Roma street-workers in Uppsala: racialised poverty and super precarious housing conditions in Romania and Sweden

Abstract Using a combination of a political economy approach for the analysis of housing with a postcolonial approach to mobility patterns of the racialised and impoverished Roma in Europe, we reflect on the relationship between racialisation and the precarious living and housing conditions of the Eastern European Roma who move from the poorer to the richer countries of the European Union. Through a qualitative and multi-sited approach to housing, we reveal the situation of permanent displacement for racialised Romanian Roma groups in both Sweden and Romania. We have followed Roma street-workers who come to Sweden for earning income in order to improve their homes in Romania, but once there, they are exposed to homelessness, harsh weather conditions, racism and discrimination. We found that the super precarious conditions of housing in Romania push the migrant Roma into even worse housing and living conditions in the destination country, completing a vicious circle of forced nomadism in which the lack of right to decent housing and permanent risk of displacement are central aspects. This article merges the interests of urban and housing researchers with those from the postcolonial tradition.

between precarious housing conditions and the decision of travelling, as well as to what extent the revenues obtained in Sweden are used for any improvement of the housing conditions in Romania. We found out that precarious housing was a central motive for the population in question when deciding to move. We refer to their migration patterns as forced nomadism, since they aim to earn some money to survive and to keep and improve their homes in Romania. The topics appearing in the analysis of the empirical material are various, but we will concentrate on the narratives on the relationship between housing precarity and circular migration, exposing the importance of the money earned through begging for improving precarious housing situations; after that, we discuss racialised poverty and the effects on Swedish anti-Roma racist attitudes and hostility on the possibility of deriving extra income from begging.

Roma persecution throughout space and time
In this section, we refer roughly to the situation of Roma in the two contexts for our study, and explain the theoretical context for the analysis, based on postcolonial race critical theory. We start from the insight that it would not be possible to approach any issue regarding the Roma population without referring to racial relations and racism. Anti-Roma racism has been studied by scholars for a long time. Wherever the Roma population lives in Europe, their presence is marked by the stigma created during long periods of racialisation and de-territorialisation, including current EU treatment of European Roma as an ethnic minority (Vincze, 2014). On this matter, Vincze indicates that by calling European Roma '"an ethnic minority without a territory" the EU pave the way for the racist argument that Roma are an "inferior race" (unable) to create their own nation-state or to act like traditional ethnic minorities, possessing a mother country of their own' (Vincze, 2014: 444). By stressing ethnicity and cultural difference instead of racial discrimination, the policies of inclusion of the Roma ignore the most important result of racialised stigmatisation, which is social and economic exclusion, the cementation of poverty and the precarisation of life.
In order to understand the precarious and vulnerable situation of Romanian Roma, it is useful to recall Balibar's (2009) observation that the unification of Europe brought about a process of re-racialisation of 'The Roma' (Balibar's quotation marks), i.e. new ways of exclusion brought by EU's both policies and discourses on the Roma when treating and policing them as an ethnic minority, implicitly leaving them outside definitions of European identities and refusing to recognise them as historically belonging to the region. This re-racialisation carries traces of a certain re-definition of European whiteness and long histories of dispossession and exclusion.
But 'The Roma' challenge and resist 'the norms of territorialisation and cultural normalisation' that the European unification has brought about (Yıldız & De Genova, 2018: 428). As we see in this study, they exercise the right to move across national borders occupying low status jobs, or, when that is not an available alternative, begging on the streets, exerting a sort of forced nomadism and confronting racist attitudes from the domestic population (Persdotter, 2019;Hansson, 2019). However, the successively generalised banning of begging in most Schengen countries has resulted in begging only being legally allowed in Sweden, Finland, and Portugal (Hansson & Mitchell, 2018;Hansson, 2019). Nonetheless, this exception could change in the near future because criminalisation of begging has been debated for some years in Sweden (Hansson, 2019), and some municipalities have already started to prohibit begging in public spaces (Henley, 2019).
Building on the theorisation of the relationship between race and class, one current debate within urban and political geography (Roediger, 2017), we bring to the analysis the concept of racialised poverty (van Baar, 2018;Yıldız & De Genova, 2018). Although this paper focuses on Romanian Roma arriving in Sweden as deprived 'EU migrants' today, it is also interesting to stress Sweden's own history of anti-Roma racism. This is important not only in order to understand persisting racialised imaginaries that exist towards this group, but to conceptualise contemporary Swedish anti-Roma racism, in part as a historical legacy, and in part as a result of the ubiquitous re-racialisation of 'The Roma' that Balibar (2009) refers to.
First records on Roma presence in Sweden date from the 1500s and during the 1600s by-laws were issued that entitled landowners to kill Roma trespassing on their lands. In more recent times entry bans were introduced on non-Swedish Roma entering the country. The most recent one dates from 1914 and was only lifted in 1945. For a long time, Roma families were not allowed to settle in any Swedish municipality for longer than two weeks, which put them in a condition of permanent displacement. Children could not be registered at schools, and healthcare, and other welfare provisions were denied to them (Hansson, 2019). Furthermore, forced sterilisations were performed on Swedish Roma in the period 1934-1976, affecting one in four Swedish Roma households under the specified period (SOU, 2016).
The post-war era marked, however, a new period with concrete emancipation attempts. For example, Roma first received access to Sweden's welfare and school system in the sixties (Montesino, 2012) and, after an important activist mobilisation (Mohtadi, 2012), in 2010, the Roma were recognised as one out of five national minorities (together with the Sami, Jewish Swedes, Sweden-Fins and Tornedalers). Romania, where all our respondents come from, also has a centurieslong history of Roma persecution and marginalisation. The abolishment of slavery in 1864 did not conduce to a real emancipation of Roma; in fact, they continued being a marginalised group in society (Boia, 2001). Mobile livelihood strategies became a means to survive and it was a way to avoid conflicts with the majority society or hostile authorities (Bunescu, 2014), forming a context of what we call forced nomadism and displacement.
In urban areas, Roma were often relegated to impoverished settlements, the mahalas (Garfias, 1984;Majuru, 2003). The 1930s and the Second World War brought along the darkest page in European Roma history. Sterilisation programmes, cultural repression, and ultimately deportations and mass execution affected not only Romanian Roma, but Roma across the entire continent at war. The massive housing production and industrialisation attempts of the socialist era did not necessarily lead to dignified inclusion and emancipation of Roma. Although they were offered jobs and became sedentarised, according to Bauerdick (2013) Roma became part of the poor proletariat and were overrepresented in low-skilled labour. Furthermore, illiteracy rates stayed high and Roma continued being housed in inferior dwellings, or settled in plots of rural land, far away from all services, where they built shanty towns (Berescu et al., 2013;Vincze, 2018). This situation was only to be accentuated in the period after 1989, when, especially during the economic growth during the 2000s, social inequalities increased (National Institute of Statistics [INS], 2016;Şoait a, 2012).
Turning the attention to Sweden, one of the countries of destination for the impoverished European Roma, a new chapter to Roma presence was to be added with the EU accession of Romania and Bulgaria in 2004. Unequal economic developments throughout the European Union and free movement of people, signified the advent of a new group of Roma in Sweden. Again, the 'new', like the 'old' Roma groups, would experience racial discrimination. One of the most explicit expressions of renewed anti-Roma racism is the recent example of the illegal police registration of Roma in Sweden during the so-called 2015 'refugee crisis', when temporary border control was reintroduced. This was done to avoid more refugees seeking asylum in Sweden. However, the intensified racial profiling was also seized as an opportunity to register all 'undesired' 'EU migrants' (Hanson, 2017;Persdotter, 2019;Schclarek Mulinari, 2018).
By 2015, begging was banned in some municipalities through local political decisions. Moreover, plans to prohibit begging at the national level currently determine the national discourse, and the latest public investigation (SOU, 2016) advises devolution in tackling the begging issue. The banning of begging in an increasing number of municipalities in Sweden is having detrimental effects in the lives of Roma street-workers, as our empirical material will show.

Roma housing precarity in Sweden and in Romania
Roma people face extreme poverty, unemployment and institutional discrimination both in Sweden and Romania (Berescu et al., 2013, Lind & Persdotter, 2017, which is reflected in their precarious housing conditions, rural and urban alike. Romania experienced a quick decline in housing production after 1989 and virtually the entire housing stock was privatised. Without any affordable public housing production and an emerging uneven economic landscape, rural areas were left with small or non-existent labour markets, whereas the new economic activities were now centring in and around Romania's big cities. Romania remains one of Europe's least urbanised countries. The intended socialist industrialisation and modernisation efforts were to a certain extent hampered by Ceausescu's autarkic policies. Consequently, many people, including 60% of Romanian Roma, live in the countryside (Berescu et al., 2013). Many Roma reside far from thriving urban labour markets and are thus much more prone to unemployment, with only 23% of people of working age active in the formal labour market. Moreover, over one million Roma live in poverty and the majority of these impoverished households are spread over roughly 2,000 segregated settlements. Only 60% have clear legal forms for the houses they inhabit (Berescu et al., 2013). The reasons for such high deprivation from adequate housing are diverse and have roots in different eras (Vâlceanu & Suditu, 2015). Besides historic forms of exclusion, including slavery as presented in section two, under socialism many Roma were housed in old centres planned for destruction (Chelcea, 2003), while others were placed in villages and towns in Transylvania in highly segregated areas (Berescu et al., 2013). Thus, forced settlements of Roma communities under socialism reinforced segregation and many low-skilled Roma had no direct access to better built state housing (Troc, 2002).
After 1989, many Roma faced a worsening in their housing conditions. New public (social) housing was never established and the result is rather dramatic. Since the foundation of The Romanian National Housing Agency, in 1998, only 31,000 units have been constructed (Amann et al., 2013). Also housing specifically targeted at Roma remained low and so far has not even exceeded 1,000 dwellings, of which the great majority were built in segregated locations and of low quality (Berescu et al., 2013). It can thus be argued that also in Romania, desegregation is not on the political agenda.
Like Romania, Sweden has also experienced a strong marketisation of its housing sector, starting in 1991, when a bunch of deregulating housing policy reforms passed through at the national level (Christophers, 2013;Baeten et al., 2017;Grundstr€ om & Molina, 2016;Hedin et al., 2012). Public housing was transferred to the private sector through selling to private companies or residents, transforming rental housing into cooperative ownership. Due to privatisation of municipal stock and low volumes in new production during the last three decades the traditional universal municipal rental housing, an effective alternative to social housing, has decreased both in numerical and relative terms (Baeten et al., 2017;Grundstr€ om & Molina, 2016). Christophers' (2013) study on the Swedish housing model demonstrates the gradual commodification resulted in a monstrous hybrid of a system mixing public and private welfare that stopped working. The country moved from a surplus in rental housing to a huge shortage in less than ten years (Grundstr€ om & Molina, 2016), a shortage that is unevenly distributed in terms of class, gender and space. Listerborn (2018) and Andersson and Molina (2003) further clarify how this model resulted in an institutionally generated segregation within the urban areas with large public housing estates.
While tenure neutrality slowly disappeared and public housing was explicitly targeted at lower income households, the ethnic composition of large public housing estates also changed dramatically, leading rapidly to a process of racialisation of the Swedish city (Andersson & Molina, 2003;Molina, 1997;Pred, 2000). Hence, even if precarious EU migrant groups could acquire a right of residence, it would be highly unlikelygiven the current endless housing queuesthat they would be housed in affordable dwellings.
After an increased migration of impoverished Romanian Roma, especially during and after the global financial crisis of 2008, public debate soon became focused on a new group of Eastern Europeans that established their presence in Swedish public spacesliterally (Hansson, 2019). This presencecharacterised by beggingreminded the Swedes that since the early 1990s no real action had been undertaken to prevent homelessness, although it has been increasing constantly since the beginning of the 1990s. Sweden is, in fact, the only Nordic country without a national plan to combat homelessness (Sahlin, 2015). Within this context of a failing strategy for shelter provision and affordable housing production, public opinion and outrage instead targeted the poor begging subject. Moreover, in rightwing and far right discourses on 'the burden' that refugees are supposed to cause to the Swedish welfare system, the housing question in terms of lack of housing, appears quite often (G€ oteborgs-Posten, 2018;Svenska Dagbladet, 2016). Housing crises seem to be a feasible argument in the articulation of exclusionist racist discourses.
In their article from 2018, Hansson and Mitchell analyse the situation of homelessness and deprivation in which Roma EU-migrants in Sweden find themselves. Building on Agamben's (1998) concepts of the camp and the bare life of the extremely excluded, the authors assert that Sweden has created a 'state of exception' for homeless, impoverished EU migrants wherein they are, due to extreme exclusion, forced into living 'bare lives', as the naked, de-humanised Other. Sweden is providing the Roma EU-migrants with significantly fewer rights, and almost no access to care, compared not only to the majority of the population, but also to other migrants to the country. They point, moreover, to the contradictions in a welfare state that denies them help with shelter appealing to the national law, at the same time as some municipalities actually helping them with shelter referring to the same law. As we will see in the empirical sections, the material collected in Uppsala for the present study corroborates Hansson and Mitchell's thesis. The conditions described were not just 'homelessness-like conditions' but true and pure homelessness.

Research design and methods
Our study is based on data gathered in three different fieldworks between 2016 and 2018. While the initial plan was to use a case study approach on Roma street-workers in Sweden, we soon realised that the place of origin in Romania also needed to be studied. This multi-site study consists of indepth semi-structured interviews in Uppsala and field observations, surveys, and short structured interviews in the villages of origin in Romania. In Uppsala in the period of January-February 2016 a total of 11 interviews (lasting between 40 minutes and 1.5 hour) were conducted with 15 Roma street-workers. The informants were aged between 19 and 50 years and were identified based on their activities on the streets. The interviewees were eight men and seven women; 13 were homeowners or they owned the land at their settlements in Romania, and two were squatters or homeless. In Uppsala, five lived in the woods outside Uppsala, four lived in caravans, three in homeless shelters, two under a bridge and one person lived in a car. Because of time constraints, the interviews (all conducted in Romanian) were integrally transcribed and translated into English by a Romanian proof service. The interviews were planned, analysed and coded by both authors, whereas Author 1 was the sole interviewer. The methodological challenge of (professional) translation of entire transcripts from Romanian to English for analysis by a person external to the research process induced a certain loss of control over the narratives. Nevertheless, the authors worked in close collaboration with the transcriber/translator to overcome such issues.
The interviews provided insights in how Roma street-workers are exposed to racism, poor housing conditions, and stress caused by irregular income from informal economic activities. After the eleventh interview we realised that various topics were relatively well saturated (within the context of our resources), as for instance the homelessness-like conditions in Sweden and the importance of informal work in Uppsala for financing housing improvements in Romania. We decided to stop the Uppsala fieldwork and proceed to a study in Romania. We identified three Roma groups in Uppsala: Rudari from Vâlcea County, Kalderash from Bac au County and Horohane from Dobruja region. We had only interviews with the first two groups as the latter showed reluctance to participate.
In Romania two fieldwork trips of a total of three weeks were included as background material (May 2016 by the first author and March 2018 by both authors), in which, among others, 100 households were surveyed from six villages (located in Vâlcea and Bac au County) that had been mentioned during the Uppsala-interviews. The structured interviews had some survey elements in them and in Vâlcea were guided by Tiberius, a former interviewee in Uppsala, who accepted the role of gatekeeper. The structured interviews lasted up to 30 minutes and were documented in extensive notes. The questions focused on migration to Sweden, quality of housing, and recent investments in housing. In the visited villages many of the earlier discussed topics concerning housing deprivation, economic hardship, and societal marginalisation became evident. All settlements were located at a far distance from local labour markets, relatively disconnected from urban infrastructure, but also in an ongoing process of rebuilding and improvement. Consequently, the field observations, surveys and structured interviews in this part of the study yielded valuable information on the importance of circular migration for the housing conditions in the settlements, the ongoing issues these marginal settlements face, and recent efforts to undo slum conditions.
The interviews in Uppsala and the observations, the surveys and the structured interviews in Romania enabled us to reflect upon the empirical data from different viewpoints. Through this multi-site approach it became evident how circular migration patterns, housing developments in Romania and ongoing homelessness in Sweden are interlinked. The quotations that follow are all extracted from the interviews in Uppsala.

Housing, the precariousness of life and forced nomadism
Ilie lives with his adult daughter, his wife and his two grandchildren in a caravan somewhere in Uppsala, at the time of the interview. Lack of sanitary services and overcrowding not only characterise his and his relatives' everyday life in Sweden as homeless, but also back in Bac au County, Romania, where he owns a house. There, his mother is waiting for them while taking care of the house. Ilie, as all the other interviewees, repeats often that they are poor and, using their own expression, 'mizerabil' [destitute]. Ilie is aware of his social condition. He and his family, like the rest of the participants in our study, moved thousands of kilometres to seek some income. Still, compared to their living conditions in Romania, almost all interviewees (13 out of 15 in Uppsala) considered it better living in a caravan, without water, needing to go to another place to wash themselves and their clothes.
As it is already well established by earlier research (e.g. Berescu et al., 2013;Berescu, 2011;Lancione, 2017;Vâlceanu & Suditu 2015;Vincze, 2018), the living conditions of many Romanian Roma in Romania are precarious. They are lacking decent housing, sanitary services, jobs and money to make ends meet for the basic needs. All these aspects were mentioned by all interviewees, but Andreea (42 years old) summarises them concisely: Andreea: I rented several places. Then, on this piece of ground from the town hall, I installed a kind of tent, made of plastic, a kind of nylon greenhouse. I covered that with blankets, so that my children wouldn't bake themselves inside this greenhouse. Sometimes we slept outside, when it didn't rain. Now we've built this 2-room house with a hallway in between and that's where we live. My children live there now.
Author 1: Do you have running water and electricity? Andreea: No. No water, no electricity. There is a source of water, with a tap, at about 400 metres distance from us and we go and fetch water with our house carts, as it's quite far. There's no electricity, because we live in the middle of a field, somewhere isolated from the village.
Scholars from the field of homelessness have largely shown that the lack of housing means not only the lack of shelter, but it represents the lack of a stepping stone to other resources that can accelerate the process of pauperisation. Homelessness affects access to jobs, health care, education etc., as well as having detrimental consequences for several social and emotional aspects of life (Mitchell, 1997(Mitchell, , 2011Mitchell & Staeheli, 2012;Sahlin, 2015). Some of the participants referred to the risk of losing custody of the children, the impossibility of taking jobs or sending children to school. Ana (43 years old) tells us about the fear of losing her children: The interviews, from both women and men, refer often to their worries about not being able to offer their children a decent life and a better future, and ultimately losing them forever in the hands of the social services. Having an address and being able to show the social authorities that the children will have a home protected them from the destruction of family ties. Poverty and segregation mean often the impossibility of sending children to school. Physically isolated allocations with bad accessibility, together with poverty, diminish mobility, which may put whole communities outside schooling as they lack money for making use of public transportation, and again, risk losing child custody.
Adrian: I only attended one semester of the 11th form and then I quit. It was difficult for my family to keep me in school: I had to commute daily from my birthplace to Râmnicu Vâlcea and the bus and some pocket money for food and such things would cost us 2.5-3 million (i.e. RON 1 250-300) [per month].
Some respondents are also confronted by the paradox that having a house and a home, but located in the wrong place, could put them in problematic, and even dangerous, situations. Adrian (20 years old) clarified this paradox: Author 1: Have you ever felt in danger, there in Valea lui Stan?
Adrian: Yes. There can be enormous floods, when it rains a lot. And water can enter the house.
Author 1: What happens with the house then?
Adrian: It gets shaken. And wet through, as it is made of wood. Furthermore, sometimes, being considered a homeowner can in itself play against the poor Roma. As Emil (33 years old) explained it: having a house of one's own, no matter the quality of the dwelling, can be used as an argument by the local authorities for blocking the access to diverse societal resources such as social assistance.
Author 1: You could, however, make a request for social aid, right?
Emil: The problem there is that they won't accept it.
Emil: Because they'll say we've got good living conditions. They'll say we've got a good house, but won't think that I can't eat from my house. I don't have an income.
The lack of a job and income makes it difficult for the interviewees to pay their local taxes, which often puts them at risk of losing their property, when they own one, as M ad alin (33 years old) explained.
We asked for a home, but they refused us at the town-hall. They said we had too many debts on the old house, because of the unpaid taxes. They don't really care about that house. If the town-hall finds someone else who's got money, that person pays and they may enter the house. That's how it works with us … There are families who live on the streets, but the ones with money can get a year contract on the house, they pay what has to be paid and they may live there.
Besides the vulnerability caused by the lack of income to pay the taxes, there is also an imminent risk of being evicted because of the lack of legal property rights. Andreea (43 years old) asked the municipality for a piece of land, which she got, but without receiving any legal documentation of her legal right to the plot on which she eventually built a small dwelling.
Andreea: They [some families in the village] bought their pieces of land. They've got [property] documents. We are 4 families who've received land, but haven't got any papers for it. We don't have the money to buy this land. The town hall refuses to give us any papers, but we do pay annual taxes for it.
Author 1: Have they mentioned any price at which you could buy your land?
Andreea: No. The town hall said they would issue the papers when it's possible. But it might take time until they give us papers, maybe 20 or 30 years. So, they can come and demolish our house any moment.
Author 1: How about the other families? Andreea: They are in the same situation, without documents.
Precarity and discrimination prompted poor Romanian Roma to leave the country for cities in Europe where they can earn money for their subsistence. Moreover, the narratives as well as observations in Romania revealed that the second most important motive for moving to other cities for begging or working in low-paid jobs was to make improvements in housing.
Adrian: We'd sleep anywhere, only to get something to build the house.
In all narratives, the housing question showed itself to be at the core of the list of needs for the Roma begging in Uppsala. The goal was either to build a new house or make reparations to the old one, as well as extensions 'to build an extra room'. The narratives include many examples of the importance of begging/getting money to improve housing conditions (repairing, expanding and building from scratch) or to acquire a home, in the case of Ana who had lost her adobe house to her oldest daughters and their husbands, with whom there were some conflicts. She did own the house but not the land, which means that she cannot sell.
At the same time, it was clear in the narratives of the interviewees that when referring to a 'place of your own' they did not necessarily meant having a property, but rather a home. Renting or owning seems to be less important for the interviewees, but at the same time, renting from the public housing system was felt as insecure, since it implied a permanent risk of being evicted as soon as you could not pay the rent, as in M ad alin's case: Author 1: So you've always lived with this fear of being evacuated? M ad alin: Yes. It's awful to rent from the state in Romania. If you can't pay, you're thrown out.
While there were several housing projects in the narratives of the interviewees, the most common goal was to build another room. That has of course direct relation to the high levels of overcrowding. For instance, some interviewees compared the size of their dwellings with the size of the small room of about six square metres in which the interview was conducted at the university building. Two rooms and a hall, seems to be the most common design of the dwellings. More than a design, it showed to be a minimum turned into a standard. The problems mentioned by the participants, i.e. the instability in ownership, the insecurity around the right to stay put in their locations, the hardships in trying to keep the family gathered, and the precarious quality of housing, among other aspects, caused us to reflect that the problem with precarity of life of the Romanian Roma cannot be solved solely by getting a place to live, not even if that place is privately owned. From this perspective, given the situation of extreme social exclusion in which the Roma live, together with the policies of abandonment on the part of the Romanian state, it became very clear that nomadism is one of the few possibilities for survival for this population. The mobility should be understood as forced displacement rather than a matter of lifestyle (Davidson, 2009), and in this particular case with forced migration as the corollary. Consequently, building on the narratives of the interviewees, we prefer to use the term 'forced nomadism' instead of 'cyclic migration', since the Romanian Roma moving around in Europe hardly have an existing alternative for survival or betterment. In Romania they lack job opportunities and need to approach the authorities for getting some help; they succeed at most in getting some low social subsidies for the poor, like the child subsidy, but most often, they get promises of help that are never fulfilled. In the next section, we refer to the lives of the participants while in Sweden, focusing especially on the racism experienced by the Roma street-workers begging in Uppsala.

Precariousness and racism in Uppsala
If the housing situation in Romania was precarious for all the interviewees, their shelter conditions in Uppsala were even worse. None of the participants lived in a dwelling at the time of the interview. Two were sleeping on cardboards and blankets under a bridge close to downtown. Five lived in self-fixed hovels in the woods around the city and in one case up to 19 miles south of Uppsala, commuting every day for begging. Four interviewees lived in caravans and one lived in a car, moving between parking places after being evicted several times.
Apart from the dismantling of the squatting places that 'EU migrants' occupied spontaneously at several locations in the city, in Uppsala there has been a total lack of engagement from the local authorities. There were some declarations of intentions and plans for offering a place for camping with sanitary facilities for the roughly 15 families that the municipality has estimated remain in the city but all initiatives have been withdrawn after political controversies from the city hall. A justification of the lack of local engagement is the fact that, at least according to the local authorities, there are no children of school age (which would probably put the local authorities in a different situation) and the arrangement of facilities would 'encourage more people from Romania (sic!) to come to Uppsala' (Sandow, 2018).
Instead, mostly religious organisations bring the Roma street-workers help in the form of charity, which the receivers appreciate. Three persons (one couple) were momentarily staying in charity shelters for homeless people, but they were not allowed to stay there during daytime having to leave early in the morning and come back in the evening. If they got ill or if the weather conditions were harsh, they still had to remain outdoors. The cost for the shelter was about three euro a night, which for them represented too much money. That was the reason why they all avoided to stay at those shelters for longer periods or at all.
Adrian. ( … ) This is what I make a day: 50-60 crowns, no more than that. I make 60 crowns, should I pay 25 for sleeping? I also have to pay something for my food, then I can't put anything on the side to do something with it at home. As it were, I keep myself alive here: I eat and sleep. What else can I do? If I don't pay for sleeping, I've got enough for food.
While being on the streets, all participants had been victims of racist harassment, though they did not always complain about it. On the contrary, they recognised the nature of the event as racism, but tended to justify or explain it away. They were instead expressing gratitude for what they called solidarity from the local population, coding solidarity as opposite to racism. But the narratives, as in Ilie's case below, reveal that they are targets of anti-Roma racist hatred on a daily basis.
Author 1: Yesterday you told me that some people look unfriendly to you, that they spit when they pass by and see you begging.
IIie: Some locals here hit my cup. Like this man. Every 3-4 days he hits my cup in which I collect the money, he curses at me in Swedish. He's one of these people who hate Romanians. This guy troubles me every 3-4 days. He's the only one. He passes by and troubles me every 3-4 days. He hits the cup in which I collect money, he spits … Moreover, racist harassment seems not to surprise the interviewees as they are almost expecting it to happen. Cristian (19 years old) expresses it categorically saying that 'it happens all the time'. Both in terms of gestures and verbal offence.
Cousins (2019) and Wing Sue (2010) refer to the different ways of coping with a lifelong accumulated experience of racist micro-aggressions, as a resilient way of adapting to a hostile environment. Adaptation and acceptance can be considered to be among such coping strategies. Remus (22 years old) prefers to talk about hatred: Remus: I feel how they look at us with hatred. Sometimes they get a sudden unfriendly frown when they see us. 2 Another manifestation of racism seems to be the recent reduced willingness to giving money to Roma street-workers in Uppsala and in Sweden in general. Indeed an increasing reluctance to help was being observed by the interviewees.
Adrian: ( … ) and we don't make money, I've told you. I don't know what's going on, but people don't give money to Roma beggars anymore. Maybe they don't have money anymore, or … I don't know, they don't give money anymore. I make 50-60 crowns [SEK] a day. That's the limit. This is an average. One can't make more than this.
In his recent work on the issue of racism against the Roma street-workers, based on an analysis of more than 1,000 articles in Swedish newspapers, Hansson (2019) translates the statements included in the articles into feelings that the presence of the Roma street-workers gives rise to among the Swedish population. Hansson concludes that the different attitudes adopted are closely related to Sweden's long history of particularly anti-Roma racism. One of the media trends in this respect has been the association of the Roma street-workers with organised criminality in charge of international leagues who make enormous amounts of money at the expense of the exploited Roma street-workers. Mass media has then intended to imbue the community with alleged organised crime, as a result of which several municipalities (e.g. Sala and Vellinge) sought to find ways to prohibit begging. Both givers and Roma street-workers became stigmatised and Hansson considers it to have had an impact on people's disposition to give.
In fact, in our trip to Romania 2018 we heard those narratives from the people we met who had been begging in Sweden before. 'People are not giving anymore'. It seems to be that elsewhere in Europe, criminalising racist discourses on 'The Roma' are having a negative impact on the general attitudes against them. Similarly, when studying two cases of accusations of (false) child theft, K ocz e and R€ ovid found that: As a result of the media campaign about the dubious origins of the 'blonde' Maria in Greece, human rights organisations have reported a backlash against Roma in Greece and beyond. In both cases, the inevitable result is that media portrayals of the Romaand the language that journalists and reporters use to describe them to media consumersserve to perpetuate the cycle of inferiorisation and dehumanisation, hence racialisation, of the Roma. (K ocz e & R€ ovid, 2017: 688) The cases studied here bring a new light to the conceptualisation of racialised poverty (Roediger, 2017;van Baar, 2018;Yıldız & De Genova, 2018). While the precarious conditions of the people included in this case are evident, and their relation to a long history of racial discrimination undeniable, with some exceptions, Romanian Roma interviewed in Uppsala did not conceptualise their experiences of exclusion and discrimination by referring to racism in the first place. Though being aware of racism, they stressed their conditions of extreme poverty, lack of education, and the impossibility of getting any jobs as the main reasons for being at the margins of society. The most common interpretation of hostility and harassment, which, in fact, everyone had experienced, was because they were poor. But the link between being extremely poor, lacking money, lacking a job and often lacking a home, was seldom put in relation to racial discrimination.
Overall, our study has clearly substantiated the fact that begging is the substitute of a state that does not provide its citizens with their basic needs, as well as a market that does not give working opportunities. That is, a state and a market acting not only impoverishing people by class but also by race; as called by Goldberg (2002), a racial state. Using Mulinari and Neergaard's distinction below, we assert that the re-racialisation of 'The Roma' (Balibar, 2009) in Romania and of the Romanian Roma in Sweden obeys the features of exclusionary racism: We consider this a mainstream, even a systemic example of what we call exploitative racism, operating by producing a usable (exploitable) racialised labour force through discursive and institutional practices.
[ … ] Exploitative racism operates through the process of racialisation that legitimises the capitalist production of profit. (Mulinari & Neergaard, 2017, p. 92) Whereas exploitative racism uses racialised groups as cheap labour, excluding racism expels some groups keeping them outside the labour market. Exclusion in this case operates at all societal Romanian and Swedish institutions, including housing and education, leaving the population in the worst of the marginalities. Using this distinction, we assert that the re-racialisation of 'The Roma' that Balibar elaborates on, and that we mentioned at the beginning of the article, in Romania and in the case of the Romanian Roma in Sweden, reflects in the first place what Mulinari and Neergaard call exclusionary racism. 'The Roma' are neither welcome as cheap labour, nor as Roma street-workers. The circulation of the Romanian Roma among the European countries including Sweden, is a precarious circulation of a people that find themselves in a condition between homelessness and super precariousness in housing, navigating in an environment laden with racism and hostility.

Conclusions
Through a qualitative, multi-sited study of housing among Roma streetworkers in Uppsala, this article reveal that one of the most common reasons for begging, besides being able to send money to the family members at home for their subsistence, was to build a house or to improve an existing one in the village of origin, where the rest of the family was waiting. In a context of total lack of housing policies to counteract precarious housing conditions among the Roma population and in the context of increased racialisation, this is an important finding. Having the chance to move abroad for earning money for subsistence and to improve their homes making it possible to get a more decent housing situation, becomes an important strategy for the Romanian Roma. Precariousness, which is a characteristic of their condition when being in Romania and also when being in Sweden, is exacerbated by the fact that all obtained income must be remitted to or brought back to Romania. Spending money on paying for a shelter in Sweden, or trying to rent, is not seen as an option for most.
Reflecting on participants' experiences in Romania and in Uppsala, it is hard for us to discern in which of the two contexts poverty is experienced as worse or better. All participants thought their living conditions in the two countries were extremely poor, which indicates that the distinction after all does not seem to be relevant. Also in the two contexts the participants experienced discrimination, but their experiences were diverse and were verbalised differently. The precarious housing conditions in the two geographical contexts were in fact linked: the reason participants had decided to migrate in search of money was to improve their housing situation in Romania; their precarious condition of life was also the reason why that money was not invested in better living conditions while being in Sweden, which, regardless how long their stay lasted, was considered to be a transitory location.
This study reveals that, in spite of the extreme pauperised living conditions of Roma in Romania, there is an almost total absence of social protection, as well as the lack of responsibility from the Romanian state, both at the central and at the local level. Given a racialised labour market operating on the basis of exclusion, in this particular case of Romanian Roma, they are more or less abandoned to a destiny of extreme marginalisation. For the poor Roma, one way to cope with this is moving long distances to obtain some income.
Likewise, their precariousness in Sweden as homeless street-workers begging is a consequence of the precariousness in Romania, but also a result of Swedish racialised policies towards Roma EU-migrants and the unwillingness of both central and local administrations to provide welfare services to this population. This was revealed by the fact that the Swedish authorities have no interest on including Roma street-workers in any welfare provision, whereas a minimum of help is offered by a couple of church organisations. Intensified precarity up to what we called super precarious conditions, leads as well to the self-fulfilled prophecy of Roma being different and wanting to stay in poverty because of their character (K ocz e, 2018), an argument that cultural racism embraces and uses to justify social inequality. In the case of Romanian Roma the precariousness of poverty is intimately related to racism, which makes this case a crude illustration of the concept of racialised poverty (van Baar, 2018;Yıldız & De Genova, 2018).
Authorities in Romania allege that the poor Roma asking for benefits from the municipality are not in need of help because they own a house. Moreover, irregularity in tenure and the lack of legal ownership is a factor of insecurity and increased precariousness. The Roma invest in building a house, and later on improving it, even though they are sometimes frightened of losing it all, as the narratives contained plenty of experiences of evictions. The notions of tenancy or ownership have no value when they do not have money to pay the rent or the taxes. So it was for several of the interviewees; some had been evicted from their homes in Romania because they could not afford to pay local taxes anymore, others were permanently afraid of being evicted. Although they were considered homeowners, they actually were not; they had succeeded in building a modest home, but it remains unclear, especially for the interviewees from the village in Bac au, whether they have the legal right to it. The precarious housing conditions are hidden behind the screen of homeownership, whereas the factual condition should rather be read as super precariousness. Besides the low quality of the dwellings, lack of urban sanitary provision, and overcrowding, almost all the interviewees felt permanently threatened by eviction and displacement. At the same time, nomadism had become a forced way of life, another variant of displacement.
Also in Sweden, living in car parks in cars or caravans, in the woods or under a bridge, meant being evicted every now and then. Whenever the interviewees stayed in the charity shelters, they were evicted on a daily basis as they always had to leave early in the morning. They could barely afford the price of the shelter and at the same time save money for the remittances, which means that they belong to the groups having the worst conditions of housing precarity in Europe.
Housing conditions are intersected by racism-here anti-Roma racismwhich calls for attention among housing researchers on the intersections between race and class within the neoliberal global order, and include marginal housing conditions as one relevant motive for migration. The case of super precarious housing presented here contributes evidence for this set of connections. Notes 1. In Romania it is still common to refer to millions when speaking of 100s of new lei (lei is plural for leu). The former leu (ROL) devaluated rapidly in the 1990s but was revalued in 2005 at a rate of ROL 10,000 to RON 1. At present, 1 euro converts to 4.75 lei. 2. Note from the translator: Remus uses a very expressive, rather poetic way in Romanian to put this: "seîntunec a pe noi", which translates literally as "they are getting dark on us", or "their faces turn dark when they see us"