaborated by critical discourse analyst Norman Fairclough (1992Fairclough ( , 2001Fairclough ( , 2003. Agencywho is made accountable or credited for a situation, and if any agent at all is implied-is another tool gathered from the same method school (Fairclough, 1992(Fairclough, , 2001. Accounts (Scott & Lyman, 1968) refers to how speakers excuse or justify something they have been (or fear that they will be) accused for. Just like blaming, this is a common device in political debates. Subject positions, i.e., how individuals position themselves and others in relation to each other, to a problem, or to society, is another useful analytical tool (Davies & Harré, 1990;Fairclough, 2001;Laclau & Mouffe, 1985). It is employed by a wide range of discourse analysts, regardless of their main affiliation. For instance, homeless people can be positioned as 'one of us,' as victims of the market, as 'deserving poor,' miserable losers-or as a burden or even a threat to society. Chains of equivalence (Laclau & Mouffe, 1985) or lists (Fairclough, 2001) refer to the repeated enumeration of attributes or phenomena, which shapes the reader's/listener's associations. Since politicians participate in debates to argue for an opinion or an action, these tools can also be used deliberately as rhetorical means.

Who's Homeless?
Municipalities in Sweden are obliged to plan for housing provision for their inhabitants and to see to it that no one within their borders suffers. The individual is entitled to assistance to be assured of a reasonable standard of living. However, there is no social housing in Sweden, no enforceable right to housing, and homelessness is not defined in any legislation.

Official Definitions
Since 1993, the National Board of Health and Welfare (NBHW) counts homeless persons, 18 years or older, a week in April every sixth year. Its current definition of homelessness is partly influenced by the ETHOS definition (see FEANTSA, 2006), which departs from the actual housing situation. Four such situations are defined as 'homeless': "roofless/emergency shelter"; "staying in institutions with no home by discharge"; sublease contract with social services; and private, short-term lodging (NBHW, 2017, p. 7). The mapping comprises persons with citizenship or a residence permit in Sweden who are open cases with the local social services, care institutions or NGOs. By this definition and its explicit elaboration, asylum-seekers, undocumented migrants and "persons from the rest of Europe who reside in the municipality but lack anchorage/rootedness there (EU/ESS-third country citizens)" (NBHW, 2017, p. 13) are excluded.
Comparisons between countries, municipalities, or over time are complicated by the fact that the recorded number of people with a certain problem is related to the number of services offered, resulting in a higher number of homeless people being counted in cities and countries with a great number of shelter places than in places where no or few homeless services are provided-the socalled service statistics paradox (Busch-Geertsema et al., 2016;Habitat, 2000;Tipple & Speak, 2005).
More importantly, the results of mappings and counting are affected by the choice of survey respondents and the ways municipalities are organised. Local and national counts both primarily target persons who are viewed as entitled to assistance to accommodation through social services, and when guidelines for such assistance change, so do the statistics. Today, a common interpretation of the Social Services Act is that the social authorities have no obligation to provide accommodation for adults who are 'only' homeless, or for whom other agencies are responsible. As a result of this institutional delimitation, homeless people counted through the local authorities or NGOs often have additional problems (e.g., with substance abuse or mental health).

Other Houseless Categories
Four large groups of people born abroad are excluded from regular local and national homelessness mappings even if their housing situations would be defined as homelessness according to the NBHW.
The Migration Agency is responsible for asylumseekers' accommodation and subsistence. In January 2020, 40,312 persons were registered in the reception system. Of these 42% were staying in reception centres ('ABO') and 55% in private accommodation ('EBO'; Migration Agency statistics, 2020). The latter group is usually lodging with relatives and acquaintances, often in overcrowded conditions. They get no accommodation allowance and in certain residential areas, they may also be deprived of their allowance for daily subsistence.
Refugees who have been granted residence permits and are registered in a municipality are called 'newly arrived' for the first couple of years (the establishment period). Previous EBO households often remain under similar conditions as when waiting for asylum, while former ABO-residents since 2016 are assigned to municipalities that are expected to settle them permanently and are obliged by law to accommodate them for at least two years. Even if their housing conditions are deficient and insecure, they are not regarded or counted as 'homeless.' Citizens of other EU/EES-countries without work in Sweden are allowed to stay in the country for three months, but no central or local authority is obliged to see to it that 'vulnerable EU-citizens' have accommodation. Since Romania and Bulgaria became members of the EU in 2007, several people from these countries have come to Sweden to beg. Surveys estimate their number to be 4,500-5,000 ("Antalet utsatta EU-migranter," 2019). Although a few charities and NGOs in the big cities, sometimes with limited municipal subsidies, have arranged night shelters and provisional facilities for hygiene, rest, and food for these migrants, they are generally left to themselves, staying in tents, caravans, cars or sheds in circumstances that could be categorised as sleeping rough. They are not entitled to regular social services or shelter.
EU-citizens can be rejected before three months have passed if they have "shown to be an unreasonable burden to the welfare system according to the Social Services Act" (SALAR, 2014, p. 2). Since this group is part of the same European Community but not registered as Swedish inhabitants, its situation is similar to 'the alien or non-residential poor' and the vagrants in the medieval European city-states (Geremek, 1994), who were the first ones to be excluded, punished and sometimes physically expelled when towns and cities regulated begging.
Because of the high number of rejected asylumapplications in recent years, the number of undocumented migrants is assumed to be several tens of thousands. Many are quite well-anchored in municipalities where they may have lived and worked for many years (see, e.g., Holgersson, 2011). Nevertheless, the police are expected to expel people whose asylum applications are rejected or who enter Sweden without the right documents or without applying for asylum.
'Structurally homeless' people, whose lack of homes is appreciated to be caused only by poverty or shortage of affordable housing, are in some cities not eligible for homeless accommodation but may get temporary shelter in emergency cases. The fact that the vast majority of structurally homeless families come from abroad confirms the tendency to exclude non-Swedish people without homes from the 'homeless,' as Simmel would define this word, i.e., from those who (should) get assistance as homeless.

Whose Homeless?
When Simmel asked where the poor belonged, he referred to who was responsible for providing them with support and relief. The formal definitions and categorisations of people without homes are in part related to authorities' official obligations, but in the public debate, boundaries are also drawn between 'ours' and other homeless. The latter may be declared to be somebody else's homeless, but sometimes it is left open if anybody at all is responsible for them.

'Our Homeless'
'Our,' like 'belonging,' has a dual connotation of property and community. It is sometimes aimed at underlining ownership and implies exclusion, as in the following post: People exploit the free movement in the EU to come here and support themselves through begging….It is a very bad solution for all parties involved that poor people from other countries are begging on our streets and outside our shops. (Post 4, Mikael Eskilandersson, SD, 2017, May 17) By emphasising that streets and shops are ours, this speaker indicates that the poor foreigners are transgressing, illegally staying in places that are not theirs.
Being someone's property is associated to protection and care but may also entail subordination, while belong-ing to a group, a 'we,' connotes to inclusion and solidarity. These meanings of 'our homeless' will be highlighted first. Next, I will explore how homelessness in the public debate is being associated with other individual problems and, finally, what functions this expression may fulfil for the political opposition.

Our Protégés and Friends
The expression 'our homeless' is sometimes used by NGOs, especially by charity organisations. In such a context, 'our' has a somewhat paternalistic flavour and connotes to protection, but at times also to nationalism. When blamed for only providing soup and Christmas presents to "Swedish citizens" and for being connected to the extreme right party (SD), the leader and founder of the organisation Vid din sida (By your side), declared that her only focus was to "help our homeless pensioners" (Skoglund, 2019). In recent years, SD, in particular, has used this expression in Parliament debates and statements, as well as in motions. In a debate article, journalist Carl Öström (2016) warns against this party's possible plan to put refugees' and asylum-seekers' needs against the ones of Swedish homeless: Hasse is a homeless Swede and has spent the last night on commuter trains….Naturally, people's indignation about newly arrived refugees' temporary sleeping places upsets Hasse and other homeless people. It is about standard-what is regarded as unacceptable for refugees, is commonplace for the homeless. (p. 6) Note that the author himself distinguishes between 'refugees' and 'homeless.' Later, he argues that it is "about time for other political parties to revise their treatment of our homeless" (Öström, 2016, p. 6).
'Our homeless' often translates into 'homeless Swedes.' A debate article with the heading "Who Engages in Homeless Swedes?" had a similar argument: A generous Swedish refugee policy provides many advantages. We show solidarity and compassion, enrich and internationalise our society, make our population younger and improve our demographics. But if we do not take concrete, effective measures for our own, already vulnerable groups we will add fuel to xenophobia and antagonism grows. (Swärd & Eriksson, 2015) None of these texts argues against better provisions for non-Swedish people without homes, and 'refugees' constitutes one of the 'groups of homeless people' in the last quote, but they are nevertheless discursively excluded from 'our' homeless or groups.

Substance Abuse, Misery-But not Crime
Combinations of attributes that appear so often that one of them gives immediate associations to the others con-stitute 'chains of equivalence' (Laclau & Mouffe, 1985) or 'lists' (Fairclough, 2001). Although the context and syntax may help to define the relations between the words, such as if one property causes, includes or is caused by the other, the nature of this relationship remains obscure while the association is consolidated. Not least in the Parliament debates, 'homelessness' is combined with several other problem terms and qualities representing misery: Because of substance abuse, mental illness, debts, insufficient economy, relational problems etc., about 35,000 people in Sweden are homeless. In the first quote, homelessness is presented as a result of other individual problems, in the second as coexisting or alternating with these problems and in the third as caused by faulty drug policy, but they all underline the association of homelessness with individual problems, especially substance abuse.
Although 'homeless' is often combined with drug abuse, it is not connected to crime. On the contrary, homelessness is sometimes clearly dissociated from criminality. This is from a debate on punishment: Almost all these criminals live on welfare today. None of them is homeless or has to commit a crime to get an income or their daily food. (Post 47, Adam Marttinen, SD, 2016, December 7) There are no homeless criminals today, but these persons [the criminals] have, in most cases, an apartment… (Post 49, Adam Marttinen, SD, 2016, December 7) Note the objective modality-this is presented as facts, not as estimates or personal opinions. By presenting homelessness and criminality as mutually exclusive conditions, this speaker seems to argue that criminals deserve no compassion and, indirectly, that homeless people do.

Homelessness as Signifying Government Failure
In Parliament, 'our homeless' may be used to underline a national duty in contrast to other demands and commit-ments, as in the following post from a debate about the fee to the EU: The number of homeless people is increasing. Mental health is declining among young women. Cancer patients die while waiting for surgery. We have blasts, shootings, executions. Let us secure our own streets before we build roads in other countries. Let's warm our homeless people before sending tax money to the EU. Let's give our elderly food before we fight injustices in Europe. Let's first save Sweden, before we save the world. (Post 26, Dennis Dioukarev, SD, 2019, December 18) While he places himself somewhat above the listeners, like a preacher addressing his parish, this speaker rhetorically lists social problems neglected by the Government to the benefit of the EU and the rest of the world. Our streets, our homeless, our elderly and Sweden are all positioned as something we should care for in the first hand.
The SD's political goal is to stop refugee immigration, which is seen as the cause of domestic homelessness. In the following quote, the same speaker gives his point through comparing the sheer numbers of homeless people (whose difficulties are emphasised with concrete pictures) to the number not of human migrants, but of 'residence permits': We talk about 107,000 residence permits-there are 30,000 homeless people in Sweden. They sleep on park benches and stand in food queues. Where is the justice in this? (Post 45, Dennis Dioukarev, SD, 2018, Homelessness is highlighted as a symptom of government failure by other political parties with other agendas, too. The following excerpt is from a Liberal's argument for deregulated rents and more owner-occupied dwellings: I can guarantee that people are not homeless in Oslo, Copenhagen, Helsinki and Reykjavik….Go for study visits in Oslo, Copenhagen and Helsinki and see for yourself whether people are homeless there. They are not….It is, in practice, ever so easy to get housing in Oslo or Copenhagen. (Post 127, Robert Hanna, L, 2016, March 23) Note the objective modality regarding the non-existence of homelessness and easy access to housing in other countries' capitals ("people are not homeless," "they are not"), which is rather underlined by the subjective additions ("I can guarantee"; "see for yourself").
In summary, '(our) homeless' is generally used for homeless Swedish residents who are mentioned with a certain degree of compassion, sometimes reinforced by combining and integrating the term with other in-dividual problems, resulting in an image of miserable, helpless, maybe old people, worthy of sympathy, pity and help. This impression is further strengthened by the fact that they, unlike other houseless groups, are not associated to-and occasionally clearly dissociated from-criminality. As homelessness is an indisputable problem, the term is also used rhetorically to indicate the Government's failure to protect, care for and house its citizens.

Other Countries' Homeless
So where do the homeless belong, if they are not ours? If 'homeless' only refers to the target groups of the municipal social services, who is responsible for accommodating other people with similar deficient housing situations?
In Parliament, as well as in many public declarations, the responsibility for homeless people from other countries is often explicitly placed on the countries of origin, like in the following statement by the organisation for municipalities and regions, SALAR (2015, p. 2): The individuals who beg in Sweden are a result of the discrimination against Roma people in Romania and Bulgaria. SALAR thinks that international cooperation must be reinforced in order to put pressure on the concerned countries to take responsibility for their citizens.
'Vulnerable EU-citizens' are repeatedly discussed in Parliament, but not because of their homelessness. Rather, they are positioned either as acting subjects (see below) or as victims of their home country's neglect.
Several of the EU member countries do not do enough to support their citizens. Instead, they continue to discriminate against and force citizens-especially Roma people-to lasting exclusion, where the only alternative becomes to go to countries where there is an opportunity to earn a little for themselves or for those who remain at home. (Post 10, Ola Johansson, C, 2017, May 17) Note that homelessness is not mentioned in these quotes, only citizenship, ethnicity, begging and discrimination in the countries of origin. Even though beggars from other EU-countries are indeed homeless when in Sweden, and this situation is described, they are rarely ascribed this attribute. The point made is that Swedish authorities are not responsible for other countries' citizens.

Nobody's Homeless
This section is about two related phenomena: First, the tendency to pass a given responsibility on to other actors or institutions and, secondly, the fact that problems caused by the market tend to be left without any political measure if these imply interventions in property rights. In a neoliberal society, which supports free markets of labour, housing and capital and where the individuals' free choice and responsibility for their situation are underlined, no statutory body is accountable for market failures.

The Dislocated Newly-Arrived
Although municipalities have the ultimate responsibility to see to it that those people who stay within its borders do not suffer and are obliged to receive and accommodate newly-arrived refugees who are assigned to them by the Migration Agency, several municipalities try to pass over these responsibilities to others. Many of them require as a condition for accommodation that the assigned refugees seek and accept housing of all kinds throughout the country, and evict them after two years if they fail. Even before two years have passed, some place their assigned refugees in neighbouring municipalities that are expected to provide accommodation and support, if needed, when the state allowances end.
The news agency TT has reported that several newlyarrived homeless people have been dislocated from Stockholm to municipalities in the north of Sweden ("Dumpad I Kramfors," 2019). In January 2020, the chair of the councils in 113 (of 290) municipalities met with the Minister of Public Administration to complain about the 'social dumping' of unemployed refugees. This tendency was previously highlighted in Parliament through a written question on "Export of socially vulnerable people": Several homeless persons from the Stockholm area claim that they have been forced by various social service administrations to move and to sign leases in, for instance, Hagfors [municipality] if they want continued support from the social services….The pattern is that one moves from larger cities at the end of the establishment period to smaller municipalities that already have a tough situation regarding unemployment and integration. What measures do you intend to take to overcome the problem of exporting socially vulnerable people to smaller municipalities? (Written question 2018/19:702 by Mikael Dahlqvist, S) While the social services in Stockholm are positioned as the acting perpetrator, and the minister is urged to act, both the small municipalities and the dislocated refugees are positioned as victims of this city's strategy. Note that the speaker called the latter homeless, but the minister (from the same party) to whom the question was addressed did not. She answered that although it is not acceptable that municipalities pass on their responsibilities to others or force people to move, "the Government is prevented from having views on how a municipality should act in an individual case or giving instructions on how laws and other regulations should be interpreted" (Lena Hallengren, S, 2019, June 12). The minister's account was an excuse-she could not intervene-but at the same time, a justification-these people were not the Government's responsibility.
So, although they have residence permits and are assigned to and registered in a certain municipality, refugees risk deportation to other municipalities. Sometimes neither the state nor any municipality is willing to include them as 'their homeless.'

'Structurally Homeless'
In the last decade 'social homelessness' has been statistically separated from 'structural homelessness.' The causes of the latter are judged to be structural, i.e., shortage of affordable housing, poverty and landlords' requirements. As mentioned above, social services tend to reject the structurally homeless as not eligible for accommodation. In a Parliament debate on child poverty, this problem was highlighted: A problem is that the structural homelessness, which is due only to the lack of housing and money, has increased. Can the Government develop how you intend to work with this? (Post 37, Rasmus Ling, MP, 2018, February 8) The minister answered that "the municipal social services are commissioned to prioritise families with minor children" but that she was "following this issue closely, because it is extremely important that children do not experience this" (Post 38, Åsa Regnér, S, 2018, February 8). Again, the municipalities alone are responsible. 'Homeless children' is not a state issue.
However, the concerned cities' social services, in turn, claim that they are only accountable for the 'socially homeless,' often called 'our target group.' Hence, no statutory body at the local or central level accepts responsibilities for this group-they are nobody's homeless.

If not Homeless-What?
As shown above, many houseless people are discursively excluded from the 'homeless' category, even if their housing situation and inability to improve it fits well with the NBHW definition of homelessness. Paraphrasing Simmel (1908Simmel ( /1965, the homeless are only those who are entitled to support as homeless. But how are the other categories named and characterised, and how are they morally related to (our) homeless?

Rivals and Enemies of Our Homeless
Newly arrived migrants, asylum-seekers and EU-citizens are not only excluded from the 'homeless' concept but sometimes also positioned as antagonists to the "homeless Swedes"-especially by the SD party: Young people…cannot leave their homes due to the shortage of housing…students forced to stay in tents all over the country. Swedish Television reported recently that residents of a nursing home in Värmdö [municipality] had to stay in containers….At the same time, Sweden has in recent years received hundreds of thousands of asylum-seekers, most of whom without real refugee reasons, which is also the main explanation of the current situation….We can see today how municipalities give precedence in the housing queues to so-called newly arrived, while other municipalities choose to offer them single-family houses and owner-occupied flats. In many municipalities, the newly arrived are also offered expensive hotel rooms. This occurs at the same time as the proportion of homeless Swedes increases. The situation is deeply unfair and discriminating….Minister Eriksson, what measures are you prepared to take to improve the situation for the Swedes who are displaced in the housing market? (Written question 2016/17:730, Rickard Jomshoff, SD, 2017 In this post, the blame is primarily put on the municipalities as agents that 'choose to offer' attractive housing and 'expensive hotel rooms' to the newly arrived, but also on the state for letting too many asylum-seekers in. The repeated 'at the same time' underlines both the asymmetry and the correlation between refugee reception and "homeless Swedes." While the state and municipalities are positioned as perpetrators, and asylumseekers and the newly arrived as privileged favourites, "homeless Swedes," young students and nursing home residents are positioned as victims. The image of old, frail people living in "containers" and refugees ("without real refugee reasons") in luxurious housing makes up a 'contrast structure' (Smith, 1978), aiming at underlining the injustice to "homeless Swedes." EU-citizens, too, have been positioned as favoured competitors concerning shelter places in comparison to (Swedish) homeless people, who are again positioned as being discriminated against, as in the following letter to the editor: What about our homeless?
Why are Swedish rough-sleepers treated differently than poor people from Romania/Bulgaria? During the whole of the 2000s (15 years), it has been reported that in Gävle [municipality], there is no room for homeless people in the local shelters, there is no accommodation. But now suddenly places have been arranged for foreign EU-citizens who have travelled here! Now we want to see that Swedish homeless also get accommodation, immediately.
Justice for us Swedes! ("Våra hemlösa då?," 2015) The absence of responsible agents is striking-people are "treated," problems are "reported," "there is no accommodation"-but the words "now suddenly places have been arranged" indicate that somebody is deciding and acting after all, and implicitly, could have acted before. "Swedish rough-sleepers," "homeless people," "Swedish homeless" and "us Swedes" seem interchangeable; all are positioned as disfavoured to the benefit of "foreign EU-citizens," the antagonists that do not belong, but "travelled here." Again, 'homeless' is only used for Swedes, and 'us Swedes' at the end of the letter further underlines the antagonism.
These EU-citizens are occasionally positioned as dangerous enemies to Swedish homeless people, as in the following post: These beggars enter staircases, they beg, they pollute in parks and woods…they attack and steal from our homeless. And the Government just stands watching all this. It's shameful! (Post 45, Kent Ekeroth, SD, 2016, February 4) Here, the victim position of 'our homeless' reinforces the beggars' positions as villains and antagonism between the two groups is established while the Government is positioned as a passive bystander.

Illegal Squatters
Although the last quote is somewhat odd, the general image of EU-citizens as a source of crime and disorder prevails in Parliament. In 2015, the Government appointed a national coordinator to propose measures for this group. However, his final report had no suggestions on housing or accommodation but was rather occupied with eviction measures: The message from the Swedish society should be clear. EU-citizens are welcome here, at the same time Swedish legislation shall be applied. It is prohibited to reside in parks or other public places or on private land. (National Coordinator for Vulnerable EU-Citizens, 2016, p. 9) In the Parliament debates, too, EU-citizens are primarily characterised as intruders and a nuisance for the police to act against: Mr Speaker! Huts, tents and caravans-temporary settlements on private land prevent the owner from using the land. Often the landowner has to clean, sanitise and restore the land afterwards. But whoever settles on someone else's land without permission is guilty of a crime. (Post 1, Caroline Szyber, KD, 2017, These temporary settlements are not associated with homelessness, but with violation of ownership. The word 'homeless' is not found at all in the bill that suggests facilitation of 'removals' of these settlements, and in the 50-page-long Parliament Committee report (Sveriges Riksdag, 2017) discussing the bill, 'homeless' occurs only once-in a critical motion by the Left Party. Eventually, the Parliament accepted new legislation that facilitated eviction of houseless EU-citizens from private and public land.

Criminal Outsiders
If criminals are sometimes compared to homeless people as two mutually exclusive categories, vulnerable EU-citizens are, on the contrary, often associated with various crimes and illegalities. A Parliament debate on an interpellation by a member of the Moderate Party provides an example: The Swedish Prime Minister has said that begging in Sweden shall and must cease….Despite this, the problem with crime coupled to the vagrant EU-migrants continues. In the city and county of Stockholm, it has gone so far that people who work in park cleaning have to carry assault alarms with a direct connection to SOS Alarm… (Interpellation 2016/17:571 by Jesper Skalberg Karlsson, M) No backing or reference is given to the association between begging and park cleaners' assault alarms. In the debate, more crime images were added: I want to highlight the part of begging where criminal actors control vulnerable people's lives, expose them to human trafficking and oppress them for their economic gain….Men, often the head of the family, who force wife and children to beg in Sweden….We must recognise that many activities surrounding begging and beggars are criminal and problematic. (Post 30, Jesper Skalberg Karlsson, M, 2017, June 26) The Minister of Home Affairs replied that "above all, we shall sharpen the enforcement of the law" (Post 31, Anders Ygeman, S, 2017, June 26), and the interpellant added that it would not be sufficient to prohibit begging, since "most of the things these people are occupied with are already illegal" (Post 32, Jesper Skalberg Karlsson, M, 2017, June 26). While the begging women in his previous quote (Post 30, Jesper Skalberg Karlsson, M, 2017, June 26) are positioned as victims of men's trafficking and oppression, they are in Post 32 themselves positioned as criminal actors alongside their husbands. Thus, the equivalence between begging and crime is gradually established.
Undocumented immigrants, too, recur in the Parliament debate as an outsider group, associated with crime. According to the quote below, they need to be more strongly monitored and deported more efficiently to avoid a "permanent shadow society": Sweden currently has no control over the number of people staying illegally in the country….If this is not stopped, it will lead to a permanent shadow society that can be characterised by vulnerability, exploitation and crime….Many choose to lead a life beside the Swedish society. For them, income often comes from illegal work and criminality….What does the Government intend to do to see to it that more of those who have been rejected, return instead of staying illegally in Sweden? (Post 6, Christian Holm Barenfeld, M, 2018, February 8) Although the undocumented migrants are initially presented as victims of exploitation, just like the begging women above, they are primarily positioned as criminal actors themselves: They earn their living through "illegal work and criminality," remain "illegally" and "choose" a life aside. In her answer, the Minister of Labour Market and Establishment confirmed the seriousness and scope of the problem, called it "unacceptable," and declared that "the Government is intensifying its work in terms of resources for the police to enable them to execute expulsions of people" (Post 7, Ylva Johansson, S, 2018, February 8).

Conclusion
The official definition of homelessness in Sweden departs from certain deficient housing situations, but in actual counting, reporting and mapping of the homeless population in Sweden, several groups are explicitly excluded in the survey instructions and/or implicitly excluded since they are not attended to by the social workers who report homeless people in the surveys. Just like Simmel (1908Simmel ( /1965) wrote more than a century ago about the poor, 'the homeless' are in practice delimited to those who get (or should get) assistance because of their housing situation. Asylum-seekers, undocumented migrants, newly-arrived refugees and vulnerable EU-citizens from abroad are not eligible for shelter, either because they are targeted by other institutions, such as the Migration Agency and the police, or because they are defined as the responsibility of their home countries-they are others' homeless, not 'ours.' Just like in 19th century Sweden, the blame and obligations are placed on the jurisdiction in which the problem first emerged. The same kind of reasoning is often applied to asylum-seekers and, outside the Parliament, sometimes even to refugees who have been granted asylum, who may be told to "go home." Also, and despite still being called homeless, many migrant families are defined out from the social services' target groups as 'structurally homeless.' As victims of poverty and a deficient housing market, they are nobody's responsibility.
This institutionally motivated and regulated exclusion of houseless people in Sweden from the category 'homeless' is reflected and reinforced in the political debate. Analyses of questions, replies and debates in the Parliament show that the term 'our homeless' often refers only to Swedish citizens, especially when they are contrasted to migrants. Reflecting the tendency in social services, 'homeless' is often associated with misery and problems like substance abuse and mental illness (although not with morally condemned qualities like criminality and violence).
According to Simmel (1908Simmel ( /1965, 'the poor' are both inside and outside the community, which is obliged to assist them, but being outside is also a kind of membership in the group (cf. 'our' homeless). But-to paraphrase him-if they are only homeless, where do they belong? Simmel concludes that they belong to the "largest effective circle," which "has no other outside it to which to transfer an obligation" (p. 127). In his view, this was the nation-state, but in our time it would rather be the EU. However, this community does not provide housing for its citizens.
While the 'homeless' are embraced with certain compassion and empathy in the discourse, other terms like EU-citizens, beggars, refugees, migrants, newly arrived etc. do not seem to carry the same kind of protective imperative. On the contrary, some of these categories tend to be associated with trafficking, crime, disorder and violation of property rights. When '(our) homeless' are positioned as victims of these other groups' crime, or as put aside or discriminated against to their benefit, the Government is urged to take side through exclusionary measures, e.g., remove them from public and private land, expel them from the country or stop them from entering Sweden. In this respect, 'our homeless' is used as a tool in xenophobic rhetoric.
From this limited, explorative study, it is clear that a new political discourse on homelessness has emerged, which ascribes more weight to ethnic belonging and national origin than to the actual housing situation and needs. This discourse is in line with the different treatment of groups with similar, deficient housing situations in social services and national legislation. Both the institutional change and the new discourse reflect and might reinforce nationalist and xenophobic sentiments. For social research into homelessness, it is an urgent task to follow the development of this discourse in political assemblies at national as well as local levels.