Winding Paths through School and After Young Swedes of Migrant Origin Who Failed in Upper Secondary School

Departing from careership theory, this chapter aims to add to knowledge about school careers and school-to-work transitions of young people with a migrant background, here referring to individuals born outside or in Sweden whose parents are first-generation immigrants.1 More specifically, we want to increase the understanding of the processes resulting in a school failure, and the subsequent process when the young person tries to enter the labour market and/or strengthen his/her educational qualifications. The analysis is based on life-history interviews with 27 Swedish young adults (21–23 years old), most of them from refugee families of non-European origins. All of the respondents left school without completing upper secondary education. Few previous studies have focused on both the school years ending in dropout and subsequent careers of early school leavers, as reported by the young adults themselves (for exceptions, see e.g. Henderson et al., 2007). As discussed in the following section, political discourses, official statistics and research often portray gloomy prospects for young early school leavers generally, and those of migrant origin particularly. However, our research provides more nuanced indications, as the young adults interviewed described not only difficulties in school

and later life but also factors that could facilitate their careers, particularly if schools and other institutions provided more professional and timely support.

Notes on school failure, dropout and school-to-work transitions
Rather than making a decision to drop out from school at a certain point of time, the young person usually follows a process over several years where multiple factors interact and result in a school failure. These factors are related to both the student and the family, to educational practices, and policies (Dale, 2010;Lundahl et al., 2017;Rumberger, 2011). Learning problems, low previous academic achievement, disability and mental problems are factors at the individual level associated with higher risks of school failure (Casillas et al., 2012;Myklebust, 2012;Quiroga et al., 2013). Scarcity of socio-economic resources, e.g. low incomes and educational levels and high mobility as well as social problems in the family and neighbourhood, are also strongly correlated to young people leaving school without upper secondary qualifications (Østergaard Larsen, Jensen & Pilegaard Jensen, 2014;Rumberger, 2011). A factor that has received less attention until recently is that school itself may contribute to failure (de Witte & Rogge, 2013). Inability to detect early warnings, insufficient provision of individual support, neglect of bullying, and in some cases even discrimination against or harassment of students by staff are school-related factors that have been shown to contribute to truancy and poor academic results (Cornell et al., 2013;Elffers, 2012;Lamote et al., 2013). Higher proportions of students of migrant and ethnic minority backgrounds leave school without complete grades than other students, but several studies have shown that socio-economic conditions strongly contribute to such differences (Elffers, 2012;Lundahl & Lindblad, 2018).
Nearly all Swedish teenagers continue to some kind of upper secondary education-academic or vocational-after leaving compulsory school. The programmes are normally of three years' duration, except shorter, so-called introductory programmes (before 2011 'the individual programme') intended to help students to become eligible for a three-year programme. Special upper secondary schools, targeting young people with a mild learning disability, provide four-year programmes. In 2017, almost 90% of young people with Swedish-born parents were eligible for one of the academic national programmes and slightly more than 90% for a national vocational programme. The corresponding figures for students with foreign-born parents, who were born in Sweden themselves or had attended a Swedish school from grade 1, were slightly lower. However, less than half of those who had arrived in Sweden at a later stage had eligibility for an academic or vocational upper secondary programme (Skolverket, 2018a) and were hence referred to an introductory programme. The proportion of school-age children of migrant background has increased faster than the proportion of the whole migrant population, especially since 2005. This group constituted approximately one quarter of all students in compulsory schools in 2016 (Lundahl & Lindblad, 2018). Within the group of children with a non-Swedish background, the boys had somewhat lower qualification rates than the girls (Skolverket, 2018a). In 2017, 78% of students with Swedish-born parents had obtained a degree within four years of starting at upper secondary education, while the corresponding proportion among students with a migrant background was 50%. Gender is also a significant factor, as girls had around 5% higher success rates than boys (Skolverket, 2018b).

Employment difficulties hit early school leavers particularly hard
The European and Nordic labour markets of the early 2000s offer young people with no secondary qualifications far fewer job opportunities than a few decades ago. Industrial restructuring, offshoring work to low-wage countries and automation of production and services have reduced numbers of entry jobs considerably. After the latest economic crisis, youth unemployment rates and numbers of those not employed, in education or training (NEETs) have stayed at high levels. In addition, part-time and insecure employment has risen considerably due to 'flexibilisation' (Blossfeld et al., 2008;OECD, 2015). While young people of non-European origin tend to be over-represented among both young unemployed and NEETs in the European Union, the proportions of young NEETs in Sweden, both Swedish and foreign born, are considerably lower than the European Union average (Figure 4.1). A plethora of international studies and statistics show that dropping out and leaving school early increase risks of unemployment, unsatisfactory working conditions, poverty and poor health (e.g. Brunello & De Paola, 2013;Lamb & Markussen, 2011). Youths with an immigrant background are over-represented among the non-completers, as are youths from homes with little economic and cultural capital, and disabled young people (Dale, 2010;Rumberger & Lim, 2008). As shown, these categories may be overlapping. In the 2010s, Sweden granted more refugees asylum than most other European Union countries, in proportion to its population, and accepted the most unaccompanied asylumseeking minors in absolute numbers. Sweden displays a considerable achievement gap between students born in Sweden and students with foreign backgrounds, particularly those who arrived after the age of seven (Lundahl & Lindblad, 2018). A general finding that immigrants, particularly refugees, tend to be at the margins of the labour market in most countries can be partly explained by lower-than-average educational attainment (OECD, 2010). However, socio-economic background and being a refugee, migrant or guest worker are also contributory factors (Rumbaut & Komaie, 2010). In 2017, the unemployment rate for foreign-born people eligible to work in Sweden was among the highest in the OECD countries (15%), and most different from that of the native-born population (OECD, 2018).
In Sweden, as in the rest of Europe, a 'transition machinery' has developed during the last 25 years that is intended to prevent and manage school failure and dropout, and facilitate young people's transition to the labour market (Brunila et al., 2011;Kurki & Brunila, 2014;Lundahl & Olofsson, 2014;OECD, 2016. Also see Chapter 7 in this book). However, the unusually farreaching decentralization and marketization of education, and disappearance of many 'entry-level jobs' in Sweden have probably increased the difficulty of transitions. Young people now have to navigate through bewildering jungles of schools and educational choices (Lidström, Holm & Lundström, 2014), and the local variations in provisions such as special needs education, career counselling and youth schemes are unreasonably large (Lundahl & Olofsson, 2014).

Theoretical framework
The theoretical framework applied here rests on careership theory, which is highly valuable for analysis and elucidation of individuals' career decisions in an agency-structural perspective (Hodkinson & Sparkes, 1997;Hodkinson, 2008), and conceptual formulations of 'otherness' (Hall, 1994;Anthias, 2002;Balibar, 2004;Trondman, 2007). Drawing on concepts formulated by Pierre Bourdieu, particularly field, habitus and capital, careership theory recognizes that young people commonly have fewer resources than other actors when proceeding through education and work. The basic careership model has the following main components: the horizon of action, the field, turning points and routines. The horizon of action denotes what an individual considers to be desirable and possible, based on her/his habitus as well as structural opportunities. Within their horizons of action, people are held to make pragmatic-rational choices rather than technical-rational ones. The field is the environment in which career choices and transitions from school to work take place, and various actors with different amounts of capital strive to achieve their goals. Life careers consist of alternating turning points (when individuals make decisions and change paths, resulting in changes of their habitus) and routines (periods between turning points when people have a series of experiences and evaluate and deal with them). Turning points may be structural, self-initiated or imposed by others, while routines have varying degrees of alignment, with decisions taken at turning points and the decision maker's intentions. In a revised version of the theory, Hodkinson (2008) emphasizes the importance of learning as an integrated and central part of people's progressive career construction. Furthermore, there is less distinction between routines and turning points; careers are regarded as being constructed and developed gradually during the individual's positioning in various career-related fields. Moreover, the career construction process may be largely described in terms of learning, and as both embodied and social.
Drawing on Said (1978), the concept of otherness helps making visible the mechanisms that create dichotomies, 'us and them' , and how knowledge of 'the other' is produced as something outside of the discourse of (Western) normality, where ethnicity or cultural identity are seen as essential (Hall, 1994). Not belonging to the majority, being an immigrant often means being constructed as subordinate and a social problem. Using otherness as a tool to study the exercise of power, beyond and behind the use of ethnicity, culture and gender in interaction with social class, helps efforts to understand how inequality arises (Balibar, 2004). The construction of 'immigrantness' as a social problem can be seen as a practice of thinking, as a grammar, upheld by historical and contemporary public narratives of the other as something worrying and troublesome. The structural categorization of immigrants as the others is experienced at individual level by those exposed to this construction, and who have to deal with the consequences as personal dilemmas (Trondman, 2007). How young people see themselves, and with what and whom they identify, can be analysed using their narratives, i.e. recorded stories about the social fields in which they live and how they relate to assigned collective identities as 'us (native)' or 'them (immigrant)' (Anthias, 2002).

Design
This chapter is based on data obtained in a research project, 'Unsafe Transitions' , in which inter alia career narratives of 100 young Swedes in their early twenties (mean age: 21.5 years) who left school without complete upper secondary qualifications were analysed. It should be stressed that they probably constitute a positively biased sample as they were mainly recruited via career counsellors and thus only included people known to the authorities. Nineteen former students from special upper secondary schools were included, as students with this background tend to encounter similar transition problems to the rest of the group: their grades are not fully recognized either in the labour market or in further education. This chapter primarily considers careers through and after school of 27 of the 100 respondents (14 males, 13 females), all of them first-or second-generation migrants, but they are also compared and contrasted to careers of the young people in the larger group.
The analysis rests on the young people's own narratives of their life-journeys, as expressed in in-depth interviews. In 15 cases (nine males, six females), 3 it was possible to conduct a second follow-up interview a year later, but here we concentrate on the first set of interviews. The narratives are seen as expressions of the young people's experiences and understanding of their social reality, at the time of the interviews, and how they identified, constructed and positioned themselves in relation to the outside world. With this as a starting point, the young people's life stories allow investigation of the connections and relationships between the individual narratives and social context shaping their careers (Kohler Riessman, 2008;Somers & Gibson, 1994).

Career narratives of the young adults
The group of 27 young adults was heterogeneous in several ways. The parents of 18 of the young people originated from a Middle East country (Turkey, Iraq, Iran, Syria, Lebanon or Palestine). Almost a third came from other parts of the world; three were born in other parts of Asia, four in Europe and two in north-eastern Africa. Almost all of the parents were refugees. With a few exceptions, the parents' educational levels were low, and they had subordinate positions in the labour market, although most of the respondents' fathers had gainful employment or ran small businesses. A large proportion of the mothers stayed at home and took care of the (often large) families. Thus, most of the young people had limited access to economic, cultural and social capital (Behtoui, 2004(Behtoui, , 2013, which is a significant factor for young people's school achievement according to various previous researchers (e.g. Rumberger & Lim, 2008;McGrath, 2009;Dale, 2010;Grönqvist & Niknami, 2017). The involuntary migrations of their families also presumably reduced their social and economic capital. Their families were highly important to the young people, most of whom lived with their parents at the time of the first interview. Statements about the family like 'always number one' , 'without the family you have nothing' and 'the family is the first priority' were recurrent. Bilal reflected on the centrality of the family as follows: Friends will come and go: colleagues will come and go, jobs will come and go. Family members are the only ones that will be there all your life. (Bilal)

The school years
All 27 young adults had incomplete secondary education, but there was substantial variation in the shortcomings preventing their receipt of a certificate of completion when they left school. Sixteen of them had attended Swedish education from grade 1 of primary school and on and therefore had more favourable initial education conditions than the other respondents had.
The scarcity of various kinds of capital affected the young people's relations to the educational field in several ways. One was that the parents had difficulties helping their children because of their lack of fluency in Swedish and often limited education.
Nothing, they don't even know what the social science program is. If I had said natural science program, then she would have said: 'What do you become? Ah, a physician' … then she would have said 'Take that' . (Tara) Sabrine felt a strong responsibility to show her parents that they made the right decision to leave everything in order to give the children a better future in a new country. In retrospect, she is glad of their support, particularly from her mother. However, the wish of the parents that she should attend the natural science programme did not turn out to be the best choice.
None of the young adults had memories of getting any substantial help from a career counsellor when choosing an upper secondary programme, even though schools are required to provide such support. Here one may add that careers education as a complement to individual and group counselling is not an explicit part of the Swedish curriculum, as is it is (for example) in Finland, Norway and Denmark.
Most of the young people also had to position themselves in relation to categorization as 'immigrants' . Some of them described how the difference making manifested as discrimination or even racism from teachers and other staff in upper secondary school, which resulted in dropout or change of school. In contrast, many teachers at compulsory level did not construe them as immigrants or regard 'immigrantness' as a social problem. Instead, they considered the individual student's needs and preconditions, in line with educational legislation more generally. For example, Zemar spoke warmly and at length about a teacher in the preparation class he was placed in as a newly arrived seven-year old child: I've never have had any teacher like her. She helped me with everything in school, I learnt Swedish from her. … She liked me a hell of a lot, and I liked her a hell of a lot. She was a terrific teacher, the best I have ever had. … I will never forget her. (Zemar) In upper secondary school the study problems accumulated, while support from school decreased. The provision of special education and other support in school was particularly scarce and insufficient at upper secondary level, where there were higher expectations for students to be selfmotivated and autonomous. For example, Amir did not realize the importance of studying: 'So I didn't take it seriously, I just went there. … Also, I didn't have a clear goal in mind. ' Nobody in school talked with him about this. 'No one told me that if you get so many credit points you can make it to that [programme], it is good if you try hard-nobody, nobody said anything. ' Playing football was a source of joy when the studies went badly. Amir attended the first and second grades in upper secondary school twice, and skipped classes quite often. He left upper secondary school after five years without complete grades.
Several of the young people worked during their studies to contribute to the family finances, which adversely affected their presence in school and time for homework. I remember not being able to attend the religion class because it clashed with my working hours at the job. I was afraid to tell the teacher, you know, so I said nothing. So I got a Fail in religion and in Swedish A. Swedish A started directly after the religion class. Sometimes I could make it, but I was late so I just got scolded every time I was there. They never understood my story. (Aras) For some of the young women, school was a space of freedom from parental control, a possibility to get experiences that were not allowed at home.
And … if you want to do things, and it doesn't have to be about seeing guys. It may just be about doing something fun. Then you do it during school hours. Then it is in school and nobody knows, you know what I mean? … Yes, from grade seven to upper secondary school, I can tell you it was the best time of our lives. (Hawa) Collaboration and communication between school and the family generally seem to have been insufficient or non-existent, in cases of high truancy in compulsory school and, more frequently, upper secondary school. Schools seem to have failed in several respects to cooperate with the parents to increase the young people's possibilities to succeed in school and make wellfounded education and future work choices. Consequently, most of them had a lonely journey through school, and most starkly through upper secondary school.

Entering the labour market: delayed turning points and limited agency
For our interviewees, leaving upper secondary school appears to have been less of a clear turning point and more of a continued routine of uncertainty and dependence on the family for support. This was particularly true for the 11 young people who were not in employment, education or training (NEET), or alternated between temporary employment, youth schemes and unemployment (here called fragmented transitions) during the first two to three years after upper secondary school (Figure 4.2).
Ten of the young people were mainly working, mostly in service jobs they obtained through their families and families' social networks-commonly jobs that may be regarded as blind alleys with little advancement opportunities (Kim, 2013). Six young men and women had spent most of this period in adult education to acquire missing grades, and in some cases had taken shorter vocational training courses. The rest (eight males and three females) were in the 'NEET/fragmented group' , so more females than males were working or improving their educational status (Figure 4.2). 4 The first year was the most difficult for most of the school leavers. After this initial period most of the young adults made choices to improve their situation. Maryam wanted to work and earn money after leaving school, but found it difficult to get a job. Instead, she stayed with her family for a year, doing nothing special. She got a temporary job at a restaurant via a friend, and later at a home for elderly people, with her mother's help. Maryam said that these jobs meant nothing to her. She began to sit at home again with no money of her own, losing self-confidence. 'Everything went downhill' , she concluded. Following a suggestion from a friend, she applied for a job at the Post Office. She got the job and felt that life started to improve; she began to 'come out from this depression' . Thus, failing upper secondary school resulted in transitions between jobs and education, but to varying degrees, corroborating research that points to the need to recognize limitations in agency and blurring of routine periods and turning points (Evans, 2007;Hodkinson, 2008). However, the young people commonly regarded themselves as the architects of their lives. For example, Rezan, who stayed at home doing nothing for one year after leaving upper secondary school, emphasized the importance of his own agency for a successful career: Hard times give motivation. … Exactly. Yes it was like that. At least you have to do something. No-one's going to come to you and say, 'Here, can you take this job, please?' Everything is about yourself. You have to do it on your own, like. (Rezan) For those who initially experienced unemployment, interspersed with shorter periods of employment and youth schemes, the routine of failing seems to have continued after school years. The young people who worked generally had rather unskilled jobs in the service sector, although some were assigned tasks that were more demanding. The young men and women in the first set regarded their jobs as a means to get temporary economic support and had plans for the future that included getting an education, sometimes combined with moving to other places. For the whole group, starting and finishing some form of adult education was recognized as an important turning point. Three years after their upper secondary school period, 5 more than half of the group had used adult education, of varying duration, as a stepping stone to the future.

Otherness
The young men and women described themselves as being talked about and treated as 'immigrants' , a characterization that they had to relate to actively, even if they had been born in Sweden and lived there all their lives. It was a part of the routine periods of their lives and everyday learning on their social position that affected habitus and horizons of action. The structural form of difference making was most obvious in school, and most virulently expressed in forms of discrimination and racism, resulting in some cases in dropout or change of school. One of the young men, Bilal, described having to discuss, almost daily, what other people defined as 'immigrant problems' . He had been affected, he said, by discrimination and generalization because of his non-Swedish looks, although he did not regard himself as an immigrant but as a Swede who had been born and grew up in Sweden.
Tesfay, who had been born in East Africa, had experienced outright racism because of his skin colour. Therefore, he did not want his children to grow up in Sweden, which he characterized as a racist country, but planned to move elsewhere to get a job before having any children. Sirwe perceived the city where she lived as racist, and compared it to another city in the same region that she believed was far more open-minded; for example, youths with a foreign background were more commonly employed there.
The young people's social spaces were reflected in their geographical locations. 6 Places where the young people lived and went to school were characterized by absence of 'the Swedish'-Swedish neighbours, Swedish classmates and the Swedish language. The young adults also argued that it would probably have been far easier to get entrance to Swedish society if they had lived in a city area with more Swedish inhabitants. Their narratives reflect increasing segregation of the housing market in Sweden. Living in poor-quality, stigmatized and/ or publicly subsidized 'housing projects' or 'sink estates' , as most of the young interviewees did, exacerbates feelings of otherness, particularly for inhabitants of non-European origin. This coincides with accumulated poverty, low employment and income levels, low proportions of young people who are eligible for upper secondary school, and remoteness from places of power and influence.

Pragmatic choices-changes in habitus and horizons of action
The young people's careers had been constructed and developed in fields where they had subordinated positions, based on their families' mostly limited social, economic and cultural capital, their own short education and limited experience, and the otherness they encountered. Against this background, their educational and labour market career choices may be understood as pragmatic-rational, enabled and limited by the resulting horizons of action. All their narratives reflect development of their horizons of action from the time they left school, and pragmatic-rational decision-making that changed their positions. Nevertheless, the choices were often constrained by heavy restrictions of actorship and opportunities due to their school failure and relatively scarce resources. Clearly, what happens in school and later in the labour market strongly affects the horizons of action of young people such as our interviewees. The learning and interaction that occurs within the routine periods are both crucial for understanding the process that results in school failure and the subsequent, extended period of establishment in working and adult life. Focusing on learning, regardless of whether it is learning to fail or improving positions in the field, enables elucidation of changes in horizons of action and habitus.
Comparing the careers of the young women with those of the male respondents adds further complexity. A larger proportion of the young women than the young men in our study had been working and/or attended education during the first years after leaving school; fewer of the women had experienced a period of NEET or oscillation between unemployment, youth schemes and brief jobs (Figure 4.2). However, many of the young women had even more restricted space for autonomous action than the young men, which limited their future opportunities. During the school years they were, for example, often expected to take care of younger siblings and help with other domestic tasks. Maryam's experience is quite typical in this respect: Yes, but it turned out like that because I'm the oldest. And mum and dad, yes dad's working all the time so I have to help mum. Pretty much with cooking, cleaning and helping with homework. Check where everyone is, and training times and such. (Maryam) Some of the young women also described their parents as being reluctant to allow them to move to another city for further or higher education, or even being actively opposed to the idea.
To summarize, the narratives of the young people show that school failures, intermittent attendance and dropout 7 are complex and extended processes that are influenced by education, family, and access to power and capital. The young people also encounter difference making through the predominant images and discourses of 'immigrantness' as a social problem and by being located in specific socio-geographic places that delimited their horizons of action.
The family is highly significant and, in most cases, provides security and continuity. The family's present and future situations are crucial elements of young adults' habitus and identity, which also affects their choices. Hence, their individual horizons of action are strongly influenced by their families' opportunities and horizons of action, indicating a need to recognize collectively enabling and limiting horizons of action rather than just individual ones.

The school years
A comparison of the group of young people with a migration background and the larger group of 100 respondents revealed common denominators as well as several interesting and somewhat unexpected differences. The 100 young adults described a downward spiral of negative experiences during their school years that reduced their self-confidence and motivation (see also Lundahl et al., 2017). According to their narratives, most of them felt that schooling had been quite successful and largely met their needs until they reached the last years of compulsory school. In most cases the students stayed in upper secondary school for three or more years but failed to varying extents. This appears to have been at least partly related to the fact that support from school and teachers' commitment to individual students increasingly declines as students progress through the current Swedish education system; in upper secondary school students are expected to be able to be independent and self-disciplined. Another recurrently mentioned aspect of their schooling is being treated as different, by both peers and teachers; subordination and powerlessness are prominent aspects in their narratives. However, the young people did not generally perceive themselves as victims or blame non-completion on this. Rather, they commonly related their failure to their own shortcomings, for example laziness and lack of motivation and seriousness.

Career patterns after upper secondary school
For many of the 100 non-completers, leaving upper secondary education was not the positive turning point that they had hoped. At the age of 21-22 years, more than half of them (54%) had spent the last two to three years in a fragmented ('yo-yo') transition pattern or as NEETs. A third (33%) had work as the main activity after leaving school, and only 13% had mainly attended education. It is notable that a considerably higher proportion of the 27 young people of migrant background had either been employed and/or studying than of the respondents with Swedish-born parents (Table 4.1).
Several factors may have contributed to the finding that the proportion of NEETs was lower among non-completers of migrant background than among those with a Swedish background. One is that parents of many of those of Swedish origin were divorced and/or had problems of unemployment, illness and substance abuse. Accordingly, the lives of these youths had been frequently disrupted by break-ups and moves. In contrast, families of most of the young people with a migration background were intact, and they did not describe their parents as having personal problems or being heavily hit by unemployment.  Moreover, many of the last-mentioned respondents had extensive networks of family and relatives in their neighbourhood, and they had rarely moved during their school years, excluding the migration from another country. However, the young women in the migrant group in general enjoyed less freedom than the daughters of Swedish-born parents did; they had more domestic duties and more restrictions than the Swedish young women had. Compensating for this by using school as a 'free zone' sometimes impaired their studies. Many respondents in the larger group also reported that they had attention deficit hyperactivity disorder (ADHD), dyslexia or similar disabilities that were diagnosed very late, sometimes after 12 years in school. In other cases, learning and concentration difficulties were discovered early, but the young people characterized the school's efforts to treat and compensate for such difficulties as insufficient or non-existent. Consequently, their self-confidence decreased, and they perceived themselves as failures in school. They felt treated as different, both by peers and teachers. In contrast, very few of the young people with a migration background spoke about disability, learning difficulties or bullying as the major problem, but instead often related to their immigrant background when they described being treated as different.
The career patterns after upper secondary school also depended on the time spent in upper secondary education (Figure 4.3). If we look at the whole group of 100 young adults, a higher proportion of those who spent less than two years in upper secondary education were in the NEET/fragmented category two to three years after leaving school. Thus, spending at least two years in upper secondary education apparently decreased risks of being in this category, probably at least partly because time in upper secondary education is correlated with chances to get more passes, and filling gaps in subjects or grades, often parallel to gainful work. Those with a migration background had similar career patterns to each other, or to other groups. However, to a greater extent than other youth in the study, they succeeded in getting jobs even at an early stage-but often ones that risked becoming dead-end jobs. An important contributory factor seems to be that they had already worked during the school years, mainly in service jobs they obtained through their families' social networks (Lindblad, 2016).

Conclusions and discussion
In line with previous research, the study shows that school-to-work transitions are clearly extended, complex and insecure processes for young people without at least upper secondary qualifications (Dale, 2010;Lindberg, 2012Lindberg, , 2014Settersten & Ray, 2010;Stokes & Wyn, 2007). Our findings further indicate that the length of time spent in upper secondary education affects career patterns and time required for establishment in the labour market, as well as the opportunities available to young people (despite exceptions enabled by individual agency and support, informal and/or institutional).
The 27 young adults' narratives portrayed two different, rather evenly distributed transition patterns. Almost half of them obtained employment directly, often in unskilled service jobs that they had already tried to help their families financially while attending school. However, this was also one of the reasons for failing in school. The other half entered a fragmented transition pattern, spending most of their time unemployed or in youth schemes, with limited choice alternatives and agency (see Evans, 2007). Nevertheless, the narratives also suggest that young adults' careers gradually developed, and their horizons of action expanded. They learned from their efforts to establish themselves in the labour market and their participation in the field. They realized that they wanted and needed to change their position in the labour market, which often prompted them to participate in some kind of education (Hodkinson, 2008). Other important factors promoting changes included informal and institutional support, the family (which strongly influenced, positively or negatively, young people's education and work choices) and associated social networks.
When the 27 young people considered options, structural conditions and individual agency interactively dictated what was regarded as possible and impossible. Social class and ethnicity closely intertwined in their positioning in the socio-geographic space and the difference making they encountered. Gender may also be an important factor, as shown by our finding that the young women had less space for autonomous action than the young men. However, our findings also indicate that, when individual agency was limited, most obviously after failing in school, opportunities for enhancing agency and successful transitions could still emerge, especially if institutional support was available. Access to flexible adult education and other institutional services, e.g. targeted and individually adapted youth schemes and employment support, are thus highly important to avoid 'locking in' low-educated youth in low-status jobs or unemployment. This crucial aspect of careership has however received relatively little research attention, especially in Swedish contexts.

Discussion
As is evident from a host of research (see above), structural conditions, such as availability (or lack thereof) of well-developed and equal services in school, e.g. special education and career counselling, second-chance education, and decent jobs at the labour market for young school leavers, are essential for positive school-to-work transitions. Here we want to point to some aspects that became visible in our study and indicate measures and approaches that may contribute to strengthen the agency of the young people concerned.
In the narratives of the 27 young men and women in our study, the family constitutes a centre of life, and the young adult has to consider its needs and demands. Thus, collective pragmatic-rational processes, rather than just individual ones, may sometimes be involved in career-related decision-making. It may be important to understand, not least for educational actors, that conditions, values and ideas about the world that significantly differ from expected or dominating discourses may influence young people's thoughts about what is desirable and possible to do in the future. The young people's choice to work parallel to school studies is based on such deliberations that are products of a social and historical context. Hence, a better understanding of the life conditions of young people from families with a migration history, and improved relations between schools and parents, could empower students during their school years. In career counselling, such aspects are part of a multicultural approach, founded on insights that clients may have values and conceptions of the world that differ from the counsellor's in significant respects, which affect clients' thoughts about choice: what exists, what is possible and what is impossible (Chen, 2008;Sue & Sue, 2008). Applying such an approach in schools and other institutions that deal with young people is essential, inter alia, treating other's 'life space' respectfully, with interest and questions, in order to understand their conditions and thus facilitate improvements in their position and development (Peavy, 1997).
Furthermore, increasing teachers' and heads' insights into the life conditions of students with migrant backgrounds could be highly beneficial. For instance, there appears to be too little awareness of parents' lack of knowledge about the educational system and language problems, the students' difficulties in getting help and a quiet place for doing homework, and their need to contribute to the family's finances by working. Dissemination of such knowledge is essential for the introduction of effective new solutions, e.g. coaching to improve structural factors and encouragement, and help with homework in school, to prevent dropout and low or incomplete grades. In addition, organized instruction about educational and vocational options may be essential for preparing students to take rational, informed choices for the future, particularly in cases where parents have little relevant knowledge. This is relevant for many students generally, not just those with migrant backgrounds. Our overall conclusion is that prevention of school failure for many of the 27 young adults in this study would not have required much investment in resources and efforts. That is good news.

Notes
1 The chapter emanates from the research project 'Unsafe Transitions' , funded by the Swedish Research Council (ref. no 2009-5964), and Lindblad (2016). 2 Foreign born here refers to being born in a country outside of the EU 28. 3 The falling-off had several causes, only some of which were known, e.g. movement to another country or illness. In other cases, the respondents could not be reached, because either they did not answer at all or contact details were missing (sometimes because of movement to another country). In some cases, the young adults agreed to participate but repeatedly postponed the follow-up interview. 4 However, it is possible that we failed to recruit young women for our study who continued to assist their families with childcare and other domestic work after failing to find a job in the labour market. For example, more female than male non-native-born Swedes aged 15-24 were in the NEET category between 2009 and 2017 (except in 2011) according to Statistics Sweden (2018). 5 Students 'normally' finish upper secondary education at the age of 19, but many continue for a fourth year, e.g. due to a change of programme or attending special upper secondary school. At the age of 20, youth education ends, and young adults who want to complete compulsory and/or upper secondary education are referred to adult education. 6 The increasing housing segregation and its connection to ethnicity in Sweden have been reported and analysed by several authors, e.g. Bunar and Sernhede (2013), Molina (2005) and Szulkin and Jonsson (2007). 7 Our study of 100 young adults shows that 'dropout' is often a misleading concept; most of the respondents attended upper secondary school for three years or longer, but often with high rates of truancy, and without reaching the goals and or leaving with a pass in all subjects.