Horizontal Advantage: Choice of Postsecondary Field of Study Among Children of Immigrants

Abstract

Educational expansion has raised the influence of sorting across postsecondary educational fields on children's future life chances. Yet, little is known about horizontal ethnic stratification in the choice of field of study among children of immigrant parents, whose parents often have moderate absolute levels of education relative to native-born parents but tend to be positively selected on education relative to nonmigrants in the origin country. Using rich administrative data from Norway, we study the educational careers of immigrant descendants relative to the careers of children of native-born parents. Our results show that children of immigrants from non-European countries have a higher likelihood of entering higher education and enrolling in high-paying fields of study compared with children of natives, despite having poorer school grades and disadvantaged family backgrounds. However, immigrant parents' positive selectivity provides limited insight into why children of immigrants exhibit high ambitions later in their postsecondary educational careers. These findings document a persistent pattern of horizontal ethnic advantage in postsecondary education in which ambitious children of immigrants are more likely to enter into more prestigious and economically rewarding fields of study than their fellow students with native-born parents.

Introduction

Successful incorporation of children from disadvantaged immigrant families into the educational systems of Western immigrant-receiving countries is a key challenge in the twenty-first century (Alba et al. 2011; Drouhot and Nee 2019; Heath et al. 2008). Because of their immigrant parents' disadvantages, such as limited language proficiency, neighborhood disparities, and ethnic discrimination, children of immigrants often face difficulties in terms of low school achievement and heightened risks of early school-leaving (Heath et al. 2008; Kao and Thompson 2003). However, these children often exhibit higher transition rates into academic upper secondary educational tracks and postsecondary education relative to children of native-born parents with comparable socioeconomic origins (Glick and White 2004; Heath and Brinbaum 2014; Jackson et al. 2012; Keller and Tillman 2008). Prior research pointed to high aspirations as a key explanation of immigrant-background students' bold educational choices (Friberg 2019; Jackson et al. 2012; Kao and Tienda 1998). Unpacking this phenomenon of high aspirations—referred to as immigrant optimism (Kao and Tienda 1995), second-generation advantage (Kasinitz et al. 2008), or an immigrant drive (Portes and Rumbaut 2001)—is central to improving our understanding of future ethnic stratification and assimilation across immigrant generations.

In recent decades, large-scale immigration has coincided with an expansion of higher education and inflation in educational degrees across Western societies (Alon 2009; Gerber and Cheung 2008; Lucas 2001). With growing numbers of university students, horizontal differences across postsecondary education fields of study have gained importance for graduates' future earnings, occupational attainment, and career outcomes (Borgen and Mastekaasa 2018; Kim et al. 2015; Reimer et al. 2008). Children from advantaged family backgrounds not only complete more education (Jackson 2013) but also tend to enter more prestigious fields of study (Ayalon and Yogev 2005; Davies and Guppy 1997; Hällsten 2010; Helland and Wiborg 2019; Kraaykamp et al. 2013). Prior research from Europe and North America reported that immigrant-background students often have a high propensity to enter higher education (Glick and White 2004; Heath and Brinbaum 2014; Jackson et al. 2012; Keller and Tillman 2008). However, much less is known about horizontal ethnic stratification in postsecondary education. In this study, we ask whether children of immigrants are more or less likely to enroll in prestigious educational fields with higher economic returns than children of native-born parents.

We investigate differences between children of immigrants and children of native-born parents in postsecondary education entry, sorting into different educational fields of study, and the expected labor market returns to their chosen educational fields. Specifically, we explore whether horizontal ethnic inequalities in the choice of postsecondary educational field can be explained by earlier academic achievement and differences in family background related to immigrant parents' length of education and premigration social status in the country of origin. We draw on high-quality microdata from Norwegian population registries that provide detailed information on children's educational careers and parental resources.

There are several reasons to expect ethnic differences in the choice of postsecondary educational field. First, immigrant parents may lack local institutional knowledge about how to succeed in the labor market and may encourage their children to choose well-known fields of study that carry high prestige in the country of origin (Chiswick 1988; Jonsson and Rudolphi 2011). Second, immigrants and their native-born children may expect to face ethnic discrimination in the labor market, which could increase motivation to select educational fields that provide economic security and high earnings (Heath et al. 2008; Salikutluk 2016). Further, children who grow up in low-income immigrant families may also be more prone to focus on the material and economic benefits of education (Inglehart 1997), thus gravitating toward educational fields that likely provide secure career outcomes and higher earnings.

However, recent studies have argued that immigrant descendants' high educational aspirations are related to the positive selection of their immigrant parents on traits such as perseverance, health, and education relative to nonmigrants in their country of origin (Feliciano 2020). Positive educational selectivity among immigrants has been documented across a broad range of destination countries (Engzell and Ichou 2020; Feliciano 2005b; van de Werfhorst and Heath 2019) and has been found to contribute to high aspirations and educational advancement among the native-born second generation (Engzell 2019; Feliciano 2020; Feliciano and Lanuza 2017; Ichou 2014). Expanding on this literature, we hypothesize that children of positively selected immigrants are more likely to make ambitious choices in postsecondary education and sort into fields of study with higher economic returns than are children of natives.

Norway has an ethnically diverse and large immigrant population, with a foreign-born population share that is comparable to that found in France, Germany, the Netherlands, and the United Kingdom. This immigrant population is characterized by different selection patterns owing to considerable inflows of labor migrants and refugees over the past decades (Brochmann and Kjeldstadli 2008; OECD 2022; Statistics Norway 2023). Norway is of comparative theoretical interest for horizontal educational stratification among children of immigrants: it allows for investigating how the phenomenon of immigrant optimism in education unfolds in a context relatively unrestricted by institutional obstacles. Free tuition, affordable loans, and generous subsidies to students in Norway suppress economic barriers to university education, and lack of early-ability tracking reduces the influence of academic achievement at earlier stages of schooling (Bol and van de Werfhorst 2013; Jackson et al. 2012). Overall, this institutional landscape should expand the role of other determinants of educational choices, such as student ambitions and parental premigration status, throughout children of immigrants' educational careers in the Norwegian choice-driven school system.

Background and Previous Literature

Research on educational inequalities between children of immigrant and of native-born parents reveals complex and divergent patterns, depending on whether the focus is on achievement outcomes (e.g., grades or test scores) or educational decision-making and completed levels of education. Children of immigrants from some regions tend to perform worse on standardized tests and receive poorer school grades than their average nonmigrant counterparts, although their achievements are often comparable to those of native peers from disadvantaged families (Heath and Brinbaum 2014; Schnepf 2007). For example, children of Turkish and North African immigrants often lag behind in European countries, whereas children of Hispanic origins tend to be disadvantaged in the United States (Drouhot and Nee 2019; Heath et al. 2008). By contrast, children from other regions, often of Asian background, commonly outperform natives and have been referred to as model minorities (Lee and Zhou 2015; Sakamoto et al. 2009).

Interestingly, progression to higher attainment levels reveals a different picture. Children of immigrants often excel relative to native-born parents' children with similar educational achievements and socioeconomic backgrounds (Heath and Brinbaum 2014; Jackson et al. 2012). Notably, comparatively high shares of immigrant students continue into postsecondary education despite poor achievement levels (Glick and White 2004; Heath and Brinbaum 2014; Jackson et al. 2012; Keller and Tillman 2008). Although this research has increased our understanding of educational career progression from primary school to university-level education (Glick and White 2004; Heath and Brinbaum 2014; Jackson et al. 2012; Keller and Tillman 2008), few studies have explored ethnic differences in college choice and educational fields (Ma 2009; Song and Glick 2004; Xie and Goyette 2003).

However, rapid growth in the number of students in postsecondary education over the past decades has led to an increase in the number of individuals earning degrees, and thus horizontal differentiation in higher education arguably matters more for graduates' future earnings, occupational attainment, and other employment-related outcomes today (Borgen and Mastekaasa 2018; Kim et al. 2015; Reimer et al. 2008). Children of less educated native-born parents tend to complete less education relative to those from better-off family backgrounds (Breen and Jonsson 2005) and often select educational programs with shorter duration, lower earnings, and less prestige (Davies and Guppy 1997; Hällsten 2010; Helland and Wiborg 2019; van de Werfhorst et al. 2001). Whether immigrants' children resemble disadvantaged children of natives or instead display higher academic ambitions by sorting into postsecondary educational fields with higher prestige and better economic payoffs are key questions that are largely unanswered.

Several mechanisms could explain the high educational ambitions often found among immigrants' children. One proposed explanation emphasizes that immigrant parents anticipate ethnic discrimination and blocked opportunities for their children in blue-collar occupations and low-wage segments of the labor market where they themselves have been employed (Heath et al. 2008; Salikutluk 2016). If immigrant parents perceive the labor market for university graduates to be more meritocratic, they may encourage their children to pursue high degrees to avoid unemployment and low-paying jobs (Jonsson and Rudolphi 2011). Moreover, the anticipation of discrimination and blocked opportunities may also cause immigrant parents to steer their children into study programs with more secure access to employment and better earnings prospects. For example, evidence suggests that licensed occupations, in which access is limited to individuals holding specific educational credentials (e.g., health professions and law degrees), reduce immigrants' earnings disadvantage relative to natives (Drange and Helland 2019). The labor market for university graduates is often more meritocratic because the impact of social origins on college-educated workers' career outcomes is reduced (Breen and Jonsson 2007; Hout 2012; Mastekaasa 2011).

A second, perhaps paradoxical reason to expect high educational ambition among children of immigrants is related to immigrant parents' lack of knowledge about culture and institutions, which lowers their ability to help their children navigate the educational system (Bleakley and Chin 2008; Casey and Dustmann 2008). Such lack of cultural and institutional know-how may also predispose immigrants to encourage their children to select educational fields that are well known to carry prestige in most countries (Chiswick 1988). Occupational prestige is similar across countries (cf.Treiman 1977). Further, as Xie and Goyette (2003:490–491) argued, immigrant children may strategically adapt to existing social and economic structures by placing a higher premium on the “instrumental value, rather than the intrinsic value, of formal education” and thus favor “fields of high demand in the economy, as their preferred channel of mobility.” Relatedly, many children of immigrants experience economic hardship and insecurity while growing up and may be more prone to emphasize so-called materialistic/survival values as opposed to self-expressive values, which include a focus on nonmonetary rewards, such as creativity and self-fulfillment (Inglehart 1997). According to this argument, individuals from disadvantaged backgrounds may be more focused on avoiding economic insecurity, which might prompt them to enter educational fields likely to provide personal financial gains and a secure income for their family and other close kin.

However, a lack of knowledge about the host country's educational system may lead children of immigrants to make (overly) ambitious choices by underestimating the requirements and demands for attaining higher degrees in prestigious fields of study (Birkelund 2020; Dollmann and Weißmann 2020; Tjaden and Hunkler 2017). Thus, ambitious choices are not necessarily transformed into the completion of the chosen education. Further, high ambition among children of low-income immigrants could be blocked by expensive tuition and living expenses at the elite universities offering more selective and advanced fields of study. However, the lack of parental financial resources should be less relevant in egalitarian educational systems that limit economic barriers, which largely allows for the manifestation of high educational aspirations and ambitious educational choices (Jackson et al. 2012).

A third possible cause of ambitious choices of postsecondary educational fields is related to children of immigrants' attempts to regain their parents' premigration social status. It has long been acknowledged that immigrants are a selected group relative to nonmigrants (Lee 1966; Ravenstein 1885), as captured by Lee's (1966:56) famous statement that “migrants are not a random sample of the population at origin.” Recent studies have argued that positive self-selection among immigrant parents—who often have high educational levels relative to their nonmigrant compatriots in their home country—is a key explanation for their children's high educational aspirations (Engzell 2019; Feliciano 2020; Feliciano and Lanuza 2017; Ichou 2014). For example, immigrants who experience downward occupational mobility upon arrival (Akresh 2008; Chiswick et al. 2005) may retain a subjective reference group and family culture that reflects their social status before migration (Engzell and Ichou 2020). In a situation of transnational status inconsistency—referred to as the status paradox of migration (Nieswand 2012)—immigrants may transmit to their children a set of cultural dispositions, beliefs, and other nonmaterial resources that primarily reflect their earlier class position. For example, Feliciano and Lanuza (2017:215) argued that “[i]mmigrants' children may internalize their parents' pre-migration class orientations and parents' narratives of hardships, propelling them to vindicate their parents' sacrifices through their own success.”

Suppose migrants leave their origin country with the aim of achieving socioeconomic progress after settling in a new country. If they are barred from achieving this goal, they may transmit their ambitions to their children (Louie 2012), and the relative position in the status hierarchy before migration may serve as the benchmark below which they do not want their children to fall (cf.Boudon 1974). However, immigrants' children may need to complete substantially higher absolute levels of education than their parents to retain—let alone improve—their parents' premigration socioeconomic position (Engzell 2019; Feliciano and Lanuza 2017; Ichou 2014). Importantly, in the context of educational expansion, children of positively selected immigrants may need to seek opportunities along horizontal dimensions within higher education and aim to enroll in more selective fields of study that yield high status and economic returns to regain or improve their parents' relative status in the origin country. A positive association between immigrants' educational selectivity and their second-generation occupational status aspirations (Engzell 2019) will likely raise the likelihood that their children will also pursue prestigious educational trajectories along horizontal dimensions.

Children of immigrants, however, face potential barriers to reinstating their parents' premigration social position by enrolling in prestigious educational fields because doing so requires both high ambition and high achievement. In this regard, Boudon's (1974) classical distinction between the primary effects (i.e., educational achievement inequality) and secondary effects (i.e., educational choice inequality) of family background on educational careers is also highly relevant for studies of ethnic stratification in schooling (Engzell 2019; Heath and Brinbaum 2014; Jackson et al. 2012). School grades or standardized achievement tests are central criteria for accessing high-status selective fields of study. Immigrants' children often lack access to socioeconomic resources and cultural and institutional knowledge that facilitate academic achievement and ease access to more selective study programs, implying disadvantages in entry to higher education. Nonetheless, positively selected immigrant parents may transmit skills and other nonpecuniary resources to their children that aid their academic achievements. For example, economists have argued that immigrants often are self-selected on skills and abilities, which can explain upward earnings assimilation in the years after facing initial barriers in the labor market (Borjas 1987; Chiswick 1978). Empirical research has found immigrants to be positively selected on skills and health (Akresh and Frank 2008; Chiquiar and Hanson 2005)—factors that likely can positively affect their children's academic achievements.

Previous research on the link between immigrants' educational selectivity and their children's schooling has primarily focused on completed levels of education. Feliciano's (2005a, 2005b) groundbreaking studies used aggregate data to examine the likelihood that a U.S. immigrant was more or less educated than those who stayed behind in the country of origin and showed that this selectivity could explain group-level differences in schooling between immigrant descendants of different national origins. Feliciano and Lanuza (2017) demonstrated that individual-level measures of parental educational selectivity help explain why most groups of U.S. immigrant descendants complete more education than children of nonmigrant White Americans. These educational advantages are largely mediated by parents' and children's aspirations and expectations. Similarly, Ichou (2014) demonstrated that the relative educational attainment of immigrants to France predicts their children's educational attainment after adjusting for postmigration socioeconomic resources. Using Swedish data, Engzell (2019) demonstrated a positive influence of parental educational selectivity on continuation into academic upper secondary education tracks, career aspirations, and attitudes toward schooling among immigrants' native-born children in the second generation. Despite these important contributions, little is known about how sorting across postsecondary educational fields among children of immigrants is related to their parents' educational selectivity. Addressing this knowledge gap constitutes a key aim of the present study.

The Norwegian Context

Norway has a large, ethnically diverse immigrant population and an institutional context characterized by low economic inequality, strong welfare-state institutions, and an open educational system (Esping-Andersen and Myles 2011; OECD 2022). Early waves of labor migration from Pakistan, Turkey, and Morocco started in the late 1960s. Immigration shifted toward refugees arriving from Vietnam, Chile, and Iran in the late 1970s and into the 1980s, and from Yugoslavia, Somalia, and Iraq in the 1990s (Brochmann and Kjeldstadli 2008). In 1970, immigrants and their native-born children made up only 1.5% of the Norwegian population, but this share increased to 20% by 2023 (Statistics Norway 2023). The relative size and ethnic diversity of the Norwegian immigrant population are comparable to those found in other major immigrant-receiving countries in Western Europe (OECD 2022).

Immigrants arriving in Norway as labor migrants and refugees, often with low absolute levels of formal education, tend to experience low employment and labor market difficulties (Bratsberg et al. 2014). Further, child poverty rates are high among many immigrant minorities (Galloway et al. 2015). Children of immigrants from Pakistan, the Middle East, North Africa, and Latin America typically experience disadvantages in school grades and early school-leaving, and children from many East Asian backgrounds outperform natives (Bratsberg et al. 2012). Differences in upper secondary completion rates between children of immigrants and natives have steadily declined since the mid-2000s (Bratsberg et al. 2012; Reisel et al. 2019). Overall, children of immigrants from most national origins complete more education than their foreign-born parents, and their improved educational attainment is converted to considerable intergenerational progress in the labor market (Bratsberg et al. 2014; Hermansen 2016; Reisel et al. 2019).

Equal access to free education is a defining feature of Norway's educational system, which may mitigate adverse consequences of early-life disadvantage on children's life chances (Imsen and Volckmar 2014). Public schools do not charge tuition and are considered high quality (Lauglo 2010). These public schools absorb most students at all educational levels, including students in compulsory education (primary and lower secondary school, grades 1–10), upper secondary education (a four-year vocational track or a three-year academic track leading to a general university and college admission certification), and postsecondary education (three-year bachelor's, two-year master's, and doctorate degrees). No formal ability tracking occurs throughout lower secondary school, and even though students compete for admission to their preferred upper secondary educational track primarily on the basis of prior grades, all children are legally entitled to enter upper secondary education regardless of academic achievements. Consequently, Norway constitutes an institutional context in which academic ambition and educational choice have more influence on children's academic careers relative to settings where students from disadvantaged backgrounds are constrained by early tracking and lack of financial resources.

Data and Variables

Data

We use administrative microdata from Norwegian population registers, which provide an anonymized system of unique personal identifiers allowing us to link information on immigration background and educational outcomes. Further, we link children to their parents, providing information on parental education and other sociodemographic family background characteristics. We link the administrative data to the Barro-Lee (2013) database, which contains information on educational attainment in 146 countries.¹ The information in this database is obtained from UNESCO Statistical Yearbooks, United Nations Demographic Yearbooks, and other national publications. By combining information on individual educational attainment from administrative registers in Norway and educational distributions in origin countries from the Barro-Lee data, we develop a measure of parents' educational status relative to their nonmigrant compatriots in the home country (described later). Prior studies on educational selectivity have used the Barro-Lee data compilation (Engzell 2019; Feliciano and Lanuza 2017; Ichou 2014), considered the most comprehensive data source for information on educational attainment.

We restrict our sample to two groups of children in the 1985–1997 birth cohorts: (1) children of two Norwegian-born parents and (2) children of two foreign-born parents who were born in Norway or immigrated before age 7. We exclude children with parents from different birth countries (i.e., parents of mixed origin). To study postsecondary education enrollment and field-of-study choices, we follow individuals' educational careers up to age 23. Following Engzell (2019), we exclude children whose immigrant parents arrived before age 25 because it is less plausible that younger parents' highest educational attainment was obtained in their country of origin. Finally, children of parents from Kosovo, Bosnia-Herzegovina, and Somalia are excluded because the educational distribution in these countries is not available in the Barro-Lee data set. With these restrictions, our analytic sample contains 673,674 observations. Table 1 presents descriptive statistics separately for children of immigrant parents (n = 19,999) and children of nonmigrant parents (n = 653,675).

Variable Measurement

Our empirical focus is on the educational careers of children of immigrants relative to children of native-born parents. We use a dichotomous measure of children's immigrant background, differentiating between children with two foreign-born parents and children with two native-born parents. Further, we distinguish eight sizable immigrant origin groups: children of Pakistani, Iraqi, Vietnamese, Iranian, Sri Lankan, Chilean, Turkish, and Indian immigrants, and two groups with origins from other non-Western countries and other Western countries (see section A of the online appendix for group sizes).²

We construct a set of educational outcomes measured at age 23. First, postsecondary education enrollment is a dichotomous indicator of whether the individual is enrolled in or has completed a postsecondary degree (yes = 1, no = 0). Second, we study horizontal ethnic stratification by differentiating between enrollment in eight fields of study in postsecondary education: (1) education; (2) law; (3) business and administration; (4) health, shorter programs (e.g., nursing and social work at the bachelor's degree level); (5) health professionals (e.g., dentistry and medicine at the master's level or higher); (6) science and engineering; (7) arts and humanities; and (8) social sciences and other programs. We define an individual's field of education as the first enrolled postsecondary education or currently enrolled education at age 23, whichever is highest. When education at first enrollment and education at current enrollment match in level but are in different fields, we use the field of the currently enrolled education.

Third, we calculate the expected earnings in each field of study to proxy for the individual's expected labor market outcomes based on the chosen field of study. We measure the field-specific expected earnings for each individual using information on the average annual earnings of graduates who are 16 years older than the individual and hold the same educational degree as the individual has chosen. To construct this measure, we extract the older graduates' pretax annual wages from employment and income from self-employment (capital income and social welfare transfers excluded) from highly accurate tax records and average the graduate's annual earnings between ages 35 and 37. Then, we rank the graduates on their earnings (including zero earnings) relative to others in the same birth cohort for males and females combined. This method yields a symmetric variable that captures each graduate's percentile rank (from 1 to 100) in the cohort-specific earnings distribution (Chetty et al. 2014), which is unaffected by inflation trends. For each birth cohort, we collapse the relative earnings rank by a detailed measure of education distinguishing between educational attainment levels and postsecondary fields of study (i.e., three-digit educational codes from the Norwegian version of the International Standard Classification of Education, distinguishing between 105 detailed fields of study) based on their highest completed educational degree at age 35. Finally, we match the average field-specific earnings rank of the older cohorts to the individuals in our analytic sample using our measure of the postsecondary field of enrollment.

Table 2 provides information on the narrow fields of study covered by each of the eight categories, as well as the average expected earnings rank, grade achievement level (discussed in detail later), and student distribution by immigrant background for each field. Students' prior grade achievement (z-standardized values) is higher in educational fields with higher expected earnings. Notably, children of immigrants are highly concentrated in health professional degrees, representing roughly 10% of those seeking such degrees compared with approximately 3% among all postsecondary education enrollees.

Figure 1 shows box plot distributions of students' field-specific expected earnings by immigrant background for the eight broad fields, where within-field variation in expected earnings reflects immigrant–native differences in sorting across the 105 detailed fields of study. Graduates with health professional degrees, including medical doctors and dentists, have the highest expected earnings rank, with median ranks above the 80th percentile in their cohort-specific earnings distribution. Degrees in business and administration, law, and science and engineering also have high expected economic returns, followed by degrees in the social sciences and other programs. Finally, graduates from the arts and humanities, education, and shorter programs in health have expected earnings close to the 50th percentile in their cohort-specific earnings distribution. Overall, immigrant–native differences in expected earnings are relatively small in each of the eight broad fields of study, but children of immigrants enroll in detailed fields of study with notably lower expected earnings in law and the social sciences.

We use two key measures of family background as explanatory variables. First, parental years of education refers to the mean years of parental education when the child was age 16. We recode parental years of education from educational attainment levels, distinguishing nine attainment levels ranging from no education or only primary education, to different levels of secondary and tertiary education, and up to individuals with doctorate degrees (section B, online appendix).

Second, we measure parental educational selectivity following Ichou's (2014) rank-based approach. Thus, we first recode the parental education level when the child was age 16 to the seven educational levels in the Barro-Lee data set. Then, we calculate each parent's educational rank relative to their same-sex and same-age (nonmigrant) reference population in the origin country for the five-year measurement period that overlaps with the year the child turned 16. As Engzell (2019) described, an individual's selectivity is given by

where $pj$ is the proportion of the reference population in the origin country with educational level $j$⁠. Thus, a parent's selectivity value is the sum of the percentage of the parent's origin-country reference population with a lower education level plus half of the percentage of the reference population with an education level equal to the parent's. We use the mean of the mother's and father's values to calculate the individual's parental educational selectivity. Using the same measure, we also calculate parental educational selectivity for children of native-born parents from the Barro-Lee data for Norway.³

Figure 2 displays the relationship between the distribution of educational rank and years of completed education among immigrant parents (cf.Engzell 2019). Panel a shows a strong positive correlation between the individual-level values of the two measures (rho = .681); here, each dot refers to the two aspects of parental education for each child of immigrant parents. Panel b displays the same relationship aggregated to national-origin groups. Those groups above the horizontal line (i.e., the median of the cohort-specific educational rank distribution in the origin country) are, on average, positively selected on education relative to the nonmigrant reference category in the country of origin (y-axis). Groups to the left of the vertical line are, on average, less educated than native-born parents of nonmigrant backgrounds in Norway (x-axis).

Despite the high individual-level correlation, origin groups with high average parental educational rank relative to the origin country often have considerably less parental schooling than the children of Norwegian-born parents. For example, immigrant parents from Pakistan, Vietnam, Iran, Iraq, Chile, and India are positively selected on education. However, except for Chilean and Indian immigrants, they have completed fewer years of education than the national average in Norway. In contrast, Turkish immigrants are negatively selected and have less education than native-born parents, whereas Sri Lankan immigrants have less education than native-born parents but seem not to be selected from their origin country. Immigrants from India are hyperselected (Lee and Zhou 2015): on average, they have higher educational ranks than nonmigrants in their origin country and more education than native-born parents in Norway. Figure 2 documents that a high degree of country-level educational selectivity often corresponds well with the length of schooling in the absolute sense.⁴

To measure students' prior educational achievements, we use students' grade point average (GPA) at the end of compulsory lower secondary education. The GPA is a composite of teacher-assigned grades and exam grades obtained in the year when most children turn 16. It is available for all children, regardless of their later educational attainment. The GPA is standardized within graduating cohorts, with a mean value of 0 and a standard deviation of 1.

Finally, all regression models reported here adjust for basic sociodemographic background characteristics. These covariates are child's sex, mother's and father's birth year, number of siblings, birth order, information available for only one parent in registers, and birth cohort.

Results

Ethnic Inequality in Postsecondary Education Enrollment

Figure 3 summarizes results from ordinary least-squares (OLS) regressions (linear probability models). Using information on complete birth cohorts, the models estimate gaps in the likelihood of postsecondary education enrollment relative to natives for all children of immigrants and separately by country of origin. We present coefficients from four model specifications. Model 1 adjusts for sociodemographic background characteristics (child sex, mother's and father's birth year, number of siblings, birth order, whether information is available for only one parent, and birth cohort). In the remaining models, we successively add controls for parental years of completed education (Model 2), parental relative educational rank (Model 3), and students' earlier grade achievements (Model 4). Models 1–4 allow us to assess baseline immigrant–native gaps in the probability of enrolling in postsecondary education and whether these gaps are attributable to differences in parental education and students' academic achievement. The coefficients are reported as marginal effects (probability changes in percentage points), where the probability of postsecondary enrollment (y  = 1) is assumed to be a linear function of the set of predictors (Wooldridge 2010).

The results show a substantial ethnic advantage. Children of immigrants have considerably higher chances of postsecondary education enrollment than children of native-born parents. If we consider all origin countries combined, children of immigrants have a probability of university-level education enrollment that is 9 percentage points higher, on average. We observe a comparable pattern for most origin countries—Pakistan, Iraq, Vietnam, Iran, Sri Lanka, India, and other non-Western origins—with a likelihood of entering postsecondary education that is 8–21 percentage points higher. However, children of Turkish and Chilean immigrants are considerably less likely to enroll in postsecondary education, whereas children of Western immigrants do not differ from the natives.

Once we adjust for parental years of education, the pattern of ethnic advantage is even more pronounced. A plausible explanation is that parents' educational attainment is not as good a proxy for parental resources transmitted across generations in immigrant families as it is in native families (Luthra and Soehl 2015). For instance, immigrant parents may have less education because of a lack of opportunity in their origin country rather than a lack of interest or ability, whereas native-born parents with similarly low levels of education may be more negatively selected on skills or other unobserved traits. In this case, parental relative education might better reflect immigrant children's socioeconomic background and thus better explain the high enrollment rates. Next, we therefore adjust for parental educational selectivity. Interestingly and in line with the foregoing argument, controlling parental relative educational rank reverses the increased gap found in the model with only parental years of education, suggesting that the ambitious enrollment choices of children of immigrants, relative to those of children of natives with similar parental years of education, are partly due to parental educational selectivity. Nonetheless, substantial immigrant–native gaps in postsecondary enrollment remain after we adjust for both measures of parental education.

Finally, we add controls for prior educational achievement to assess whether differences in enrollment persist when we compare children of immigrants and children of native-born parents with similar achievement levels. This model isolates the role of educational choice (secondary effects) relative to the influence of prior academic achievements (primary effects) in explaining the immigrant–native enrollment gaps by immigrant background. However, adjusting for prior grade achievements does not alter the pattern substantively, and considerable immigrant advantages in postsecondary enrollment remain. Children of immigrants from all origin regions combined are close to 10 percentage points more likely to enter university-level education than natives, net of all adjustments. However, we find comparable or larger advantages of 7–22 percentage points for children of Pakistani, Iraqi, Vietnamese, Iranian, Sri Lankan, Turkish, Indian, and other non-Western immigrants. Only children of Chilean and Western immigrants are less likely than natives to enter postsecondary education once we adjust for family background and prior educational achievements. Thus, Figure 3 reveals a striking and persistent pattern of immigrant optimism. Children of most immigrant origins are substantially more prone to continue to university-level studies. Differences in parental educational selectivity, parents' absolute length of schooling, and students' prior academic achievements are limited in explaining this substantially greater likelihood.

Ethnic Inequality in Choice of Postsecondary Field of Study

Next, we assess horizontal ethnic inequality in sorting across different postsecondary educational fields. Figures 4 and 5 summarize results from OLS regressions (linear probability models) predicting enrollment in a given field of study for children of immigrants relative to children of natives conditional on having entered postsecondary education.⁵ We regress binary indicators of whether a student enrolled in the given field on immigrant background. The coefficients reflect the percentage-point gap in the likelihood of enrollment in a given field of study for children of immigrants relative to natives. Figure 4 reports results for the four lowest-paying fields—defined by the median of the field-specific expected earnings—presented in ascending order from panel a (lowest; arts and humanities) to panel d (highest; social sciences and other). Figure 5 similarly reports estimates for the four fields with the highest expected earnings, presented in ascending order from panel a (lowest; business and administration) to panel d (highest; health professionals). Each panel starts with results from the model specification with basic adjustments and then moves on to models successively adding controls for parental years of education, parental educational selectivity, and prior grade achievement.

The results reported in Figure 4 show that children of immigrants are less likely than natives to enroll in the four fields of study with the lowest expected earnings. This pattern is most pronounced for shorter health degrees (e.g., nursing and social work) and education (e.g., preschool, primary, and lower secondary school teachers). With the exception of immigrant-background students of Chilean and Western origins, this pattern of lower enrollment probabilities is quite consistent across all other non-Western origin groups. Adjusting for parental years of education, parents' relative educational ranks, and earlier grades does not account for these field-specific immigrant–native enrollment gaps and, in fact, slightly accentuates the field-specific negative enrollment gaps relative to comparable students with native-born parents.

In contrast, Figure 5 documents that the probabilities of enrollment in the four highest-paying educational fields for children of immigrants are similar to or higher than those of natives after we make basic adjustments. This pattern is most pronounced for enrollment in health professional fields, which include advanced degrees in medicine, dentistry, and pharmacy. On average, children of immigrants have a probability of pursuing a health professional education that is roughly 9 percentage points higher than that found among native students. Given that almost 12% of immigrant students and less than 4% of native students in postsecondary education are enrolled in health professional fields (cf.Table 2), immigrant students are more than three times as likely to pursue health professional educations as are natives. Broken down by country of origin, this pattern of overrepresentation in health professional fields is similar for all origin groups except for children of Chilean and Turkish immigrants. The remaining fields of study follow a general pattern of higher enrollment among children of immigrants, but variation occurs in which national origin group is overrepresented in business and administration (Pakistan, Vietnam, Turkey, and India), law (Pakistan, Iraq, and Iran), and science and engineering (Vietnam and Sri Lanka). Overall, sorting across fields of study among children of immigrants reveals horizontal ethnic advantage relative to natives in postsecondary education.

We next assess whether parental and student characteristics explain this horizontal advantage. Adjusting for family background and prior academic achievement does not provide a general account of the observed patterns, although we find some variation in the immigrant–native gaps across fields of study. For example, differences in parents' relative educational rank contribute to higher enrollment in business and administration among children of immigrants relative to natives, particularly for children of Pakistani, Vietnamese, and Turkish immigrants. Immigrant selectivity seems to matter considerably less for the three remaining fields of study: law, science and engineering, and health professionals. Perhaps most strikingly, adjusting for parental educational rank seems only to accentuate the enrollment advantage that children of immigrants have in the health professional field.

Our results do not provide clear support for the argument that highly ambitious choices and enrollment in higher-paying educational fields among children of immigrants can be attributed to selective patterns of immigration and their parents' relative social status in the origin country. In contrast, other types of explanations—such as strategic choices anticipating future discrimination or more instrumental views of the value of education—may explain why children of immigrants gravitate toward high-status fields of study with high economic returns.

Ethnic Inequality in the Expected Economic Returns From the Chosen Field of Study

How does sorting across educational fields of study translate to future ethnic differences in the labor market? To address this question, we examine immigrant–native inequality in expected economic returns due to sorting across fields of study among postsecondary education enrollees. Figure 6 summarizes the results from OLS regression models regressing students' expected earnings rank on immigrant background using the same model specifications reported earlier.

Children of immigrants enroll in educational fields with higher economic returns relative to natives after we adjust for sociodemographic characteristics. On average, across all origin groups, children of immigrants enroll in educational fields with expected earnings roughly 1.7 percentile ranks higher compared with natives. For children of immigrants from Pakistan, Vietnam, Sri Lanka, and India, the ethnic advantage in field-specific expected earnings is considerably higher, with a 3.7 to 4.7 higher percentile rank relative to children of natives. Children of Chilean immigrants have markedly lower expected earnings than natives—a negative gap of 5.8—whereas children of Western immigrants also enroll in fields with lower economic returns. The remaining origin groups—from Iraq, Iran, Turkey, and other non-Western countries—enroll in fields with similar or slightly higher expected earnings than children of natives.

How are the estimated gaps affected by adjusting for parental years of completed education, parental educational selectivity, and earlier grades? Controlling for any of these background factors minimally accounts for the immigrant–native gaps in field-specific expected earnings. In most cases, the ethnic advantages for children of immigrants increase once we control for family background and earlier academic achievements.

Overall, these results show that most second-generation immigrant minorities sort into postsecondary educational fields where, all else being equal, they can expect to earn higher wages after graduation than comparable children of natives. However, these results again lend limited support to the argument that immigrants' educational selectivity can explain their children's choice of economically rewarding fields of study.

Discussion and Conclusions

Given the rapid expansion of higher education in recent decades, surprisingly little research has addressed horizontal ethnic inequality in sorting across postsecondary fields of study and the consequences of this sorting for future life chances among children of immigrants. In this study, we use administrative data from Norwegian population registers to investigate differences between children of immigrants and native-born parents in their likelihood of postsecondary enrollment, sorting into different fields of study, and the effect of choice of educational field on their economic prospects. Further, we assess whether immigrant–native differences in family background and earlier academic achievements can explain patterns of horizontal ethnic inequalities in postsecondary education.

We report three key findings. First, we show that children of immigrants are more likely to enroll in postsecondary education than children of native-born parents. Second, and most importantly, children of immigrants have a systematically higher likelihood of self-selecting into fields of study with high expected earnings, such as degrees in business and administration, law, science and engineering, and health professions (e.g., medicine and dentistry). On the basis of ethnic differences in choice of postsecondary educational field, children of immigrants have higher expected earnings than children of natives. Third, differences in earlier academic achievements, parental educational attainment, or premigration social status and skill selectivity do not explain these educational advantages.

Overall, our findings reveal a persistent pattern of horizontal ethnic advantage: despite being disadvantaged in terms of earlier achievements and parental resources, children of immigrants are more likely to enter postsecondary education and pursue more prestigious, economically rewarding fields of study than children of native-born parents. Our findings show immigrant optimism extending beyond that documented in earlier research focusing on continuation into upper secondary academic tracks or higher educational attainment (Glick and White 2004; Heath and Brinbaum 2014; Jackson et al. 2012; Keller and Tillman 2008). Because horizontal stratification across fields of study is increasingly important for graduates' careers (Borgen and Mastekaasa 2018; Kim et al. 2015; Reimer et al. 2008), these salient ethnic advantages in postsecondary education may be central to continued second-generation socioeconomic progress in Norway and other immigrant-receiving countries (Drouhot and Nee 2019; Hermansen 2016).

The ambitious postsecondary educational choices observed among most children of immigrants signal rapid intergenerational assimilation and declining ethnic stratification, but this conclusion should be viewed with caution. If ambitious educational choices are based on a miscalculation of their skills relative to the academic rigor required in the highly demanding study programs they enter, immigrant-background students might run into what has been called an optimism trap, where they face higher risks of noncompletion and later experience adverse labor market outcomes (Birkelund 2020; Dollmann and Weißmann 2020; Tjaden and Hunkler 2017). Yet, admission to high-prestige study programs in postsecondary education tends to be highly selective, drawing on prior grades or test scores as the key selection criteria. In such cases, negative skill selection may be a less relevant factor. Further, graduating from educational fields with high expected earnings does not automatically translate into labor market success (Hout 2012). Given that children of immigrants often face ethnic discrimination and other barriers in the labor market (Hermansen 2013; Quillian et al. 2019), a key task for future research is to assess whether they reap the same economic returns from high-prestige educational degrees as their fellow graduates with native-born parents.

Moreover, immigrant parents' educational selectivity does not explain the ethnic advantages in postsecondary education entry and sorting into highly rewarding fields of study seen among children of immigrants. These children seek opportunities along the horizontal dimension in postsecondary education, but we find no strong evidence indicating that these choices reflect strategic adaptations aimed at reinstating their foreign-born parents' premigration social position (Engzell 2019; Feliciano and Lanuza 2017; Ichou 2014). One alternative explanation for immigrant descendants' ambitious educational choices could be related to immigrant parents' lack of fine-tuned contextual knowledge about the intrinsic value of educational fields (Chiswick 1988; Xie and Goyette 2003), which may lead immigrant parents—irrespective of educational selectivity—to encourage their children to choose more transnationally well-known, economically instrumental, high-status educational fields. Another possible explanation of ambitious choices may be strategic behavior in response to expected discrimination and other barriers in the labor market (Heath et al. 2008; Salikutluk 2016). For example, relative to native children, of non-Western immigrants are much more likely to enroll in health professional fields of study. These fields often lead to licensed occupations, and occupational licensure often reduces ethnic inequality in the labor market (Drange and Helland 2019).

Our findings should be interpreted with caveats in mind. First, our measure of parental educational selectivity does not capture (positive) selection relative to nonmigrants in the origin country on other unobserved characteristics, such as class position, health, or personality traits (Feliciano 2020). We measure selectivity on parents' educational attainment without information on their field of education or their occupational specialization before migration (Potochnick and Hall 2021). Through cultural socialization, these factors are likely to affect children's educational choices (Helland and Wiborg 2019; Jonsson et al. 2009). Further, we measure parents' selectivity as their education relative to the national distribution of educational level in their origin country, not relative to local distributions. Consequently, our measure may be biased regarding parents' self-perceived educational status from a local perspective.

In conclusion, this study documents an optimistic scenario of a horizontal ethnic advantage in postsecondary education among children of immigrants in the Norwegian welfare state context. However, this pattern might not generalize to a broader set of immigrant-receiving countries. The Norwegian educational system is open and comprehensive, lacks early formal tracking, and has few economic barriers to entering high-prestige study programs—factors that arguably provide more room for choice-based mechanisms allowing high-aspiring children of immigrants to advance into postsecondary education (Jackson et al. 2012). Thus, Norway is a best-case scenario, where immigrants' children's ambitious educational choices and potential consequences for labor market careers are likely to be less constrained than in less egalitarian educational systems. Future comparative research is needed to establish the extent to which immigrant-background students' ability to maneuver into high-prestige postsecondary educational fields is more or less constrained by the opportunity structures found in different institutional contexts.

Acknowledgments

Earlier versions of this paper were presented at the International Sociological Association Research Committee 28 (RC28) on Social Stratification and Mobility's Spring Meeting 2021 in Turku; the Nordic Demographic Symposium 2022 in Oslo; the 2022 Norwegian Demographic Association annual meeting; and seminars at the Nordic Institute for Studies in Innovation, Research and Education, University of Oslo, and Stockholm University. We thank Anthony Heath, Magnus Bygren, Nicolai Topstad Borgen, Gunn Elisabeth Birkelund, and participants at the aforementioned events for helpful comments. This research was funded by the Research Council of Norway (grants 236793, 287016, 314249, and 300917) and NordForsk (grant 95263), and is part of the Center for Research on Equality in Education (CREATE; funded by the Research Council of Norway, grant 331640). The study used Norwegian administrative data provided by Statistics Norway. For confidentiality reasons, these data are not publicly available, but they can be accessed upon receiving relevant approvals.

Notes

References