Compensatory School Effects and Social Capital

: The aim of this study is to explore whether and how school-based social capital (SSC) may increase or reduce inequalities in the academic achievement and well-being of students from different backgrounds (class, gender, and ethnicity). SSC here refers to those qualities of social relationships and the degree of interconnectedness between students, teachers, and parents that can support the educational attainment and social adjustment of young people. As the results of our ethnographic studies indicate, there is a significant association between SSC and school composition—i.e., the class and ethnic background of students in a school. The association indicates the stratification effect of social capital in schools as a predominant pattern. In a school with students from higher socio-economic backgrounds, we observed more qualified and motivated teachers, an intensive parental involvement, and, consequently, more constructive and friendly relationships among students. In contrast, in a school located in a disadvantaged area, the social relations were quite the reverse. We then found a third category. In a school with children from lower social class backgrounds placed in an immigrant-dense area, highly committed school staff were able to create an emotional closeness and trust between them, the pupils, and their parents. School-based social capital in this context introduced a sense of solidarity and created a pro-educational climate.


Introduction
As a public institution, schools are more accessible than family … as an appropriate unit of intervention (Garbarino 1992, p. 121).
As the sociology of education demonstrates (see, e.g., Behtoui 2017), a profound understanding of the differences in the educational attainment of pupils is not to be found exclusively at the 'micro level' in factors like family resources (class background) and the pupils' individual characteristics (e.g., gender, migrant background, or learning difficulties). The 'macro level' of the education system can also generate schools with different compositions and resources, which, in turn, can create varying school environments and can increase or decrease inequality among students in many ways.
In addition to factors at these two levels, one should also consider the significance of the 'mesolevel' factors, i.e., the social relations in schools (Behtoui and Neergaard 2016). The meso-level factors of social institutions (like school or neighborhood) are the immediate social contexts in which young people's education and transition to adulthood occur. As Field (2008) mentions, this tangled web of social relationships with other individuals in these contexts affects and (re)forms young people's aspirations, opinions, behavior, and plans for the future. This social environment of schools (in which young people spend the greater part of their days) is a primary site of social interaction, and is, therefore, a main source of access to social capital for them.
The aim of this study is to explore whether and how SSC-school-based social capital (the mesolevel factors)-may increase or reduce inequalities in the academic achievement and well-being of students with different social class and migration backgrounds. School-based social capital (SSC) is defined in this study as the resources embedded in the standing social networks in a school that may support the educational attainment and social adjustment of pupils (cp. Virtanen et al. 2013). Since young people spend a considerable part of their waking hours in school, most of their social contacts are concentrated in the school environment. In this way, schools are an important setting where pupils develop social networks through interpersonal interaction with both their peers and with teachers/other adults, as well as through an interconnectedness between their parents and school. In line with this definition, SSC in this study is operationalized as those aspects of the social contexts of schools that include: (a) relationships between the students and adults in the school, (b) the pupils themselves, and (c) the parents and the schools, as well as, finally, (d) the rate of students' involvement in school-based extra-curricular activities. In what follows, we first review the various definitions of social capital related to education and previous research in this field. After a short discussion of the methodology, we present our findings.
We find (hardly surprising) a significant association between SSC and school composition-i.e., the social class and ethnic background of students. In a school with superior educational results and students from higher socio-economic backgrounds, we observed more qualified, committed, and motivated teachers, an intensive parental involvement, much less physical and verbal violence in relationships between students, and, consequently, greater school-based social capital. In contrast, in a school located in a disadvantaged area with poor educational results, social relations exhibited quite the reverse. Our conclusion is, then, that SSC is strongly associated with a school's composition. Furthermore, SSC (like other forms of social capital) is more likely to function as a mechanism that facilitates the intergenerational transmission of advantage.
However, we found an example of a third category. In a school with children from lower social class backgrounds placed in an immigrant-dense area, the highly committed school staff were able to create an emotional closeness and trust between staff, pupils, and parents. SSC in this specific context brought about a sense of solidarity, created a pro-educational climate, and operated with counter-stratification effects.

Theoretical Framework and Previous Studies on the Social Elements of Schooling
For young people, school is equally important as an arena for gaining knowledge (hard skills) and for social learning (soft skills) through relationships with other young people and adults (Valenzuela 1999). The social environment of schooling-how the different actors in a school system are related and interact with each other-strongly affects the quality of educational attainment. Consequently, SSC is one of the most central concepts in any understanding of the social elements of schooling and education.
As theories on social capital within an organization emphasize, well-connected individuals in a school, workplace, or community bring a sense of solidarity to all members of these collectives, facilitate the coordination of action, and generate trust as well as common norms and values (Putnam 2015). Resources in the various social networks of these collectives can bring benefits to all individuals embedded in them, as underlined by Portes (1998). The socialization of adolescents in a school context where students feel that they receive fair treatment, where there is an emotional closeness between students and between them and their teachers, where students participate in common affairs at school, and where they feel that the work of schooling is meaningful has a significant impact on their attitudes and behavior. On the other hand, in a fragmented school milieu, when students are divided into rival or unfriendly groupings, where bullying, racism, sexism and violence, excesses, and harassment are tolerated, all pupils feel less safe, the flow of information is limited, and trust is in short supply. In such an environment, people are less likely to trust each other and cooperate (Kawachi et al. 2013). Coleman's (1990) definition of 'social capital' emphasizes the density of social ties in each organization or collective (e.g., a workplace, school, or community) and their capacity to enforce some shared norms. 'Closure' was the concept that Coleman 'used to refer to mutual knowledge and social ties between members of a collective who support each other and sanction deviance' (Portes and Vickstrom 2011, p. 462). Portes (1998, p. 10) explains the source of this kind of social capital-the 'bounded solidarity and enforceable trust'-and maintains that the mechanism behind the effects of collective social ties is the development and enforcement of some specific norms, as well as the improvement of trust and social control among all members of the group.
Thus, for our study, we can define the collective social capital of a school as the quality of the relationship, bonds, and trust between staff, pupils, and parents as well as between the pupils themselves 'that support educational attainment and should have implications for social adjustment' (Parcel et al. 2010, p. 831). Collective social capital, in this case, affects the entire group through an effective flow of information, the enforcement and development of pro-educational norms and aspirations, and the improvement of the trust and social control that ultimately create a more inclusive and cohesive school climate and a greater sense of school belonging (Van Rossem et al. 2015). SSC is taken to be an important dimension of a school's culture, which encompasses common practices, patterns of relationships, and learning, as well as rituals within a school. School culture, in its turn, refers to 'the visible and habitual ways in which meanings are shared' in a school and deals with student and staff perceptions of school life socially, emotionally, and academically (Wrigley et al. 2012, p. 102).
Social capital, as Bourdieu (2001) defines it, deals with the problem of the production and reproduction of social inequality and, like other forms of capital, is synonymous with power. Stanton-Salazar (2001) underlines the stratified character of SSC as the predominant pattern and writes that, in school settings, teachers tend to possess the cultural capital that characterizes the middle classes. Accordingly, they have a strong feeling of affinity and more responsive and tighter relations with middle-class white pupils and their parents, thus favoring them (compared with those from workingclass minority groups). Middle-class parents, furthermore, enter the education system with the knowledge and skills that facilitate their relationship and communication with the school personnel (Lareau and Muñoz 2012;Weininger et al. 2015). Thus, higher amounts of SSC are associated with the socio-economic status (SES) composition of a school's intake (Thrupp 2001). Segregated schools (with a high proportion of pupils with fewer family resources) are 'structured' in ways that impede access to social capital-something that is taken for granted by middle-and upper-middle-class youth (Stanton-Salazar 2001).
Stanton-Salazar (2011) simultaneously draws our attention to the possible counter-stratification effect of social capital. This happens when young people from disadvantaged groups have been able to gain access to resources beyond their immediate (ego) social networks through institutional mediation (teachers and other adults in school), institutional resources (e.g., funded programs), or the leaders of extra-curricular activities outside the school (Behtoui 2019).
To illustrate the impact of SSC, Coleman and Hoffer (1987) wrote about the better educational outcome of students in Catholic schools (compared to public schools) as a consequence of a cohesive, supportive social system. Different types of bi-directional interpersonal relationships in these schools (between students, between teachers, between teachers and students, and between the school and the parents) provide resources that enrich the sources of information and control. Even if empirical results on the impact of common norms have been mixed (see Carbonaro 2006), the hypothesis remains valid.
Gülen-inspired schools for the children of Muslim communities in Turkey and other parts of the world are another example of schools at which pupils achieve academic excellence (Agai 2007). The common faith for students, families, and teachers in these schools promotes trust based on common values, which is useful in supporting educational achievement and pro-social behavior. The 'shared values that inhered in these environments' may be critical to their success (Horvat and Pezzetti 2018, p. 44).
As Valenzuela (1999, p. 28) writes: Positive social relations at school are highly productive because they allow for the accumulation of social capital that can then be converted into socially valued resources or opportunities (e.g., good ratings, a high school diploma, access to privileged information, etc.). Virtanen et al. (2013, p. 74), in summarizing recent empirical findings within this field, write that SSC 'has been associated with better academic achievement among students'. Most recent research on SSC has been through quantitative studies, where one or several features of SSC, social ties, or interconnectedness (as characteristics of social capital) are measured. The networks that connect students, parents, and teachers to each other and the ways in which adults transmit information and norms to youth, together with students' participation in leading school or extra-curricular activities, are among these measures (see Dufur et al. 2013 for a summary of these measures). Summarizing the qualitative research in this field, Dhillon (2015) writes: The main methods used to gather qualitative data [in this field] are observations, interviews, and focus groups, though other methods, such as documentary analysis, diaries, journals, and logs are also common.

The Swedish Context
The objective of promoting social equality and justice through mass education gained priority in Swedish educational policy after the Second World War and remained a top priority up until the 1990s. Swedish schools during this period were characterized by good opportunities for the children of disadvantaged classes (Erikson and Jonsson 2000). This historical experiment demonstrated that schools can make a difference when it comes to rectifying and resolving unequal social conditions. However, since the 1990s, the Swedish education system has undergone comprehensive restructuring, which includes decentralization, privatization, freedom of school choice, and the application of a New Public Management Agenda in state education (Daun 2020). This restructuring, in combination with growing segregation in the housing market, has led to school segregation and to deterioration in the academic performance of Swedish pupils, particularly in marginalized neighborhoods . The increased segregation of students (based on their class and migration background) explains almost the entire and widening gap between the different schools in recent years (see Skolverket 2018 for a summary of current Swedish research).

Data and Methods
The data are drawn from ethnographical case studies that were accomplished in three schools in Stockholm County and that involved participant observation and interviews (Buchanan and Bryman 2009). The schools were selected from a list provided by the Siri database of the Swedish National Agency for Education. We divided the schools located in Stockholm County into three distinct categories-depending on the educational results during the three previous school yearsand chose one from each category. Classes in the final year of compulsory schooling (young people between 15 and 16 years old) were the focus of our study. The average merit rating from Grade 9 (completion of which is central for their application to upper-secondary school or gymnasium) and the proportion of pupils who achieved the knowledge requirements in all subjects were indicators used to determine the position of each school in these categories. As a result, we selected the following three schools to carry out our research: (1) East School, located in a low socio-economic neighborhood, with the lowest final average merit rating that improved their educational results during this period; (2) South School, a school located in a low socio-economic neighborhood that had the same stable poor educational performance during the preceding three years; and (3) North School, from the category of those with the most stable, best educational results.
Combining participant observation and interviews allowed us to gain a multifaceted insight into everyday practices in the schools. Data collection for this project was carried out in different periods between 2016 and 2020 by the two authors of this article. The project was approved by the Swedish Ethical Review Authority. As specified by the general standards for ethical research, all participants were informed about the purpose of the project prior to taking part. In all disseminations of the findings, the names of participants and distinguishing features of the schools are changed or omitted.
We were particularly interested in the way in which each school rose to the challenge of supporting the educational progress of pupils and the measures that had been taken to achieve this aim. Post-lesson (formal and informal) interviews with pupils and teachers and participant observation were used to find patterns of relationships between the different actors (Dhillon 2015). Observation took place both in classrooms and outside (in communal spaces), during lesson and break times, in staff rooms, and at parental meetings, and we followed the young people during their extra-curricular activities outside the school.
The formal interviews followed a semi-structured template and were taped. We took notes about our observation of the various forms of interaction in these different spaces during and after each day of observation. The opportunity to participate in several different environments over a long period enabled us to gain a broad and multifaceted understanding of the relationships between school staff and students (Beach 2010). Combining participant observation and interviews allowed us to carry out a multifaceted exploration of the everyday practices at each school and to gain an understanding of how social relations affect the social bonds and trust between school staff, the pupils, and their parents.
Field notes and interview data were then coded (with the QDA software NVivo 11) for the different themes of central importance to our research questions (Saldaña 2015). Firstly, we classified the data according to: (1) where the interaction took place (e.g., in the classroom), (2) between which actors (e.g., teachers and students), and (3) which subjects they addressed (e.g., emotional support). Various themes subsequently emerged from the data material-for example, "the aesthetic, architectural, and sensory dimensions of schools", "student absence rate and records", and "interactions between pupils".
In the following sections, we outline our ethnographic findings from the three case studies, with our main focus being how social relationships shape the quality of SSC in each setting. The following themes were in focus for this paper: neighborhood characteristics and school results, staff turnover and stability, teachers' opinions on the characteristics of pupils and their parents, physical or verbal violence, and pupils' influence and participation.

South School
South School is located in a marginalized suburb in the south of Stockholm County. Only 30 percent of the inhabitants in the age group of 20-64 years have post-secondary education and an average annual income 290,000 SEK (around 27,500 EUR). Over 70 percent of the residents were born abroad, the majority in Iraq and Syria.
During our fieldwork, about 500 students were studying at this school (from Grade 1 up to Grade 9). Their final average merit rating for Grade 9 (210 points out of a maximum of 340) was lower than the average rating of Swedish lower-secondary-school students. Only 60 percent of pupils achieved the knowledge requirements in all subjects and were qualified to enroll in the national secondary school program. Students' educational performance was stable over the three years of observation (and with the same poor results).
The concept of continuity cannot in any way be associated with this school. For three years, we witnessed two principals leave, the second one telling us 'I guess only 25 to 30 percent of the teaching staff here are the same as three years ago'.
Joni, 16 years old, described how their class had had several teachers in natural science subjects and that each of them taught in their own way. Joni's classmate Angelica confirms this and continues, 'We had to adapt to each new teacher's way of instruction, and it is not an easy job-some of us give up'.
The school's principal, Birgitta, highlighted both the school's and the neighborhood's poor reputation as the cause of this extreme turnover of staff: '[People think] that it is difficult to work here. They fear that it will be noisy and quite violent, with unmotivated students and an unpleasant situation'. With the arrival of the latest principal, frustration arose among a group of committed teachers who felt that their efforts were no longer valued. They chose to leave the school, thus worsening the situation even more.
The high turnover of school staff is partly due to their negative attitude: 'I will leave this school as soon as I get a chance'. There are also some teachers who get an offer of a higher salary from other schools and therefore leave. A further problem is that, in general, schools have difficulty in hiring qualified teachers. Some of the staff (particularly in marginalized areas) have no teaching qualification and/or do not master the subject that they teach. They are thus not a stable group. We met several such teachers at South School who had applied for jobs here because they could not get one anywhere else. One of them (a math teacher), did not master her subject at all. It was shocking to sit in her classes and watch her giving incorrect instructions to the class and confusing the students.
Nonetheless, there were some teachers who wanted to work there, but found themselves confronted with a paradox between their social engagement as a teacher and their private interests. Janin was one such an example-a science teacher who was very popular with his pupils, who described his lectures as very instructive and stimulating. His students were dreadfully disappointed when he abruptly left the school after a year. Janin left South School, as we realized later, for a private school with high-performing students, which offered better pay and working conditions. The Swedish Schools Inspectorate had repeatedly made several serious criticisms regarding mistreatment at this school-e.g., the lack of a possibility for pupils to study in peace, the need for them to work in sometimes violent surroundings, the lack of a proper library and computer room, and the lack of support for students who needed it (note that the majority of the students lived in overcrowded housing estates and had difficulty studying at home in the afternoons). The school's failure to resolve the problems would lead to it receiving a final warning of closure if nothing was done.
Verbal and physical violence was really one of the most serious problems, a constant and normalized phenomenon in the school milieu. Some of the school staff took the attitude that the violence was not a big issue, saying that 'boys will be boys' and that boyish pranks were nothing serious. When, for instance, we were going along the school corridor and saw a boy aggressively pulling a girl to the floor by her hair, the teacher whom we were following rolled her eyes and sighed 'They are playing'.
We heard from both parents and school staff of their desire for greater and better communication between home and school. Rosalin (the Spanish teacher) told us that the parents were uninterested in what was going on in the school: 'They never call and do not come to parent-teacher meetings'. Welfare officer Helena said, however, that it is difficult for the parents to come to school during their working hours. However, as we observed, it was difficult to attract parents to the general meeting that was held when the new principal (Magnus) invited them-only 40 of out of 500 participated. One of those parents (Anwar), who did not participate in this meeting, told us later: I have four children in this school, I have tried to get involved but I experienced a lack of response and respect from their teachers. They assumed that we are ignorant and uninterested, which is not true.
Elif, a student, pointed out the language problem and said: Often the children have to act as interpreters for their parents during the performance review. It creates a strange situation where the teachers turned to the students instead of to their parents.
Regarding pupils' influence on the school's organization and extra-curricular activities, we can report that most of the activities like music or sport worked on the basis of an ad hoc model and were discontinued after a time. One example was the planned online school newspaper (a serious request from a group of students). The English teacher, Ayla, was supposed to lead it, but she had no experience with editorial work and did not understand why she should be the one to do the job. A dozen enthusiastic students participated in the first meeting and had many ideas for what they wanted to write about. It turned out that they did not have access to computers and, therefore, had no opportunity to start the work. Ayla's half-hearted interest was an additional reason to not continue with the plan.

North School
North School was built in the 1920s in an affluent suburb of Stockholm County whose inhabitants had an average annual income of 440,000 SEK (circa 42,000 EUR) and of whom 66 percent had postsecondary education. About 90 percent of the population were native-born Swedes.
During our fieldwork, North School had more than 650 students enrolled in Grades 6-9. The final average Grade 9 merit rating was 285 points, and 92 percent of students achieved the knowledge requirements in all subjects and were thus qualified to continue on the national secondary school program.
Most of the staff had worked there for some 15 to 20 years. The school's principal (who had held this position during the last 17 years) was enthusiastic about the further education of the school's teachers. There were a remarkable number of teachers with special educational skills in the school. Maggie, one of them, told us: 'During a particular period, I had the privilege of working only 75 percent (with full pay) in order to have time to devote to my further education studies'. Part of teachers' regular working hours was used for activities other than teaching. For example, 30 percent of the social-science teacher's working hours were used to run the school's student magazine and 20 percent of another teacher's work was to run the anti-bullying campaign at the school.
Since the school's pupils come from affluent and well-educated families, they received support with their homework either from their parents or through hired study help. However, pupils in North School were able to choose to participate in the after-school homework club supervised by a teacher. Alicia, a ninth-grader, appreciated being able to take part in this activity, even though she also had hired help with her studies at home. The math teacher, Paul, who was responsible for the homework club, told us that ten trained special educators were working with him. Karin, another teacher involved in this activity, stated that: Most students in our school attend theoretical (university preparatory) programs in uppersecondary school. There are only a few who have chosen vocational programs. Therefore, the need for help in certain subjects, like math, is considerable.
The teachers (particularly the class mentors) were always vigilant in determining whether there were pupils who had problems at school or at home. Physical or verbal violence and bullying were not openly visible in the school's communal spaces, even though a number of our interviewees said that, on occasion, some pupils had experienced subtle forms of abuse like freezing out or glaring looks. According to the principal, in one period, they had problems with online harassment. When we asked the pupils about their experience of the school as a social setting, they replied: Our school's environment is supportive and encouraging; we are seen and confirmed. However, school for us is a place for study-and not primarily a place for socializing.
The parents have a great commitment to the school and are in close contact with the staff. Parentteacher meetings were attended by virtually all parents. Each term, a 'class parent group' is set up to support each class both during and after school hours. They had regular meetings-at least three times a term-and sometimes went to the school to sit in on lessons. Networking with each other, going to parent councils, taking night walks in the area, arranging activities for the class, and planning and participating in bigger class trips were some of the tasks that these parents took on. Cohesion between pupils in the school was attained through these activities-which were effective ways to construct a sense of 'us' among them (e.g., organizing cultural activities like theatre or museum visits, school trips, and festivities arranged with the help of the parents). These latter had money, time, and large houses (sometimes with a pool) and were able to arrange parties (e.g., for birthdays), and were thus helpful factors for increasing the cohesion of the collective. Finally, we noted the queues outside the school welfare officer's room, which indicated that there were children who had mental and social difficulties and wanted to talk to her. Teachers in this school had a traditional and authoritative role. Each class thus had a specific teacher (mentor) with overall responsibility; the mentor went through the homework and exams in each subject every week, kept in touch with the parents, and followed up on the students if something went wrong. Malin, 15, described their own class teacher as one with an extra-close relationship with her pupils, one who knew them each individually. Every student had a specific seat in the classroom, which was determined by the teacher and changed at regular intervals.
The parents placed high demands on the school and on the education provided. As mentioned above, they occasionally sat in on some lessons to monitor the teaching in their children's class more closely. The teachers spoke of countless occasions when grades set by them were questioned or when they received angry emails from dissatisfied parents. The parents were also not late in contacting outsiders, such as experts in certain school subjects, if they felt that the pupils had received an incorrect assessment.
At North School, there were several extracurricular activities. The school's newspaper had Tomas (a social-science teacher) as its editor. He dreamed of becoming a journalist when he was younger and said that the school's newspaper had become a way to realize that dream. Each term, a group of students in Grades 7 and 9 worked with him and wrote about their school. The budget for the newspaper was provided by the school. The newspaper had a pronounced identity-creating function and had received several awards for its work.
The school's choir was another activity run by an enthusiastic music teacher. They performed regularly on special occasions, school trips, and other activities arranged by the students themselves. The anti-bullying group (run by the student council) consisted of two students from each class who were selected by their classmates during the first year of school. We also heard critical voices saying that class representatives were mostly a symbolic role. William, a ninth-grade student, pointed out that the student council's work was not very visible other than on these special occasions. Christina, another student, told us that the class representatives had mostly a decorative function.

East School
At East School, there were 220 students in Grades 6-9 located in an extremely deprived suburb of Stockholm County, where about 85 percent of the inhabitants were foreign-born and 36 percent had post-secondary education and an average annual income of around 220,000 SEK (21,000 EUR). In their first year (three years before), only 60 percent of pupils achieved the knowledge requirements in all subjects and were qualified to go on to secondary school and its national program. The following year, the figure increased to 67 percent, and now is about 75 percent, with 240 points as the final Grade 9 average merit rating. Most of the students in the school come from countries marked by political turmoil, like Somalia, Iraq, and Afghanistan.
The school is a very new one, started by a group of well-educated young people with an immigrant background in a bid to help adolescents with the same background in the neighborhood. 'We started our life here in this neighborhood as well, and, now that we are established professionals, we should pay something back', the school's principal (a woman in her 30s who had studied law) told us. She had recruited people like herself as school staff, committed and enthusiastic young people who saw a noble meaning in their work in the school. Their dedication was very visible when we participated in their staff meetings. The subjects of discussions were those pupils who had received low grades, those who had been absent for more than 10 percent of school hours, and ways to help parents to fill out and send the application to a charitable foundation for financial support for buying clothes.
There was a pronounced zero-tolerance policy regarding physical and verbal violence through a number of measures. These latter included the presence in the corridors of several teachers at break time, talks with those who violated the rules, the arranging of extra teaching on bullying and social relationships for those classes that had problems, the inclusion of feminist education as a part of the curriculum, and contact with families in the case of problems.
When a pupil begins studying there, the principal and a member of staff visit the pupil's home to talk with the family. The principal, Roya, told us: There is deep suspicion of persons in authority among the population in this area because of their experience of long-lasting racism and disrespect. Therefore, it is extremely important for us to demonstrate our respect and to present ourselves as people like them, to ensure that we have a common interest in the well-being and success of their children.
In this way, the school staff has been able to cooperate with and make use of all of the families' resources. They have close contact with them (through social media, phone calls, and meetings) in order to keep parents informed on how well their children are doing in the school and what they need to do in order to graduate with good grades. Each mentor has good knowledge about the academic progress and key aspects of the personal lives of her/his pupils through the regular performance review meetings with the parents. Our presence in parent-teacher meetings enabled us to observe at first hand the active participation of both parents and teachers as well as the warm relationship between them. Contact between these actors is close and continuous.
The teachers of subjects that some pupils notoriously have difficulty with (e.g., math, natural sciences, and languages) work 70 percent of their hours during the week in order to be able to give extra teaching on Sunday mornings. 'Since they live in overcrowded housing conditions and lack enough money to hire private teachers to help them, these 'Sunday classes' are extremely valuable for them', a natural-sciences teacher, Somaiya, told us. Sitting in one of her classes, we could see a restless boy who was disruptive and talked while she was teaching. After the lesson, she asked him to stay behind in the classroom. During their conversation, fighting back her tears, Somaiya referred, in a motherly tone, to a previous agreement they had, repeating over and over again: 'You have let me down'; embarrassed, the boy promised 'It will not happen again'.
The school personnel are in close contact with the police and social services in the area in case pupils get into trouble outside the school. The deputy principal, Nora, told us: One of my students was among a group young people who had been arrested by the police for shoplifting; she called me for help because she was afraid of her family's reaction, and I went there as quickly as possible to help her.
The teachers try to adjust the curriculum to the students' backgrounds and needs. For example, one of the tasks of the pupils during a Swedish-language lesson was to write a hip-hop song about 'my life, my dreams', 'because hip-hop is popular among these kids', the subject's teacher, Catharina, explained.
One frequent activity is study visits to various universities and colleges in Stockholm. 'Our purpose is to encourage our pupils to apply for further education and continue studying at university. For many of them, studying at university is not a matter of course', the school's career adviser, Aisha, told us. They also arrange regular school-sponsored trips for Grade 9 students, although, considering the somewhat tight budget, these trips are normally to cities in the immediate surroundings. The sports teachers make use of all the sports facilities in the neighborhood to teach their pupils to swim and ice skate, and arranged a one-day skiing experience (which the parents are not able to do). With the help of volunteers from the 'Save the Children' organization in the neighborhood (all of whom have an immigrant background), a termly 'girls evening' and 'boys evening' is organized at which the pupils talk, among other things, about social relations in the school and their homes.

Summary and Discussion
The successful adolescent transition to adulthood in current complex societies requires extensive social and emotional support. School children need the various resources provided by their social relationships with their proactive parents, committed school personnel, and pro-academic peers. The social capital supplied by these 'significant others' helps them both to participate simultaneously in multiple socio-cultural worlds and to go through the various stages of developmental transition (Stanton-Salazar 2001). Those who are embedded in resourceful social networks manage to get ahead and achieve a substantial measure of self-realization. School (besides the family) is one of the most important arenas for developing and providing adolescents with informational, emotional, and social support, which we have called school-based social capital, or SSC.
As our fieldwork data demonstrated, SSC in the South and North Schools operated in line with the production and reproduction of social inequalities. In the case of the South School, we observed a high turnover of school staff, which resulted in the recruitment of teachers with lower academic qualifications and seriously harmed the school environment and students' performance. The staff's high turnover was also due to the school's low status and negative reputation, as expressed by the media and through statistics published by public authorities. More qualified teachers avoid such schools and prefer to work in those with socio-economically higher pupils, because working in segregated schools in disadvantaged areas is a demanding task. Furthermore, the majority of the members of staff who do work in these schools have a negative perception of the pupils, their parents, and the neighborhood, together with low expectations of their students' academic performance. This leads to an unfriendliness, a feeling of unease, and sometimes hostile relations between staff, students, and their parents. The exclusionary attitude of the South School's staff towards the pupils' families is based on notions such as the latter's lower social class, immigrant status, and country of origin in the Global South, as well as on the criminal stigma of the neighborhood.
Families' limited resources and the low status of the community resulted in fewer resources for the school, which, in turn, negatively affected school-based social capital. Lower SSC, in turn, more negatively affected the educational achievements and well-being of students. In brief, inequality in home resources was not only compensated for by school resources, but the school context (as a consequence of school resources) also accelerated the initial inequality.
In the case of the North School, the quality and stability of the staff (including the number of certified teachers) and the development of trust and collaboration between them and parents were key factors explaining the presence of a high SSC and the observed high educational performance of pupils. High SSC-strongly associated with the socio-economic status (SES) of the school's intake of pupils and school resources (the quality and stability of the staff plus interventions of parents)-in turn, positively affected the educational performance of the students. The staff were extremely keen to create a feeling of 'us' for each class and of belonging to the school as a whole. Parental resources were of great help to the school, enabling teachers to accomplish their tasks by creating a supportive social environment for pupils and facilitating their successful educational performance and maintenance of good mental health. Parents in this school (like other similar middle-class schools) were engaged, organized, and endeavored to share responsibilities with school staff (cf. Lareau and Muñoz 2012). On the other hand, school personnel, with their middle-class values and positions, can more easily communicate with upper-and middle-class families than with those from a workingclass and/or migrant background. Horvat and Pezzetti (2018, p. 46) correctly remind us of 'the unequal distribution of power across race and social class' and explain the inherent disagreement between school staff and families from lower strata as being due to the power imbalance of people from different social classes and positions.
To summarize the reproduction effect of SSC, we can say that the negative effect of lower socioeconomic position of parents (fewer family resources) on the children's educational achievements would be even stronger in schools with less SSC. Thus, the children became subject to a double handicap. On the contrary, children from an advantaged home will do even better in a more favorable school environment (with high SSC). These two cases demonstrate that schools are not neutral sites of knowledge provision nor meritocratic springboards for the social mobility of young people with different backgrounds.
However, school should not inevitably intensify children's initial social class inequalities. Our East School case demonstrates that schools can compensate for the fewer family resources of pupils from lower social classes and their early difficult experience of schooling by providing positive SSC. This has been achieved by strengthening the emotional tie between the student and the teacher, helping students to acquire a sense of being valued, belonging, and hope, maintaining a respectful relationship with the parents, and creating a positive school climate in which staff, students, and their families work together for the common goal of supporting students' social development and academic learning-this in spite of a rarely strong bridge to mainstream life for non-mainstream children and families in segregated neighborhoods and schools. The social capital provided by school staff, peers, and family (through warm and trusting relationships) socializes pupils from lowerincome and minority backgrounds in a psychological motivational orientation that facilitates their academic learning and prepares them for mainstream adult life. Of great importance is the emotional bond between the student and the staff in such schools. Each experience of support from the adults in a school helps the students to acquire a sense of being seen and valued. Pupils' development and learning take place through their interaction with school staff in a warm and supportive environment. The school personnel's role as compensatory sources of informational, emotional, and personal support has altered these students' social networks and their parents' attitudes towards teachers. SSC of this type has allowed young people to maintain high self-regard and high levels of academic motivation, which, in turn, provide them with socially valued resources or opportunities for the future.
Moreover, the above-mentioned achievement is not restricted to the social mobility of an individual. As this case demonstrates, students with a working-class and/or migrant background can perform well if they have access to schools where there is a strong SSC to support them. The social support provided by the school staff through warm and trusting relationships can socialize pupils from lower-income and minority backgrounds into a psychological motivational orientation that facilitates their academic learning and prepares them for mainstream adult life. Each experience of care and help from adults in such schools helps students acquire a sense of being seen and valued and a sense of belonging.
Schools such as these overturn the dominant discourse of the 'supposed innate inability to learn' of young people from these groups. To be able to overcome schools' reproductive mechanisms of inequality, the institutionalization of practices such as those found at the East School can partially compensate for any initial differences and decrease inequalities. Such school environments promote pupils' development, facilitate their academic learning, and prepare them for future life in a counterstratification manner. Even if it is more likely that school-based social capital will function as a mechanism that facilitates the intergenerational transmission of advantage, in specific contexts, it can nevertheless function with counter-stratification effects.
One important limitation of our paper (like other case studies) is that it generates more questions than definitive answers. For example, are there other ways for schools in marginalized neighborhoods to cultivate leadership and develop an adult community? May the development of a more intensive relations with parents provide a strong SSC to support pupils with fewer family resources? Or, how well do students do after they graduate from the East School and advance to upper secondary school? Further research should explore these questions.