Social Exclusion and School Achievement: Children of Immigrants and Children of Natives in Three European Countries

Social resources are crucial for educational success: Social exclusion in the classroom can severely compromise access to those resources, which can have detrimental effects on educational success and thus, in the long run, socioeconomic attainment. This applies in particular to groups that are vulnerable to begin with, such as children of immigrants. Strategic knowledge about the education system, about manners, values and cultural codes that prevail at school, and access to networks of support and inspiration that promote learning and performance are typically less available to children of immigrants¹ than to their ethnic majority peers. Their parents, migrants themselves, are more likely to lack these types of resources, rendering the existence and quality of social relations and connections at school more critical and decisive for their educational career. If ethnic minority children are able to establish a network of support and resources for themselves within school, this can counteract those disadvantages. However, if ethnic minority children are socially isolated and avoided, these types of support networks are less likely to be established. This inequality of network resources then contributes to the perpetuation of already existing ethnic stratification patterns. Thus, the main question that this paper seeks to answer is: Can social exclusion help explain ethnic differences in school achievement?

While it is not a new idea to consider social network resources (especially Lin 1999, 2000), the fact that social exclusion is difficult to measure adequately on a large-scale basis has resulted in a scarcity of empirical research on its consequences. Through the utilisation of social network data, this paper adequately measures social exclusion in the form of social isolation (absence of friendship) and social avoidance (negative relationships). Using the CILS4EU data (Children of Immigrants Longitudinal Survey in Four European Countries), the analysis is furthermore able to test this dynamic in several national contexts. This study uses sociometric data, i.e. social network data, on 15,017 students in 731 classrooms in England, Germany, and Sweden. These countries differ in their institutional arrangements and in patterns of ethnic educational stratification, and this variation in contexts is particularly valuable for testing the occurrence of general mechanisms. The high number of networks and the possibility of generalising the results to populations in several countries are exceptional due to high costs and effort associated to the collection of social network data.

Research in social stratification has long focused on diverging educational success, both attainment and performance, with the traditional predictor being the socioeconomic background of the family. Following increased immigration rates in many Western societies, recent research has increasingly considered ethnic inequalities in educational success (e.g. Heath et al. 2008). While research on first generation immigrants has singled out language skills and pre-migration educational qualifications as obstacles to achieving higher education (van Tubergen and van de Werfhorst 2007), a lot of recent research on ethnic inequalities has focused on the second generation, that is, children born in host countries to immigrant parents, where other explanations need to be sought. These results show that the socioeconomic origin of immigrant groups is behind much, but not all, of the overall poorer performance of ethnic minority groups (Heath and Brinbaum 2014). A potential explanation for the remaining minority disadvantage is the social dimension of youths, as expressed in their social networks, friendship patterns, and social exclusion such as bullying and social isolation. Such a focus acknowledges that the role of peers becomes more significant during adolescence (Osterman 2000; Youniss and Smollar 1985). Since adolescents spend the majority of their time at school, the school class is a particularly significant social context for them, and as a societal ‘micro-system’ the school class harbours both positive and negative peer relations. The aim of this paper is to analyse whether negative relations may lead to disadvantage in school achievement, and whether this more severe to the ethnic minority group.

When it comes to negative social relations at school, most research has focused on bullying and its negative effects on a variety of outcomes (e.g. McDougall and Vaillancourt 2015), however, subtler forms of negative relationships are likely to play a role, too. In particular, isolation, i.e. having no friends, and being avoided by others are forms of a non-positive social environment in the class, which have not been researched as extensively as bullying. Both forms are less explicit than bullying, but can nevertheless be detrimental when it comes to individual outcomes, including educational success.

There are some factors suggesting that ethnic minorities are at a higher risk of being socially excluded. Social misfit theory proposes that individuals are avoided when they are different in one or several regards (Wright et al. 1986), which could be the ethnic or cultural background. Furthermore, recent studies find that youths with a migration background in Sweden are not only more likely to be victimised (e.g. Vervoort et al. 2008), but are also at a higher risk to face subtler forms of exclusion (Plenty and Jonsson 2017). Work on friendships and friendship networks, has shown repeatedly that friendship is more likely and more stable between individuals that are similar to each other, a phenomenon known as homophily (McPherson et al. 2001). This applies in particular to ethnic background (Boda and Néray 2015). Thus, depending on the ethnic composition in the classroom, ethnic minorities can be at a higher risk of not being able to form or initiate stable friendships if there is not much opportunity of befriending someone with a similar ethnic background.

While the need to belong is a fundamental human need that applies to people of all ages (Baumeister and Leary 1995), it has been found that belonging and social acceptance is particularly important for adolescents’ well-being (Brown 2004), relative to adults or children. Previous research has documented the negative effects of low mental well-being on educational outcomes (Nakamoto and Schwartz 2010), but there can also be direct effects from being isolated and avoided on academic achievement, since social exclusion will reduce the chance of learning from others, doing homework together, or drawing on friends’ parents for academic support, such as strategic knowledge on how to navigate the education system (Erikson and Jonsson 1996). These social network resources can be conceptualised as quantity and quality of resources that an actor can access through their location in a social network, i.e. through the people he or she knows, and who these people know (Lin 2000). Network structures themselves can also lead to more stable beneficial connections, e.g. through being embedded in close friendship groups (Block 2015). These types of resources have been found to mediate part of the effect of social origin on educational and occupational attainment (e.g. Jencks et al. 1983), and may counterbalance disadvantages created by lower parental socioeconomic resources. However, access to beneficial social resources is often also compromised by initial inequality, which can aggravate these already existing inequalities. Since personal networks tend to be segregated along a variety of personal characteristics, such as gender and age, but also socioeconomic status (McPherson et al. 2001), there is an increased risk for initial inequalities to be reinforced: Those who are already at an advantageous position will be able to promote this more due to access to the social resources that this network position carries, and those who are already disadvantaged cannot exploit their social contacts to the same extent, and are left behind even more (Lin 2000).

Classrooms, in which children and adolescents spend a lot of time, provide a potential opportunity for gaining access to social network resources. Although classrooms tend to be sorted by ability and parental background, due to residential segregation or in some countries due to a tracked education system, they are still an important environment for associating with others in the same life situation but with a potentially different background: ethnic as well as socioeconomic. Therefore, school classes have the potential to provide a platform for the access of valuable social network resources, but social exclusion can inhibit the access to such resources.

There are several different ways in which pupils can be socially excluded, such as isolation and avoidance. Isolation here is understood as the absence of personal links (e.g. friendship ties), and can either be understood with reference to everyone in the classroom (complete isolation), or just with reference to the majority students in the classroom (isolation from majority students). Avoidance implies negative relations, such as dislike.

With the utilisation of social network data to measure social exclusion on a large scale, this analysis manages to explicitly test a mechanism that has received a lot of attention in theoretical works, but which has not received much empirical attention, due to lack of suitable data.

The main finding of this study is that social exclusion is related to lower grade point average, which applies to both majority youths and those with a migration background. Although various concepts of social exclusion were considered, and a tendency for a negative association with GPA was seen throughout the analysis, the only significant results were found for those measures that capture exclusion from the majority students in the classroom: being isolated from majority youths, and being avoided by majority youths. This implies that access to “native social resources” matters more, which is in line with theoretical considerations on the beneficial knowledge on manoeuvring the educational system that majority youths are more likely to possess, relative to those with a migration background, as well as to the value of classmates’ help with the school work. The possibility of reverse causality cannot be excluded here, so that the association reflects the avoidance of classmates who are underachievers, possibly because of their behaviour, but the analysis manages to control for a number of items indicating school rejection, such as skipping classes and arguing with teachers. The analysis did not show, however, that being socially excluded is worse for migrants, although this test had to be crude to maintain statistical power.

This study set out by asking whether social exclusion can account for some of the disadvantages in grades that children of immigrant origin experience in relation to their peers of majority background. Although the findings show some signs of this, the impression is that this hypothesis was not supported. While exclusion is important for grades, and more children of immigrant origin are socially excluded, the relations are not strong enough to explain the grade difference that was found.

There are limitations to this research. While it is assumed that the mechanism that is investigated here is a general one, the analysis could only include data from three countries, and all of them are European. Ethnic composition of the migrant population varies due to different migration histories of the countries. Since social network data is not a standard feature of large-scale cross-national surveys, future analyses that consider other national contexts depend on the collection of more data of this kind.

Furthermore, this analysis can only point to correlations, while the process linking social exclusion to achievement is likely to start before youths entered secondary school. Collecting and analysing longitudinal data on significant peers and the family from an early age would rectify this limitation. Lastly, more research is needed on the actual mechanism on how social exclusion works, to support and help interpret the initial results that are found in this paper. Next to limited access to social resources, lower mental-wellbeing is another potential pathway by which social exclusion could negatively affect school grades.

However, there are various reasons that render the framing of this study plausible and its findings valuable. While the cross-sectional analysis limits the claims that can be made, there are strong theoretical foundations that build on various strands of previous research, as presented in the literature review. The social dimension is important to include in considerations of educational success, and students who are socially excluded are likely to lack valuable resources, which plays out negatively in their educational success. Most of these combined theoretical implications have not been tested explicitly in the past, due to unavailability of suitable data. Using a large-scale cross-national survey that includes social network data, and can thus accurately measure social dimension of educational success, renders this study particularly valuable.