Traditional versus experimental pathways to university: Educational aspirations among young Swedes with and without an immigrant background

ABSTRACT In this paper, we investigate the different ways in which young people with and without an immigrant background, who have in common that they aspire for a higher education, picture their nearest future. We distinguish between two major pathways to a university education: a traditional pathway (direct transition from upper secondary school to university) and an experimental pathway (where university plans are mixed with other plans, such as work and travel). Data come from a school survey of 685 final-year students in 10 upper secondary schools in Malmö, the third biggest metropolitan area in Sweden. We find that it is much more common among youth with an immigrant background to plan for a traditional pathway, compared to youth of Swedish background who instead are more inclined to plan for experimentation. This difference remains when controlling for factors related to school performance and parental socioeconomic background.


Introduction
Studies on educational aspirations were first conducted in the United States and date back to the 1960s. These studies have repeatedly shown that young people with an immigrant background have higher educational aspirations than their native-born counterparts, despite generally lower level of school performance (Downey, Ainsworth, and Qian 2009;Kao and Tienda 1998;Mickelson 1990). This is also illustrated in many other countries, such as Germany (Salikutluk 2016), Norway (Hegna 2014), and the United Kingdom (Strand and Winston 2008). These studies often share the assumption that young people aspire to continue to university along what Milesi (2010) has called a 'traditional pathway', which refers to a direct move from upper secondary education to university in a step-by-step progress in education. Studies on educational transitions have however showed that the path to a university degree is frequently much less linear than assumed in these models. Young people often delay entry to university, shift between school programmes, or re-enter education after a certain age (Milesi 2010;Rindfuss, Swicegood, and Rosenfeld 1987;Shanahan 2000). So far, studies on educational aspirations have not used these insights from the literature on educational transitions to examine aspirations beyond the traditional pathway. Thus, we know very little about how young people who share the aspiration of a university education potentially differ in how they picture their near future, to what extent they see themselves taking a traditional (direct) pathway to university, and to what extent they intend to do other things before or in parallel with their university studies.
The aim of the current study is therefore to examine potential differences amongst young people who have in common that they have a university education in mind. We distinguish between two major pathways, namely a traditional and an experimental pathway. The former refers, as stated above, to a direct transition from upper secondary school to university, whilst the latter refers to a path where higher education is combined with, or preceded by, other activities such as work and travel. In this paper, we examine what young people who are still in secondary school see themselves doing within the next two years. We are thus looking at aspired educational pathways, rather than the actual route or transition from one level of education to another.
When looking at these differences in aspirations, we will focus on differences between young people with and without an immigrant background. 1 We argue that there are several reasons to expect differences to exist.

A culture of experimentation
Within the field of developmental psychology, 'experimentation' is often conceived of as a general characteristic of the transition from adolescence to adulthood. In his theory on emergent adulthood, Arnett (2000) argues that experimentation generally takes place in the early and mid-twenties. In this period of life, young people explore their self, experience frequent changes in their life situation and are surrounded by rising uncertainties. While Arnett (2000) as a developmental psychologist, has emphasised expected similarities between young people (at least across Europe and the United States), sociologists have instead focused on the extent to which contextual factors such as institutions and culture lead to strikingly different experiences in the transition from adolescence to adulthood (Billari and Wilson 2001;Buchmann and Kriesi 2011).
In their comparative study on the role of institutions (i.e. family, educational systems, labour market and state) for youth transition, Cook and Furstenberg (2002) argue that youth transition in Sweden and the Nordic countries, much more than in other European countries and the United States, is characterised by a culture of experimentation. In Sweden, institutions actively support young people's autonomy. The state provides generous student loans and housing benefits for low-income groups that enable young people who have not yet entered the labour market to become financially independent, reducing their reliance on family. Many young people in Sweden and in other Nordic countries therefore leave the family nest in the early twenties or even before, which is 'earliest-early' in Europe (Billari 2004). 2 Therefore, these opportunities allow young people to make more autonomous decisions about, among other things, their educational pathways.
In addition to higher education, work and overseas travel are two major areas through which young people explore themselves in Sweden. It is common to take time off from education to travel abroad before, as well as during, university education (Frändberg and Vilhelmson 2011). Similarly, many young people go in and out of different jobs during their twenties, a phenomenon that has been referred to as 'job shopping' (Johnson 1978), or take a break in education, before or during university, to get some work experience. A consequence of being 'experimental' in these respects is thus that entry to, and exit from, university is delayed. 3 Being an immigrant in Sweden Generally speaking, immigrants are self-selective groups in the sense that people often move to create a better future for themselves and their children. Kao and Tienda (1998) formulated the 'immigrant optimism hypothesis', referring to the fact that immigrant parents and their children generally have very positive attitudes towards higher education and are optimistic about the possibilities of advancement. Many immigrant parents view education as a vehicle for upward mobility (Heath, Rothon, and Kilpi 2008) and 'educational success' is often something they expect of their children, and that the children feel obliged to achieve (Hardway and Fuligni 2006). A large body of research supports the immigrant optimism hypothesis, showing that firstand second-generation immigrants have higher educational aspirations than native youth (Feliciano and Lanuza 2016;Hirschman 2001). We can also see that these aspirations, at least in Sweden, manifest in higher university enrolment among young people with an immigrant background (Skolverket 2012). Considering that children of immigrants generally can be assumed to have less exposure to the host country's culture of experimentation (see above), we may expect high educational ambitions within this group to translate into aspirations to enter and exit university as soon as possible, i.e. to adopt a traditional pathway.
In addition, we know that the post-school period is especially insecure for children of immigrants. Previous studies in Sweden highlight that such young people, particularly those with a non-Western background, are more likely to face adverse conditions in the labour market (Rooth 2002;Schierup, Paulson, and Ålund 1994). For example, key actors in gatekeeper positions have been found to discriminate against applicants with foreign-sounding names (Bursell 2012;Rydgren 2004;Schierup, Paulson, and Ålund 1994). More generally, people with an immigrant background run a higher risk of stigmatisation, i.e. they are more often faced with negative stereotypes, such as being inferior, less developed and not modern (Bredström 2003;Herz and Johansson 2012;Torres 2006). It is not unlikely that such adverse conditions, rejection, exclusion and disapproval prevent young people from exploring their selves and, more concretely, prevent them from delaying entry to, and exit from, university. The effect may be directe.g. reducing the chances of getting a job to finance a longer trip or allowing for a break in studiesor indirecte.g. reducing the confidence and trust in the future needed for a more playful and experimental attitude to life, work and education.
Another reason to anticipate a difference in the degree to which youth with and without an immigrant background plan for an experimental path to university, is the degree to which they are exposed to, and thereby influenced by, a culture of experimentation. Like studies on immigrant optimism has pointed out, many young people have 'dual references': cultural influences from both their, or their parents', country of origin, and from the country where they now live (Feliciano and Lanuza 2016;Kao and Tienda 1998). Since the immigrant population in Sweden is highly diverse when it comes to country of origin, these cultural influences can be expected to vary greatly. Nonetheless, since Sweden, to a greater extent than most countries, embraces 'experimentation' (see the previous section), we can assume that young people exposed to some culture other than the Swedish are in general influenced in a less experimental direction than those with an exclusively Swedish background.
Given that the degree of exposure to the culture of experimentation determines the extent to which young people plan for a traditional or an experimental path to university, we would also expect to find differences between firstand second-generation immigrants. Those who have lived their whole life in Sweden should, if the argument holds, be more inclined to plan for an experimental path to university than those who have spent part of their childhood in their parents' country of origin. (This type of cultural adaptation is often raised as an advantage of second-generation immigrants over first-generation immigrants in studies on immigrant optimism (Feliciano and Lanuza 2016;Kao and Tienda 1995).) On the other hand, one should also bear in mind that the 'adverse conditions' that young people with an immigrant background face seem to make little distinction between firstand second-generation immigrants. For example, actors in gatekeeper positions have been found to reject job applications from firstand second-generation immigrants alike (Bursell 2012;Rydgren 2004;Schierup, Paulson, and Ålund 1994). To the extent that second-generation immigrants anticipate adverse conditions in the near future, it may affect their aspirations. For instance, if they foresee that they will have difficulty finding a job to save money for travel, they may consider a direct transition to university to be their only (or at least best) option. If so, we should expect only minor (if any) differences between firstand second-generation immigrants when it comes to aspired educational pathways.

This study
We have discussed three theoretical perspectives relevant to an analysis of the potential differences in aspired educational pathways between young people with and without an immigrant background: the 'culture of experimentation', the immigrant optimism hypothesis and the adverse labour market conditions generally facing young people with an immigrant background. We argued that high levels of educational ambition among youth with an immigrant background, in combination with generally lower exposure to culture of experimentation, implies that they should opt for a traditional pathway to university education more often than their Swedish peers. Awareness of the adverse conditions awaiting them in the labour market may similarly steer them towards a traditional path. When it comes to potential differences between firstand second-generation immigrants, the conclusion differs depending on perspective: if the degree of exposure to a culture of experimentation is important, then second-generation immigrants should plan for an experimental pathway more often than first-generation immigrants. If the anticipation of adverse conditions matters more, however, there is less reason to expect difference since second-generation immigrants face more or less the same difficulties in the labour market as first-generation immigrants.
When analysing these differences between youth with and without an immigrant background, we identify three factors that may be expected to be intertwined with background (native/ immigrant) and aspired educational pathways: socioeconomic background, gender and school grades.
Firstly, there is a class aspect to experimentation. Although not mentioned by Cook and Furstenberg (2002), other studies have showed that socioeconomically advantaged parents more often adopt a set of behaviours that encourage young people to follow their own way (self-directedness) and to be more resourceful, i.e. to enjoy diverse interests and leisure activities (Reay 2000;Vincent, Ball, and Kemp 2004). In addition, economic resources restrict the extent to which experimentation is possible. To be able to travel, for instance, a young person needs economic support from their family or a job which enables them to save up money for themselves. For this reason, we may expect children from socioeconomically advantaged families to be more inclined to aspire for an experimental pathway to university compared to children from less advantaged families.
School grades may restrict the possibility of being accepted to a university education, and low grades could therefore potentially manifest in aspirations for a pathway here defined as 'experimental'. That is, if lower grades lead to uncertainty about being accepted to the preferred university education pathway, the individual may instead plan to combine work or travel with additional courses to improve his/ her grades (which is possible within the Swedish adult education system), and/or taking the Scholastic Aptitude test (an alternative way to compete for a place on courses at Swedish universities). On the other hand, research on 'immigrant optimism' suggests that educational aspirations are not necessarily affected by grades (Mickelson 1990;Raleigh and Kao 2010;Salikutluk 2016) and if this is also the case with aspired educational pathways, the fact that youth with an immigrant background generally have somewhat lower grades would not affect the aspired educational pathway.
When it comes to gender, studies from Australia and Great Britain show that it is more common for women than for men to take a 'gap year' between secondary school and university (Birch and Miller 2007;Jones 2004). Swedish official statistics similarly show that although women generally have higher grades and are slightly overrepresented in higher education, they tend to exit university at a later age than men (UKA 2017). Women also choose to study abroad more often than men (ibid.), possibly indicating a more experimental attitude towards this period of life. We may thus expect women to opt for an experimental pathway to a university education more often than men. On the other hand, traditional gender roles may be more prevalent among some immigrant groups (e.g. Turks and Kurds) and thus reduce the degree of autonomy available to women (see e.g. Akpinar 2003;Darvishpour 1999). For this reason, we may expect gender differences in aspired educational pathways to vary between youth with and without an immigrant background.
Based on the arguments above we set out to examine the following research questions: (1) Do young people with an immigrant background, more often than those of Swedish origin, plan for a traditional pathway to a higher education? Correspondingly, do young people of Swedish origin, more often than those with an immigrant background, plan for an experimental pathway to higher education? (2) Do second-generation immigrants, to a higher extent than first-generation immigrants, plan for an experimental pathway to a higher education? (3) If we find such differences, to what extent are they related to socioeconomic background, gender and school grades?

Participants and procedure
This paper is based on data collected as part of a larger research project involving upper secondary schools in Malmö, Sweden's third largest city. 4 The population of Malmö reached 314,000 in 2014. The city has experienced major economic and population changes since the 1970s. From the 1990s onward, the city has received a large flow of immigrants, mostly with Middle Eastern and African backgrounds. According to SCB (2016), one out of three people in Malmö has a foreign background.
In spring 2011, a survey was administered to all first-year upper secondary students in a strategic sample of eleven schools. Since the total number of schools was relatively small and since the schools differ greatly in terms of their programmes as well as the size and social composition of the student body, a strategic sampling method was considered to be the best way to attain a representative sample. The sample was thus constructed to contain as wide a range of schools, programmes and students as possible. The total number of students registered at the selected schools (1169) constituted 32% of the total number of students registered at any of the secondary schools in Malmö that year (3647). The survey was filled out in the classroom during school hours, either on paper or online.
A second survey was carried out among the same cohort two years later, in spring 2013, when the students were in their final year of upper secondary school and aged about 18 years. Of the students still registered, 708 (70%) participated in the second survey. This paper is based on data collected in the second survey. In order to make the categories more distinct we excluded 23 respondents of Western European and North American origin. This concerns a small number of countries and most of them (especially the other Nordic countries) resemble Sweden with regard to the culture of experimentation. For example, the United Kingdom has a gap year phenomenon in which young people often take a year off from their studies to focus on their personal development (Jones 2004;King 2011). We also expect this category of immigrants to experience less discrimination (adverse conditions) and that the immigrant optimism hypothesis does not apply to them to the same extent as other immigrant groups. This reduces the sample size to 685 students. The majority of students with an immigrant background remaining in the sample are born in, or have parents who are born in, Middle Eastern countries (e.g. Iraq, Lebanon, Turkey), Balkan countries (e.g. Bosnia, Kosovo and Serbia), East/Southeast Asia (e.g. Vietnam) and South American countries (e.g. Chile).

Aspired educational pathways
In the survey, the students were asked whether they were planning, within the next two years, to (i) continue on to university, (ii) find a job, (iii) travel abroad, (iv) get engaged or married, (v) stay at home and take it easy. Each was a separate question with two possible answers: yes or no. It was possible to mark any number of alternatives and they were subsequently coded yes/ no. To continue to university (55%), to get a job (30%) and to travel (44%) were the most frequently mentioned alternatives and all students in the sample marked at least one of these. The remaining options were marked by 8% (getting married) and 13% (stay at home), respectively. The operationalisation of 'aspired educational pathways' is based on the first three alternatives (university, work and travel) since they were considered well aligned with the idea of traditional versus experimental pathways. After twelve years in school, getting a job or travelling (or a combination of both) are common ways for young people to explore themselves and the surrounding world, to get a sense of independence, meet new people and learn new things. Based on the responses under these three items, we constructed a dependent variable consisting of three categories: 'plan to continue to university only' (traditional pathway), 'plan to continue to university and work/travel' (experimental pathway), and 'do not plan to continue to university' (see Table 1).
The remaining two alternatives (getting married and staying at home) did not influence whether a respondent was categorised as experimental or traditional. We consider the alternative 'stay at home and take it easy' to indicate either disengagement (Collin 2008;Lesley et al. 2012) or, in combination with 'university studies', full focus on education. Likewise, we do not find that getting engaged or married indicate experimentation, at least not to a higher extent than other forms of romantic relationships that (at least in Sweden) are much more common at this age (SCB 2016). Respondents who have marked 'get married/ engaged' or 'stay at home and take it easy' can thus be found in any of the three categories, since the categorisation is not determined by whether these alternatives have been marked or not.
Of the students who plan an experimental pathway, 95% include travel in their plans for the next two years. Of these, 40% plan to go to university and travel only, and 55% plan all three activities: studies, work and travel. Only 5% (eleven students) plan to combine studies with work only. We do not know how many of these eleven students plan to study full-time at university while working part-time, and it can be questioned whether such respondents should be categorised as experimental at all. However, as part-time jobs are common among students in Sweden 5 we would expect this number to be much higher if the respondents had interpreted 'find a job' in this way. Our cautious interpretation of the low number of respondents in this category is thus that they are primarily thinking of full-time jobs. If some are in fact traditional in the sense that they plan to study full-time directly whilst working part-time, they constitute such a small part of the sample that they do not have a major influence on the results.

Immigrant background
We distinguish between three categories. Young people born abroad, no matter where the parents are born, are referred to as first-generation immigrants (n = 133). For 'second-generation immigrants', we use a strict definition which refers to native-born children with two foreign-born parents (Portes and Rumbaut 2001). It can be discussed whether it is appropriate to label this category (n = 171) 'immigrants' since they are born in Sweden but we decided to stick to this terminology since it is widely established and applied by e.g. Statistics Sweden. The third category, assigned as a reference category in the analysis, consists of children born in Sweden and with at least one parent born in Sweden (n = 381). We use the term 'immigrant background' when referring to both firstand second-generation immigrants.
Gender refers to whether the respondent identifies as male or female. Employed parent(s) is a variable indicating whether at least one of the parents has a fulltime job (1) or not (0). In Sweden, the unemployment risk is twice as high among non-Western immigrants compared to Swedes (Arai and Vilhelmsson 2004). This means that more immigrant parents depend on employment benefits. In that sense, the measure is a proxy for economic differences among parents.
Highly educated parent(s) is a variable showing whether at least one parent has a university education (1 = yes, 0 = no).
School grade is the average grade from three courses, namely the first-level mathematics, social science and Swedish courses. These grades were included in the study since all students take these courses, no matter which programme they are enrolled in. 6 Grades are given on a four-point scale from failed (1) to excellent (4).

Results
We begin by reporting descriptive statistics for the three groups: 'first-generation immigrants' (respondents born abroad), 'second-generation immigrants' (respondents born in Sweden with two foreign-born parents) and 'natives' (respondents born in Sweden with one or two Swedish-born parents). The former two groups collectively comprise respondents with an immigrant background. Table 2 shows how young people, with and without an immigrant background, imagine the period after upper secondary school. The results include aspirations in three domains: work, education and travel. Young people with an immigrant background, regardless of whether they are born in Sweden or abroad, have higher educational aspirations than those of Swedish origin in the sense that they more often plan to continue to university (68 and 60%, respectively, compared to 49% of Swedish youth). Young people of Swedish origin, on the other hand, are more likely to see themselves working after high school (38% compared to 18 and 20% of those with an immigrant background) and/or travelling (54% compared to 32% of those with an immigrant background).
Turning then to aspired educational pathways, we find, as expected, that young people with and without an immigrant background differ in their plans. Only 15% of youngsters of Swedish origin plan to continue to university directly; the corresponding rate for first-generation immigrants is 43%, and for second-generation immigrants 39%. In contrast, the proportion of natives planning to take an experimental pathway to university is 34%, compared to 23 and 25%, respectively, for second-and first-generation immigrants. Table 2 shows major differences in school grades and parental socioeconomic background. We find that the grades of young people with an immigrant background lag behind the grades of young people without an immigrant background. While there are significant differences in parental education (16 percentage points between natives and second-generation immigrants), there is a huge gap when it comes to parental employment: 63% of those with native-born parents have two parents who work full-time, compared to only 28% of first-generation immigrants. The gender distribution is, however, very similar in the three groups.
To summarise the results so far, we find major differences between young people with and without an immigrant background when it comes to educational pathways. We do not, however, see major differences between firstand second-generation immigrants. The background of the parents (Swedish or non-Swedish) thus seems to be more important than the amount of time the young person and their family has spent in Sweden. However, the fact that there are also major differences in the socioeconomic and educational backgrounds of the parents, as well as in school grades, makes controlling for such factors essential before drawing any conclusions about the importance of immigrant background. In the following section, we will therefore continue the analysis, using multinomial logistic regression. Table 3 demonstrates a multinomial logistic regression model. The model examines whether the differences between the three groups with regard to aspired educational pathways remain when taking into consideration differences in school performance, gender and parental socioeconomic background. We choose a multinomial model since the outcome variable consists of three categories (traditional, experimental and no university plans). This means we get three models where each pairwise combination of categories (pathways) is presented. The relative risk ratio (RRR) then gives us the effect of each independent variable. Its interpretation is equivalent to that of the odds ratio (Long and Freese 2006, 391).
We will focus on the first model in Table 3: 'Traditional vs. Experimental'. The main conclusion that can be drawn from this table is that the differences remain when all other variables are controlled for. The probability for a first-generation immigrant to plan a traditional pathway to university is 3.38 times higher than the probability for a native Swede, and for a second-generation immigrant compared to a native Swede, 3.63 times higher. The results thus confirm our expectations and previous results, showing that young people with an immigrant background are more likely than those with a Swedish background to aspire to a traditional pathway. However, we do not find any indication that second-generation immigrants are more experimental than the first generation on the contrary they turn out to be (marginally) more traditional than first-generation immigrants, when controlling for other variables.
Turning to the remaining variables, parental education and employment does not have a significant influence on the aspired educational pathway. Gender, however, now becomes significant: female students are more likely than their male counterparts to plan for experimentation. We examined the interaction effect between immigrant background and gender but found no significant difference in the size of the gender gap between the three groups. In all three groups, women are more 'experimental' than men.
School grades do not have a significant effect on the pathway to which a young person aspires. From the other two models, however, we can conclude that school grades are important predictors for whether students aspire to go to university at all, irrespective of pathway. That is, students with higher grades are significantly more likely to plan to go to university than those with lower grades. Likewise, children of highly educated parent(s) are significantly more likely to plan for a traditional pathway to university, compared to not planning to go to university at all. 7 These results are in line with previous research, and may be explained by the fact that highly educated parents are generally more involved in the schooling of their children and provide them with a home environment (with access to the literature etc.) that encourages them to continue to university (see Lareau 1987).

Discussion
The current study shows that young people with an immigrant background plan to continue university directly more often than those with entirely Swedish backgrounds, who instead more often plan to mix university studies with work and/or travel. The results are thus is in line with our expectations. Among those with an immigrant background, Table 3. Multinomial logistic regression estimating aspired educational pathways, by using relative risk ratios (RRR = exp(β)). we did not find substantial differences between young people who are born in Sweden (second generation) and those who have themselves immigrated (first generation). This finding is in line with previous studies on educational aspirations, in the sense that similar results have been found for these two groups. For example, regardless of where they themselves are born, children of immigrants are more likely to choose academic programmes over vocational ones than children of Swedes (Jonsson and Rudolphi 2011). The fact that the aspired educational pathways of firstand second-generation immigrants differ little indicates that the time spent in Sweden is of less importance than immigrant background itself. In that sense, the results thus speak against the 'exposure to the culture of experimentation' explanation and for the 'adverse conditions' explanation for the differences found between the three groups. It is often claimed that second-generation immigrants have an advantage over first-generation immigrants since they are fluent in the native language and familiar with the educational and labour market systems (Feliciano and Lanuza 2016;Kao and Tienda 1998). As mentioned before, second-generation immigrants are however faced with more or less the same adverse conditions as first-generation immigrants on the labour market. If this group is generally better informed and has a more accurate understanding of how the labour market functions, this may subsequently lead them to the conclusion that the best option for them is to aim for a traditional pathway to a university education. Unfortunately, we do not have any indicators on expectations of future discrimination in the labour market, and therefore we are not able to estimate the influence of such perceptions on aspired educational pathways. However, at least one more study supports the idea that a traditional pathway to a university degree can, for some people, be an attempt to counteract expected discrimination in the labour market (Goyette and Xie 1999).
One may also generally assume that the lower the threshold (e.g. entry requirements) and cost (e.g. university fees), the more likely that higher education is the best option available for people with few other options. In Sweden, tuition is free and relatively advantageous student loans are available to cover the costs of living. For this reason, the association between socioeconomic background and educational attainment is weaker in Sweden than in many other European countries (Breen 2005). It is thus not strange if young people who expect to face adverse conditions in the labour market, and/or lack the financial means to travel abroad, plan for a direct path to a university degree.
Could there be anything problematic about the fact that youth with an immigrant background more often plan for a traditional, rather than experimental, path to university? There is no simple answer to this question and it depends to some extent on the mechanisms involvedif the differences are primarily resulting from differences in preferences, or from an adaptation to unequal life chances. More generally, however, both pathways potentially have positive as well as negative outcomes in later life. Experimentation may result in better thought through educational choices and a more successful and rewarding career, but can also be considered a waste of time and future income if it simply means that the exit from university is delayed. A period of experimentation may be joyful and rewarding for some, but delaying important decisions, such as applying to university, may lead to ambivalence and indecision for others. Since instability and exploration goes hand in hand in the early twenties, feelings of being lost and confused are common (Arnett 2000) and maximum freedom to explore oneself may not always be the best cure.
The fact that young people with an immigrant background are less likely to plan for an experimental path, trial and error in a job market, or self-exploration, can be interpreted as suggesting that they are more determined than their Swedish peersdetermined to achieve higher social status, to contribute to society or simply to avoid unemployment. On the other hand, it may not be the person with the highest degree at the lowest possible age that actually gets the sought-after job, but the one who has volunteered in a refugee camp or climbed Mount Everestnot because those experiences are relevant to the work in question but because they send signals about personality that align with general ideals about independence, creativity and determination. For example, mutual interests between the employer and jobseeker may create emotional ties, which in turn stimulate a positive evaluation of a job application (Rivera 2015). Studies have also suggested that informal methods, such as knowing influential people, are as important as formal methods, such as applying for publicly advertised positions, when trying to find a job (Behtoui 2008;Ekström 2001;Korpi 2001). In line with these studies is one showing that in Sweden (in contrast to the Netherlands) an educational degree is not interpreted as a clear sign of the jobseeker's productivity (Breen 2005).
In sum, although formal education is undoubtedly very important for the employment chances of an individual, other factors may be as important. As the job market becomes increasingly competitive for highly educated people (Bygren, Duvander, and Hultin 2013) young people with 'only' an academic degree may be at a disadvantage with regard to the informal side of recruitment. In other words, there may be a risk associated with not adhering to the culture of experimentation. On the other hand, it is possible that young people aspiring to a traditional pathway in fact manage to engage in experimentation during their studies, or that a focus on achieving a university degree is simply the most viable strategy.
For now, we may conclude that there appear to be major differences in the ways in which university-aspiring youngsters, with and without an immigrant background, picture their nearest future. The consequences of this are yet to be explored but considering the adverse conditions awaiting many young people with an immigrant background, it is important to gain deeper understanding of the reasons for, and the consequences of, differences in aspired educational pathways.
There are some limitations associated with this study that are important to note. One major limitation is the sample size. A larger sample would provide opportunities for a more thorough investigation of, for example, ethnic differences within a heterogeneous group referred to here as simply 'youth with an immigrant background'. The reader should also be reminded that this study was conducted in a metropolitan area with a high proportion of immigrants, and greater segregation than most other places. The results should therefore be generalised with caution to other contexts. Another limitation of the study is the rather crude measure of 'experimental pathway'. A more fine-tuned instrument, in which more details of the plans were measured, could provide nuanced information about the aspired educational pathways of young people. It would, for example, be valuable to be able to distinguish between those who intend to work parttime during their university studies in order to finance the latter, and those who intend to work before they enter university. Finally, we believe that qualitative investigations of the contexts in which these aspired educational pathways arise could prove fruitful in order to deepen and further develop our understanding of the processes at play. Notes 1. In this paper we define an individual as having an immigrant background if he/she is born outside Sweden or if both parents are born outside Sweden. 2. The average age of leaving the family nest is 20.3 years in Sweden, compared to 22.0 in the Netherlands, 22.7 in Ireland and 22.5 in Germany (Leopold 2012). 3. According to OECD (2015) standardised statistics, the average age of entry to university is 23.9 years in Sweden. The corresponding age in the Netherlands is 19.7, in the United Kingdom 21.4, in Italy 20.4, and in Germany 21.4. The average ages of exit from university follow similar patterns: young Swedes are older when they graduate compared to students in other countries. In Sweden, the average age at graduation is 28.3 years, in the United Kingdom it is 23.5, in the Netherlands 23.6, and in Germany 25.5. 4. The research project, 'An educational dilemma: school achievement and multicultural incorporation', was funded by the Swedish Research Council and headed by Mats Trondman, a professor of cultural sociology at the Linneaus University, and has been approved by the Ethical Board. 5. According to our calculations based on the European Social Survey (2016), one out of three students (33.6 percent) aged 18-24 have a part-time job (i.e. work less than 30 h in a week). 6. The grades are self-reported. In addition, the grades are not to be interpreted as perfect indicators of actual school performance. There are national tests that all pupils sit, but schools do not have to use the results when grading students. This has resulted in grades that are generally higher than the corresponding test results (Skolverket 2017). The grades are, however, a good indicator of a student's chances of being accepted to a university programme. 7. We checked robustness of the results in two ways. First, we ran the same model on a smaller but more homogenous subsample. The majority (74%) of respondents with an immigrant background had parents born in Middle Eastern countries (including Turkey). When repeating the analysis on this subsample we obtained similar results. The relative risk ratios for traditional versus experimental pathways are then 3.35 for first-generation and 3.36 for second-generation immigrants (reference category being 'native Swedes'). Second, we repeated the analysis with a dependent variable consisting of five (instead of three) categories: 1) university + job (n=11), 2) university + job + travel, (n=81), 3) university + travel, (n=112) 4) only university (n=187), and 5) do not plan to continue university (n=305) (see Table 1). The results remain largely the same when parental socioeconomic background, school grades and gender are kept constant. The relative risk ratio ranges between 2.52 and 6.05 in the different models. All results are statistically significant except the model where 'university + job' is used as a reference category, which is due to the low number of respondents in this category (n=11).

Disclosure statement
No potential conflict of interest was reported by the authors.