Parents’ years in Sweden and children’s educational performance

Over the last half century, the proportion of second-generation immigrants in Sweden has risen from being virtually nonexistent to approximately 10 % (Westin 2003; Bengtsson et al. 2005). While considerable country of origin and gender variation exists, the schooling experiences of second-generation immigrants are generally poorer than among children of native-born Swedes, on teacher-assigned grades as well as on national and international standardized tests (OECD 2010; Jonsson and Rudolphi 2011). This study seeks to understand to what extent immigrant parents’ incorporation experiences, here measured as parents’ years since migration (PYSM), affect the educational performance for their second-generation immigrant children.

According to assimilation theory (Alba and Nee 2003), time and generation in a host country will lead to linguistic acculturation, increased institutional familiarity, and increased economic and social capital. These processes could allow immigrant parents to be more comfortable and capable at influencing their children’s educational performance. Though previous studies consider the effect of an immigrant’s time in the country, measured as age at migration, on their own outcomes (Böhlmark 2008; van Ours and Veenman 2006; Goldner and Epstein 2014), fewer have explicitly considered whether time in the host country could be associated with intergenerational effects. Though second-generation children will have resided in the host country for the same number of years as their native peers, their immigrant parents will be characterized by widely differing exposures to the destination country. More specifically, immigrant parents vary in their amount of exposure to the host country language, culture and norms, and, as a result, their level of incorporation. The parents’ accumulated incorporation experiences in a host country could influence their ability and willingness to assist in their children’s schooling and, as a result, impact their children’s school performance.

There are several potential pathways as to how parents’ accumulated incorporation experiences might impact their children’s educational performance. The clearest example of this is a direct relationship through parents’ linguistic acculturation and children’s language proficiency (Bleakley and Chin 2008; Casey and Dustmann 2008), which then extends to their school performance and, specifically, their language class performance. Indirect mechanisms of parental incorporation that could influence their children’s educational performance include increased economic and social capital, familiarity with the school system (Kristen 2008; Turney and Kao 2009), and coming to view migration as a permanent rather than temporary move (Dustmann 2008), all of which would lead parents to being more able and willing to involve themselves with their children’s education. This impact could lead to an increase in performance in those subjects that are less host country specific. In Denmark, for example, Nielsen and Rangvid (2012) found evidence that a mother’s time spent in the country helped with their children’s Danish performance, while additional time for the father helped with their math performance, which may suggest that both specific and general mechanisms of incorporation can be transmitted across generations.

To study the impact of PYSM on their children’s educational performance, we use the Swedish Interdisciplinary Panel (SIP) database, which contains registry data on the full population born between 1973 and 1995. We analyze second-generation immigrants who graduated from Swedish primary schools between 1989 and 2011, generally occurring at age 16. We consider class grades and test scores in Swedish and math due to the varying degree of Swedish-specific skills required by these two subjects.

Typically, estimating the impact of parents’ time spent in the host country on their children’s educational performance using either longitudinal or cross-sectional data represents a considerable challenge. This is the result of selection and omitted variable bias in the relationship between when individuals migrate, when they have children, and their preferences for their children’s educational attainment, putting analyses at risk of getting (upwardly) biased results (Chiswick and Miller 1995). In attempting to avoid this issue, we exploit the linkage in the data of parents to their children, allowing for a family fixed-effects approach. In doing so, the influence of all time invariant observed and unobserved family level characteristics are cancelled out, allowing us to remove the likely source of bias.

Using multinomial logistic regression with family fixed-effects and stratifying our analysis by parents’ country of origin over nearly three decades of second-generation cohorts, this is the first study to consider in detail how the effect of parents’ time in Sweden on children’s educational performance varies by parents’ background. Analyses done on students’ course grade and standardized test scores show an impact of PYSM on grades in Swedish, but not math, across nearly all country of origin groups. That the relationship is found for Swedish performance, but not math, points towards the potential effect for intergenerational transmission of parent’s linguistic acculturation. The neutral relationship between PYSM and math performance across all immigrant groups suggests that general effects of parent’s incorporation, such as increased familiarity with the school system and social capital, are less important for children’s performance, at least as measurable between siblings. Parents’ incorporation, then, can be said to have the potential to affect their children’s Swedish, but not math, performance. Though this is true across all immigrant groups, this finding is perhaps most notable among those groups from Africa, the Middle East, and South America, for whom incorporation has been deemed “less successful” and Swedish language and culture are most distinct.

The data analyzed comes from the SIP-ENTRY database. SIP-ENTRY contains longitudinal data on the entire Swedish population born between 1973 and 1995, as well as their parents and siblings born outside the main sampling window, followed until 2012. Information on attained primary school grades and test scores is available from 1989 to 2011 and therefore constitutes the key period of interest. SIP-ENTRY also has information on demographic characteristics, including date of migration, of importance to accurately define an individual’s number of years since migration. Through the addition of the multigenerational register, family IDs have been created via linking parents to their biological children. This represents a particular strength of this study as a means of isolating the effects of incorporation and dealing with unobserved individual heterogeneity (Lawlor et al. 2009). Our study sample has been designed to consist of second-generation immigrants, for whom the birthplace and date of immigration on both biological parents are known. Furthermore, the individual must have a reported final course grade in Swedish and math, as well as on national tests for the same subjects. Lastly, this information must also be available for at least one other sibling, in order for family fixed-effects models to be estimated, as explained in more detail in Section 3.2.

The results on Swedish class grade performance show evidence of assimilation across all immigrant groups. This result provides additional support for the relationship between parents’ language ability and children’s language performance (Casey and Dustmann 2008; Bleakley and Chin 2008) but using a more causal approach across a broader range of immigrant groups than has previously been considered elsewhere. One might expect children’s school performance to improve as a result of parents’ language proficiency and their general willingness and ability to become involved in their children’s education. By focusing on just Swedish grades, however, it is impossible to say whether this is the result of language acculturation alone or in combination with other general effects of parents’ incorporation. By looking at math grades, for which parents’ Swedish-specific skill may not be as important and not so directly transferable, we attempt to get a better idea of the role of general mechanisms of incorporation. The results from the math analyses show no relationship between PYSM and math performance for any country of origin group, suggesting that the general effects of parents’ incorporation are not sufficient to impact their children’s math performance. This constellation of results suggests that parents’ time in Sweden does have an impact on their children’s educational performance but in specific rather than general ways. Areas that children benefit the most from parents’ experiences are those that are more directly transferable, namely language proficiency. Although no positive relationship is found between PYSM and math performance, this would not necessarily suggest evidence of downward assimilation or ethnic disadvantage. Instead, the overall lack of a positive relationship between PYSM and math grades leads us to conclude that parents’ incorporation has no effect in a course which requires little Swedish-specific skills.

Although the results, in principle, show a high degree of homogeneity in the relationship between PYSM and Swedish and math performance by parents’ country of origin, this is an important fact to establish on the topic of immigrant incorporation. This paper demonstrates that, irrespective of country of origin, parents develop linguistic proficiency that then helps to increase their children’s Swedish performance. However, it is unable to influence performance in a less Swedish-specific subject, math. That this relationship is homogenous is notable, considering the range of sending contexts and the changing context within Sweden that immigrants have encountered over the last half century. Our analysis agrees with previous research (Åslund et al. 2009; Nielsen and Rangvid 2012) that has found that PYSM is a useful measure to capture the intergenerational transmission of incorporation experiences, specifically, linguistic acculturation. Although our research shows that there is a positive effect of PYSM on the Swedish grade and test scores of second-generation immigrants from non-Western countries of origin, the low levels of PYSM found among these groups may contribute to their underperformance in Swedish relative the native population. These low levels of PYSM are the result of immigrants from non-Western origins arriving in adulthood at a much greater rate than those from European or North American countries, combined with earlier childbearing norms. As a result, many non-Western second-generation immigrants are born into households with parents whose incorporation experiences are often limited, even if what is accrued is transmitted to their children.

Following previous research (Nielsen and Rangvid 2012), we tested for differences by parents’ gender, which yielded no differences. Additionally, we included interactions between PYSM and sex of the individual, but it does not appear that parents’ time in Sweden affects brothers and sisters differently. We further stratified by parents’ highest educational attainment, which provided unclear results, even when considered by parents’ country of origin or Western/non-Western backgrounds. We additionally ran models that included a PYSM² term, since the impact of PYSM could diminish with time, but the variable itself was not significant and it had little impact on the PYSM term. Additionally, given the research on first-generation immigrants that stressed that those arriving early had the best outcomes, we stratified by whether the parent was a child immigrant (arrived before age 12) or not, but the relationships between grades and PYSM remained regardless of when they arrived. We also stratified analyses by family size (Additional file 1: Appendix B) across the samples, and the results are consistent. Finally, Sweden offers two language classes, a standard Swedish class and a Swedish as a second language class (SSL). The relationship between PYSM and children’s Swedish grade remains robust regardless if looking at those families with no children in SSL classes, all children in SSL classes, or at least one child in a standard Swedish class and one in an SSL class.

Although this study improves upon previous research by using a well-suited causal approach to look at the relationship between parents’ time in Sweden on their children’s educational performance by a more diverse country of origin group over a broader time frame than has previously been done, certain weaknesses exist. As was previously mentioned, the external validity to families with one child is unknowable using this approach. Although this includes a minority of the sample, it is nonetheless a weakness. Additionally, we exclude those with one foreign-born and one native-born parent, which make up a majority of second-generation immigrants from Western countries of origin. We consider here performance in two subjects, and though these were selected for specific purposes, it is uncertain whether the relationship between parents’ time in Sweden and children’s performance found here is representative for all non-language courses, or whether math might represent an outlier. Areas left to be explored include looking at a broader range of courses that might provide alternative testing grounds for the relationship between PYSM and children’s performance. Although results from this paper have stressed the role of language acculturation, and children of mixed marriages inherently have a fluent speaker already in the household, there may be some value to consider children who are born into mixed families. Finally, this paper looks at the role of PYSM in academic performance just before the age that tracking takes place in Sweden. Because Swedish students have more choice in the secondary school program they undertake, conditioned on performance, than is found elsewhere (Jonsson and Rudolphi 2011; Jackson et al. 2012), it is worth asking if parents’ time in Sweden has an effect on this decision, above and beyond its effect on school performance.