Jump to ContentJump to Main Navigation
Unequal AttainmentsEthnic educational inequalities in ten Western countries$

Anthony Heath and Yaël Brinbaum

Print publication date: 2014

Print ISBN-13: 9780197265741

Published to University Press Scholarship Online: May 2015

DOI: 10.5871/bacad/9780197265741.001.0001

Show Summary Details
Page of
date: 27 May 2020

Gender and Ethnic Inequalities across the Educational Career

Gender and Ethnic Inequalities across the Educational Career

Chapter:
(p.193) 8 Gender and Ethnic Inequalities across the Educational Career
Source:
Unequal Attainments
Author(s):

Fenella Fleischmann

Publisher:
British Academy
DOI:10.5871/bacad/9780197265741.003.0008

Abstract and Keywords

This chapter examines whether the second generation has assimilated to Western patterns of female advantage in education. In contrast to most industrialised societies, which have witnessed a change towards female advantage in education in recent decades, gender gaps in education in ethnic minorities’ origin countries vary greatly, with persistent female disadvantage in world regions where many of the minorities under study originate. Interactions between female gender and ethnic background are examined for the five educational outcomes analysed in the previous chapters, thus covering the entire educational career. The results show that gender gaps among the second generation are on the whole as large and in the same direction as among the majority population. Thus the female disadvantage found in the parental generation disappears in the children's generation and is replaced by the same pattern of female advantage that is found among majority groups in Western countries.

Keywords:   gender, educational inequality, immigrants, second generation, origin country effects, comparative analyses

Introduction

IN RECENT DECADES, WESTERN SOCIETIES HAVE SEEN a reversal in gender inequalities, with girls overtaking boys at all stages of the educational career. In parallel, they have experienced increasing migration inflows from a range of countries where male advantage in education remains very strong. A key question, therefore, is whether ethnic minorities, especially those from more traditional cultural backgrounds, have assimilated to Western patterns of female success, and at what stages of the educational career. Can the legacy of traditional gendered patterns of educational success still be seen among the second generation in Western countries? And does this vary across countries and immigrant groups? These questions will be addressed in this chapter.

Most of the education literature to date has focused either on gender differences or on ethnic or racial differences in education, usually to the exclusion of the other. The few studies investigating gender gaps in education in conjunction with ethnic or racial origin, on the other hand, look at differences between several minority groups, but do not compare them to the majority population (e.g. Feliciano & Rumbaut 2005). Thus, it is unknown whether immigrants' gender-specific performance differs from the patterns observed for the majority.

This chapter takes a different approach by capturing the interaction of gender and immigrant origin in comparison to the majority population. We examine gender differences in achievement and attainment during the school career, drawing on the analyses in previous chapters. In addition to describing migrant-specific gender gaps from a cross-national comparative perspective, we discuss variation across groups and educational contexts.

In line with the preceding chapters, we investigate gender differences in secondary, upper secondary and tertiary education for a variety of indicators of educational success. Achievement in secondary education is measured via (1) test scores or grades, whereas in upper secondary and tertiary education, (p.194) the focus is on attainment. For these stages, we examine (2) the chances to pursue an upper secondary course after completing compulsory schooling (i.e., continuation), (3) the track attended in upper secondary education, (4) completion of upper secondary education and (5) completion of tertiary education.

Since gender-specific analyses are not available for all countries considered in this volume, our account is based on a somewhat smaller selection including Belgium, England and Wales, Finland, France, Germany, the Netherlands, Sweden, Switzerland, and the USA. The study thus covers a selection of European countries representing different forms of migration such as labour and post-colonial migration (Belgium, England and Wales, France, Germany, the Netherlands and Switzerland) as well as more recent refugee migration to Europe (Finland and Sweden). European results are, moreover, compared with findings from the USA as a classical immigration country. The countries under study differ in important institutional features including their integration policies (e.g. Koopmans et al. 2012) and their educational systems (Allmendinger 1989; Kerckhoff 2001; Breen & Buchmann 2002). As documented in Chapter 2 of this volume, we compare more open systems with multiculturalist characteristics (England and Wales, Finland, the Netherlands, Sweden, USA) with systems applying more pressure on assimilation (Belgium, Switzerland, Germany, France; cf. Koopmans et al. 2012), and we contrast comprehensive with stratified educational settings. Yet, all of these Western destination countries share similar gender role orientations. Compared to many of the societies immigrants originate from, there is less emphasis on gendered task distributions.

Rather than aggregating across origin countries, we distinguish as many second-generation groups by their national origin as possible. This allows investigating differences between immigrant groups in the same country, as well as studying the same immigrant group in different countries, such as Turks in Belgium, Germany, the Netherlands, Sweden and Switzerland, or North Africans in Belgium, Finland, France, the Netherlands and Sweden (cf. Van Tubergen et al. 2004 on the importance of disentangling origin and destination countries).

The chapter is structured as follows. The next section provides an overview of previous research on gender gaps in education. We describe gender differences across different educational stages in Western destination countries, considering overall gender inequalities as well as gender differences for certain migrant groups. Subsequently, we describe gender inequalities in education in the countries of origin. The following section discusses explanations for migrant-specific gender gaps in educational achievement and attainment. We apply and extend familiar arguments on gender differences in education, such as gender role socialisation and women's returns to education, to the (p.195) situation of immigrant minorities, while also considering implications for cross-national variation. Next, we empirically examine the interaction between gender and migrant origin for the educational outcomes under study throughout this volume. After summarising the main results, we finally discuss implications for future research.

Gender gaps in education: differences between destination and origin countries

Gender inequalities in Western destination countries

A fairly consistent trend of female advantages in secondary education appears in Western societies (see Buchmann et al. 2008 for an overview). The magnitude of these gender differences, however, varies considerably across countries (Marks 2007; OECD 2009a; Bedard & Cho 2010) in line with other indicators of societal gender inequality (Baker & Jones 1993; Else-Quest et al. 2010), for instance regarding under-representation of women in STEM occupations (Van Langen et al. 2006) and prevalent gender stereotypes (Nosek et al. 2009). Inequality in performance in standardised tests in secondary school is domain-specific: whereas girls tend to attain higher test scores in reading, boys usually score higher in mathematics (Marks 2007; Buchmann et al. 2008; OECD 2009a). Moreover, females receive better grades (Downey & Vogt Yuan 2005; Buchmann et al. 2008). In contrast to these inequalities in secondary education, there are no substantial differences at the beginning of the school career. Instead, gender gaps develop over time with a growing female advantage in reading and a growing male advantage in mathematics (Downey & Vogt Yuan 2005; Entwisle et al. 2007; Buchmann et al. 2008; OECD 2009a).

Gender inequalities in reading and mathematics performance in secondary education tend to be similar across first-generation, second-generation and majority students (OECD 2006; 2009a). The size of these gaps, however, varies considerably across countries (OECD 2006; 2009a). Comparative research on migrant-specific gender inequalities in test scores, as far as we are aware, reports results according to generation status only. The knowledge about the achievements of certain minority groups is therefore still limited to single country studies (e.g. for the USA: Brandon 1991; Coley 2001; for Germany: Segeritz et al. 2010); and these studies do not always include the majority population (e.g. Feliciano & Rumbaut 2005). It therefore remains an open question whether the pattern of gender differences among ethnic minorities is uniform and similar to that of the majority population.

Over the twentieth century, inequalities between men and women in their educational attainment declined markedly (Breen et al. 2010). Females in (p.196) Western societies now consistently outperform males: they have lower dropout and higher graduation rates at the upper secondary level (Eurydice 2009: 243ff; OECD 2009a; 2009b) and they more often enroll in higher education and complete a tertiary degree (Buchmann et al. 2008; Eurydice 2009: 114ff, 245ff; OECD 2009a; 2009b). With few exceptions the once prevalent male advantage in tertiary education has disappeared (OECD 2009a; 2009b). Horizontal stratification, however, persists as reflected in the gender-specific choice of discipline (Bradley 2000; Gerber and Cheung 2008; Eurydice 2009: 250ff).

Studies on gender inequalities in attainment among immigrants and ethnic minorities are scarce. Those available, mostly from the USA, indicate that despite variation in the size of the gap across groups the female advantage usually holds (e.g. for the USA: Brandon 1991; Coley 2001; Buchmann et al. 2008; for Canada: Abada & Tenkorang 2009; for Germany: Riphahn & Serfling 2002).

Gender inequalities in the countries of origin

Gender inequalities in the countries of origin are often in stark contrast to those in the countries of destination, particularly in less-developed regions. Overall, gender equity is reported to be inversely related to poverty such that gender gaps to the advantage of males are largest in the poorest countries and among the poorest segments of the population (Wils & Goujon 1998; Birdsall et al. 2005; Grant & Behrman 2010). The female disadvantage in education is particularly large in rural areas and among ethnic and linguistic minorities (Birdsall et al. 2005).

Sub-Saharan Africa, South Asia, the Middle East and North Africa represent areas where girls are still strongly disadvantaged with regard to participation in education and adult literacy rates (Wils & Goujon 1998; Birdsall et al. 2005; Grant & Behrman 2010). Many immigrants in the countries considered in this volume trace their origins to these world regions. The destination countries under study, on the other hand, are all characterised by female advantage. The contrast between immigrants' origin and destination countries is illustrated in Table 8.1, which shows the rates of female and male adult literacy and gross enrolment. Data are drawn from statistics provided by the United Nations Development Programme as part of the Human Development Index (UNDP 2010).

Table 8.1 displays modest gender differences in adult literacy and gross enrolment rates in Southern and Eastern European countries as well as considerable female disadvantages among typical origin countries and regions such as Turkey, India, Pakistan, Bangladesh, Iran, Iraq and all of Africa. However, Table 8.1 also shows that, in the countries and regions where women (p.197)

Table 8.1. Adult literacy and enrolment in countries of destination and countries of origin.

Adult literacy

Gross enrolment

Female

Male

Difference

Female

Male

Difference

Majority (destination countries)

Belgium

N/A

N/A

N/A

95.9

92.8

3.1

Finland

N/A

N/A

N/A

105.1

97.9

7.2

France

N/A

N/A

N/A

97.4

93.5

3.9

Germany

N/A

N/A

N/A

87.5

88.6

−1.1

Netherlands

N/A

N/A

N/A

97.1

97.9

−0.8

Sweden

N/A

N/A

N/A

99.0

89.8

9.2

Switzerland

N/A

N/A

N/A

81.4

84.0

−2.6

England and Wales

N/A

N/A

N/A

92.8

85.9

6.9

USA

N/A

N/A

N/A

96.9

88.1

8.8

East Asia

China

90.0

96.5

−6.5

68.5

68.9

0.4

East Asia

93.5

97.6

−4.1

79.5

84.5

−5.0

Southeast/South Asia

Bangladesh

48.0

58.7

−10.7

52.5

51.8

0.7

India

54.5

76.9

−22.4

57.4

64.3

−11.5

Pakistan

39.6

67.7

−28.1

34.4

43.9

−9.5

South Asia

56.3

73.6

−17.3

60.4

64.6

−4.2

Southeast Asia

81.5

89.3

−7.8

68.0

69.5

−1.5

West Asia

Iran

77.2

87.3

−10.1

73.0

73.4

−0.4

Iraq

64.2

84.1

−19.9

52.1

68.5

−16.4

Turkey

81.3

96.2

−14.9

66.3

75.7

−9.4

West Asia

81.7

90.9

−9.2

75.7

73.0

2.7

East Europe

East Europe

98.3

99.0

−0.7

83.7

79.6

4.1

Ex-Yugoslavia

96.2

99.1

−2.9

82.9

77.3

5.6

Former Soviet Union

99.2

99.6

−0.4

85.2

79.8

5.4

Poland

99.0

99.6

−0.6

91.4

84.2

7.2

South Europe

Italy

98.6

99.1

−0.5

94.7

89.1

5.6

South Europe

95.9

97.1

−1.2

87.9

84.4

3.5

West Europe

Denmark

N/A

N/A

N/A

105.3

97.6

7.7

Finland

N/A

N/A

N/A

105.1

97.9

7.2

Norway

N/A

N/A

N/A

102.4

92.4

10.0

Africa

Africa's Horn

37.9

63.1

−25.2

31.2

40.7

−9.5

North Africa

63.0

81.7

−18.7

76.8

75.9

0.9

Sub-Saharan Africa

53.7

69.5

−15.8

49.1

56.0

−6.9

Caribbean

Caribbean

90.2

90.2

0.0

79.4

73.4

6.0

South America

Chile

96.5

96.6

−0.1

82.0

83.0

−1.0

Mexico

91.4

94.4

−3.0

79.0

81.5

−2.5

South America

89.0

91.6

−2.6

80.8

78.3

2.5

Source. UNDP 2010.

Notes. The gross enrolment ratio is defined as the share of all children of school age in the population that are enrolled in school; values above 100 per cent indicate delays or delayed enrolment. Whenever the data sources do not allow us to make more fine-grained distinctions regarding the countries of origin, we refer to regions. For these cases, we averaged data of single countries across regions. N/A = not available.

(p.198) are most disadvantaged, male advantages in enrolment rates are generally smaller than those in adult literacy. This attests to the changing gender gaps in the origin countries with smaller male advantages or even female advantages in younger cohorts as compared to the adult population. Given strong regional variation in gender disparities in education, it seems most interesting to investigate the relative success of sons and daughters of immigrants from areas where women are most disadvantaged (i.e. South and West Asia and Africa) in the educational systems of the destination countries. If gender gaps in the education of these second generation groups are similar to those prevalent among the majority population in their countries of residence, this would imply a sometimes dramatic decrease of the female disadvantage in education in the course of a single generation. Given the discrepancies between many immigrants' origin and destination countries, the key question of our study is whether minorities, especially those from more traditional cultural backgrounds, have assimilated to Western patterns of female success in educational achievement and attainment.

Explaining gender differences in education

General arguments to explain gender gaps in education

The arguments presented in contributions on gender inequalities in education can be summarised under two overarching themes: (1) gender role socialisation and (2) women's returns to education.

The gender role socialisation argument basically states that gender stereotypes and norms influence the socialisation of boys and girls within families and educational institutions (Buchmann et al. 2008). Parents and educators may have gender-biased perceptions of children's abilities and performance; for example, that boys have a better grasp on technical questions or on mathematics whereas girls more easily master languages (Schofield 2006). These differential perceptions can translate into biased educational expectations. Simultaneously, they may lead to behaviour that influences the intellectual growth of children for whom parents and/or teachers have biased expectations for specific competences. For instance, encouragement, praise or criticism could vary according to gender and competence realm and the toys or materials presented might be different for girls and boys. In addition, parents and teachers may ask for different responses to the presented material depending on the child's gender (Harris & Rosenthal 1985; Babad 1993). Other behavioural manifestations include sex-role modelling, uneven access to parental resources or differences in parental involvement or parenting styles (Buchmann et al. 2008). In sum, male and female students receive different (p.199) education experiences on the basis of a combination of what parents and educators believe to be appropriate gender-based behaviour (Tournaki 2003).

Gender stereotypes and norms in turn may influence students' perceptions of their own abilities as well as their educational expectations. They might create a self-fulfilling prophecy (Merton 1948; Rosenthal & Jacobson 1966). One prominent example to explain girls' lower performance on standardised tests in mathematics is stereotype threat. Stereotype threat implies that awareness of a negative stereotype about the performance of one's group in a particular academic domain impairs the performance of members of the group on tests in the relevant domain (Steele 1997).

Gender role socialisation also contributes to gender differences in school-relevant behaviour. On average, girls are found to have more social skills, to be more attentive, to put more effort into schoolwork and to be less disruptive in class than boys (Downey & Vogt Yuan 2005; Buchmann et al. 2008). These gender differences in non-cognitive skills are central in explaining why in some domains girls get higher grades even though boys achieve higher test scores (Downey & Vogt Yuan 2005; Buchmann et al. 2008: 322ff). Moreover, typical out-of-class activities of girls, such as reading or playing a musical instrument, may help them to do well in subjects like language. Boys in contrast more often engage in extracurricular activities contributing to their performance in mathematics and natural sciences (Downey & Vogt Yuan 2005).

The gender role socialisation argument seems more important in accounting for gender differences in performance than in attainment. Nevertheless, it is relevant for attainment, albeit indirectly. Prior performance as reflected in test scores or grades as well as rewards for school-relevant behaviour stimulate girls' educational aspirations for attainment, thus reinforcing the female advantage that develops over the school career (Buchmann et al. 2008). Since the nine Western destination countries under study share similar gender role orientations, we do not expect substantial cross-national differences in the extent to which girls and boys are socialised differentially within these school systems.

The second important theme in the literature on gender inequalities in education refers to women's returns to education. Increasing returns in terms of labour market outcomes and social class attainment have been described as the major reason for the disappearance or reversal of the gender gap in attainment (Breen & Goldthorpe 1997; DiPrete & Buchmann 2006). This development has been accompanied by sociocultural changes in gender roles and expectations about life course trajectories for men and women with a declining number of individuals in Western countries expressing support for traditional gender roles (Buchmann et al. 2008). The reversal is related to shifts in the structure of the labour market, such as the declining gender wage gap, (p.200) especially for women with high levels of human capital, and the decreasing occupational sex segregation (Buchmann et al. 2008). In countries where labour market returns for women are more similar to those of men, such as in the USA or Sweden, women should show more favourable educational attainments than in countries where gender gaps in the returns to education are larger, such as Belgium, Germany or Switzerland (e.g. Mandel & Semyonov 2006; Evertsson et al. 2009).

Extending the arguments to the second generation

This general reasoning on gender inequalities in education can be applied to both ethnic minorities and the majority. However, the question remains whether gender disparities in education differ between groups, especially when studying minorities from countries where gender disparities profoundly diverge from the patterns of female success in Western destination countries (see Table 8.1). We address this question by applying and extending the arguments on gender role socialisation and women's returns to education to children of immigrants from more traditional societies.

We first consider how traditional gender roles might affect the educational performance of second-generation girls and boys. Traditional orientations typically imply gendered task distributions where males are expected to succeed in the labour market and females at home (Shelton & John 1996; Van der Lippe & Van Dijk 2002). In these instances, the returns to education are higher for boys. In addition, more stringent family obligations might impede the school success of girls if household duties and caring work in the family reduce the time available for educational investments (Fuligni et al. 1999). Boys may also be better off in school if their friendship networks differ from those of girls. If females are expected to be at home, they may have fewer opportunities for establishing inter-ethnic contacts, whereas boys have more opportunities to get into contact with majority peers (Dion & Dion 2001). Since exposure is important for language learning (Chiswick & Miller 2001) and good command of the destination country's language is a crucial prerequisite for school success (Esser 2006; OECD 2006), boys might be in an advantageous position. According to these arguments, the female advantage in education that is typical for majority members in Western countries should be smaller or reversed for minorities from more traditional backgrounds. We should also see marked differences between minorities from Western countries and migrants from rural, less-developed areas.

However, this does not seem to be the complete story. More household duties and closer monitoring of daughters as compared to sons could also imply that girls spend more time at home and therefore invest more in their (p.201) education. Gender role socialisation in traditional families may also stimulate females more than males to be obedient to authorities and girls' activities out of school are often under more parental control than those of boys (Abada & Tenkorang 2009). To the extent that closer supervision and strict parental monitoring facilitate encouragement and discipline, the maintenance of these attributes promotes the academic achievement of girls (Zhou & Bankston 2002; Abada & Tenkorang 2009: 585). Higher standards for good behaviour expected from girls in the home could thus carry over into good behaviour and achievement in school (Zhou & Bankston 2002; Feliciano & Rumbaut 2005).

Finally, if children do not adopt their parents' traditional gender ideology, the implications are different for girls and boys. Pursuing higher education and economic independence are a means for girls to turn against traditional gender roles, while boys more often resort to poor performance and resistance in school (Cammarota 2004; Feliciano & Rumbaut 2005). According to this reasoning, second-generation girls should do better in school than their male counterparts so that their performance is more similar to that of females in the country of destination.

Thus, opposite expectations can be derived from the gender role argument. The relative impact of different aspects of gender role socialisation can only be disentangled by empirically investigating each of them separately. However, our data sources do not allow us to address these explanations directly. Thus if opposing processes are central to the explanation, it may well be the case that they do not show up in an overall assessment of gender disparities in education.

The second line of reasoning linking traditional origins with gender inequalities in attainment addresses women's returns to education. Female students from traditional backgrounds experience a contrast between the countries of origin and destination in the status and opportunities that can be achieved through education. Whereas attainment is less productive for females in a traditional context, the education of women in the destination country promises a greater payoff. The contrast between the countries of origin and destination thus becomes a powerful motivator to succeed (Brandon 1991; Feliciano & Rumbaut 2005; Abada & Tenkorang 2009). This explanation of the higher educational attainment of women from traditional families, however, assumes that migrant women compare themselves to women in the country of origin. Since we only consider the second generation, i.e. children of immigrants who are born in their parents' destination country or moved there before entering primary school, it can be questioned whether this assumption holds as strongly as we would expect for female immigrants of the first generation.

In addition, if certain minority groups expect employer discrimination when searching for jobs or apprenticeships upon completing upper secondary (p.202) education, their opportunity costs of continuing into higher education are lower than for the majority population—as long as the returns to education for a tertiary degree do not differ across groups (Heath et al. 2008). Yet, this point becomes relevant for our account of migrant-specific gender inequalities only where discrimination is gender specific. For example, in Germany men of immigrant origin have more difficulty finding an apprenticeship position than their female counterparts (Diehl et al. 2009). As a result, the gain of pursuing an academic career should be relatively higher for second-generation males. Female students who wear a headscarf may find themselves in a similar situation. They could be at a disadvantage in the job-searching process, which in turn might increase their motivation to stay on in education. This kind of reasoning only applies to certain groups or even certain members within an ethnic minority group and it might be most relevant for attainment in later stages in the educational career. In contrast, the argument on differential gains that females can achieve through education in traditional versus the destination societies should hold more generally. Contrary to the gender role socialisation argument, which is primarily relevant in accounting for achievement differences of ethnic minorities from traditional backgrounds, the reasoning on educational returns is more important for inequalities in attainment, in particular at later stages in the academic career. In that sense, the arguments derived from gender role socialisation may be thought of as primary effects of gender, impacting the differential performance of girls and boys in majority and immigrant families. Similarly, the gender gaps in attainment that are expected as a consequence of gender-specific returns to education (and their changes over time and in comparison to migrants' origin countries) could be conceptualised as secondary effects of gender, affecting girls' and boys' choices within the educational system (cf. Boudon 1974).

Finally, destination countries with more open integration policies, such as England and Wales, Sweden or the USA, may allow traditional gender preferences to be continued. In contrast, more dirigiste systems, such as the French or German ones, could provide a greater force for assimilation (Koopmans et al. 2012). As a result, gender inequalities for groups stemming from the same region may differ across receiving contexts. The male advantage that is typical for traditional societies might be sustained longer in open systems than in systems where the emphasis on assimilation is greater. Moreover, policies that enforce equal opportunities for women and men may add to the story. In countries such as Sweden or France, where the labour market participation of women is particularly high, females from traditional families may find greater incentives for educational attainment. In these settings, the education of second-generation girls from traditional backgrounds might become more similar to that of their female peers from the majority (p.203) than in countries where equal opportunities are not implemented to the same degree. Institutional conditions may thus work in opposite directions, as in the Swedish case where a more open political system that allows immigrants to continue gender preferences combines with a strong equal opportunity system.

Empirical findings of gender gaps in the education of the second generation

In this empirical section of this chapter, we examine ethnic variation in gender gaps for the five different educational outcomes analysed in the preceding chapters. We start with the academic performance of adolescents, looking at test scores or grades of students aged 15–16. We can compare gaps in performance between male and female students of the majority population and of numerous ethnic minority groups across nine countries: Belgium, England and Wales, Finland, France, Germany, the Netherlands, Sweden, Switzerland and the USA. In the next step, we examine continuation rates after compulsory schooling in seven countries (England and Wales, Finland, France, Germany, the Netherlands, Sweden, USA). The third set of analyses concerns enrolment in academic as compared to vocational tracks of upper secondary education. As the US school system does not distinguish between such tracks, the analysis at this step is restricted to seven European countries (Belgium, England and Wales, Finland, France, Germany, the Netherlands, Sweden). Lastly, we look at educational attainment, by examining completion of upper secondary and completion of tertiary education, each across a set of countries that could provide results (Belgium, England and Wales, Finland, the Netherlands, Sweden, USA).

Gender gaps in the second generation: Double disadvantage?

Although our main focus is on variations in gender differences in education across different second-generation groups and countries, our analysis needs to address ethnic penalties and premia as well. Ethnic penalties in education can be defined as the net disadvantages experienced by ethnic minorities after controlling for parental education and occupational status (Chapter 1 of this volume; Heath & Cheung 2007). In many cases discrepancies between second-generation groups and the majority population are indeed penalties, but some groups also show consistent ethnic premia (i.e., their members out perform the majority). Linking this information with gender gaps allows illustrating whether females or males from a certain group face an advantage (p.204) or disadvantage in addition to the penalty or premium they face as members of a minority group.

For example, in certain second-generation groups females encounter both an ethnic penalty related to their membership in the minority group and an additional disadvantage associated with their gender, in the form of either a smaller female advantage or a female disadvantage—as compared to the consistent female advantage in the majority population. We refer to this situation as a double disadvantage for girls—following the terminology that is used in studies of immigrant women's position on the labour market (Raijman & Semyonov 1997). The logical counterpart to double disadvantage would be double advantage for girls, which is the case when the minority group performs better as a whole and the female advantage in this group is larger than in the majority population. Obviously, many other combinations are possible as well, such as an ethnic penalty combined with a female advantage, an ethnic premium combined with a female disadvantage, only ethnic penalties/premia etc.

In order to summarise the wide range of findings across countries and ethnic groups under study, we graphically depict effects of gender and ethnicity and their interactions for each educational outcome, while controlling for parental education, occupational status and family composition in Figures 8.28.6 (see also Tables A8.1A8.5 in the appendix). These two-dimensional graphs show disparities between males and females on the vertical axis, while disparities between ethnic minorities and the majority are displayed on the horizontal axis. Groups with female advantage are situated above the horizontal axis and groups with male advantage below. Ethnic penalties in education are evident by placement on the left hand side of the vertical axis, whereas ethnic premia place minorities on its right hand side. Data points are labelled with a combination of numbers referring to destination countries and letters referring to origin countries (see Figure 8.1 for a legend). In addition to the data points, each quadrant in Figures 8.28.6 includes a list of those country/group combinations that are located in the respective quadrant. Coefficients located on or deviating less than 0.05 from the axis are listed to the end of the respective axis. For example, the majority populations are always located on the vertical axis because they form the reference category for the analyses of ethnic penalties or premia. These lists of country/group combinations in each quadrant provide a rough summary of the distributions and complement the more detailed picture of data points. We show and discuss significant as well as non-significant interactions and main effects, because the largely divergent number of cases across countries would disturb the comparison if we were to include significant interactions only. Non-significant interactions and main effects are enclosed by square brackets throughout.

(p.205) Academic achievement

Regarding students' performance on standardised tests or grades, we observe female advantage among the majority populations in all countries under study, with considerable gender gaps in Belgium, England and Wales, Finland, France, Germany, Sweden, Switzerland and the USA, but a negligible gap in the Netherlands. The prevalent female advantage in reading test scores and grades is in line with other findings (Buchmann et al. 2008; OECD 2009a). The absence of a substantial gender gap in the Netherlands could be related to the outcome measure under study, which is a standardised test that includes different subject matters, averaging across subjects in which girls consistently show higher achievement (languages in particular) and subjects in which boys do better (maths in particular).

As Figure 8.2 shows, the female advantage in educational achievement among the majority population is consistently replicated among the minority groups. Gender gaps in the second generation are mostly similar to those of the majority population, and most interactions between gender and ethnic background are not statistically significant. Rare exceptions to this general pattern of female advantage in academic achievement are found for students of Turkish origin in the Netherlands, Iberians in Switzerland as well as Mexicans and other Asians in the USA. In these groups, boys are found to outperform or perform on a par with girls, placing them on the horizontal axis or on the lower half of the graph. Furthermore, most groups are located in the left upper quadrant, implying that a combination of female advantage and ethnic penalties in academic achievement is the most frequently observed outcome.

Three findings from the wide range of results regarding academic achievement among all groups and countries under study can be highlighted. First, given the prevalent and persistent female disadvantage in education in the origin countries of many minorities under study (see Table 8.1), the finding of fairly consistent female advantages in the educational achievements of the second generation is remarkable and suggests that the second generation has assimilated to the gendered patterns of educational achievement in the destination countries. Second, this female advantage in test scores and grades is combined in many cases with an ethnic penalty so that the second generation is often lagging behind the majority in terms of academic achievement. Third, the fact that the same ethnic minority performs differently in different countries both with regard to ethnic penalties and gender gaps means that explanations focusing on origin country effects find little support in our data. (p.206)

Gender and Ethnic Inequalities across the Educational Career

Figure 8.1. Legend for Figures 8.28.6.

(p.207)
Gender and Ethnic Inequalities across the Educational Career

Figure 8.2. Gender differences in standardised test scores or grades at age 15–16 (OLS regression).

(p.208) Continuation

Are there gender differences in continuation of full-time education after compulsory schooling and, if so, do they vary across groups and countries? Figure 8.3 shows that gender differences in continuation are modest among the majority, but to the advantage of girls. Finland forms an exception where a non-significant male advantage in continuation is found among the majority population. The pattern of prevalent female advantage applies equally to most minority groups in most countries. There are some exceptions, however, where boys more often continue in full-time education than girls (e.g. ex-Yugoslav, North African and Swedish-speaking minorities in Finland; Greeks in Germany; Caribbeans, Chinese and Bangladeshis in England and Wales; second-generation blacks in the USA). Double disadvantage for girls is found among ex-Yugoslavs and North Africans in Finland. We also find cases of a double female advantage (e.g. Sub-Saharan Africans in England and Wales as well as East and Southeast Asians in Sweden). At the same time, Greeks in Germany, Caribbeans and Chinese in England and Wales as well as second-generation blacks in the USA combine the female disadvantage with an ethnic premium. In contrast, Sub-Saharan Africans and Caribbeans (i.e. Dom-Tom) in France as well as the Polish second generation in Germany and North Africans in Sweden combine female advantage with an ethnic penalty. Again, most groups are located in the left upper quadrant of the graph, resulting from the combination of female advantage and ethnic penalties in continuation rates.

Overall, variation in gender gaps is more limited than ethnic variation with regard to continuation in full-time education after the completion of compulsory schooling. At the same time, second-generation groups cluster together according to the country in which they are located. We observe, for example, consistent ethnic premia among the various second-generation groups in England and Wales. This contrasts with the results for Finland and the Netherlands, where only ethnic penalties are found. With the exception of Greeks, the finding of prevalent ethnic penalties also applies to the second generation in Germany. In Sweden, on the other hand, the results are most mixed with ethnic penalties among Turkish, Chilean, Nordic and other Western groups and ethnic premia among Iranian and several Asian second generation groups. The minority groups in France are evenly split between ethnic premia (South Europeans and North Africans) and penalties (Sub-Saharan Africans and Caribbeans), but all groups share a female advantage that is larger than in the French majority population. Finally, findings in the USA are most mixed as second-generation groups are found in all quadrants except the male advantage-ethnic disadvantage (lower left) quadrant. (p.209)

Gender and Ethnic Inequalities across the Educational Career

Figure 8.3. Gender differences in continuation (probit regression).

(p.210) Summarising, we note that female advantages are modest among the majority populations in the countries under study. Many minority groups show ethnic penalties or premia in continuation, but in most cases gender gaps are similar to those of the majority. A notable exception to the overall trend of assimilation to the gender gap of the majority is found among second-generation blacks in the USA, where boys have a significantly greater likelihood to continue than third or higher generation white boys, whereas girls in this minority group are much more likely to drop out than comparable girls from the reference group. With the exception of Turks, we find that the outcomes within one ethnic group vary across destinations.

Track in upper secondary education

The school systems of most European countries distribute students over academic and vocational tracks at the upper secondary level. Since upper secondary education is often not compulsory, the analysis of the track is restricted to the subsample of students who continue to pursue full-time education after completing compulsory schooling. As Figure 8.4 shows, majority girls in all countries under study more often attend academic tracks than majority boys. This female advantage is replicated among most minority groups. Exceptions are ex-Yugoslavs and Italians in Germany, Turkish minorities in Belgium as well as Caribbeans in the Netherlands; in these groups, boys more often attend academic tracks than girls. While the Turkish second generation in the Netherlands does not show a significant male advantage, the non-existent gender gap in this group implies a relative disadvantage of Turkish girls, compared to the female advantage of the majority population and other minority groups.

In stark contrast to the results presented so far, ethnic premia in attendance of academic tracks are more common than ethnic penalties. The prevalence of ethnic premia highlights the selectivity of the analysed population, which excludes all those individuals who left full-time education after the compulsory stage. Because of these advantages, most groups are found in the upper right quadrant of Figure 8.4 (i.e. a combination of female advantage and an ethnic premium). Exceptions include ex-Yugoslav and Swedish-speaking minorities in Finland as well as the second generation from the former Soviet Union in Germany, where we find ethnic penalties in the attendance of academic tracks. Moreover, as Figure 8.4 shows, the female advantage is larger for most minority groups compared to the majority. In England and Wales and Sweden, this larger female advantage almost consistently goes together with an ethnic premium, thus showing a double advantage for girls. Among most second-generation groups, ethnic advantages in track attendance are more pronounced than female advantages. (p.211)

Gender and Ethnic Inequalities across the Educational Career

Figure 8.4. Gender differences in track followed in upper secondary education

(probit regression).

(p.212) Completion of upper secondary education

Regarding the completion of upper secondary education, Figure 8.5 shows female advantages among the majority populations in all countries except the Netherlands.1 This pattern is replicated among the various second-generation groups in all countries, with the exception of second-generation whites, blacks and East Asians in the USA. However, the size of the female advantage is not equal across groups and is combined with strong ethnic variation in completion rates. Similar to the findings regarding academic achievement and in contrast to the findings regarding attendance of academic tracks, ethnic penalties are more common than ethnic premia at this point in the school career, placing more groups on the left hand side of the vertical axis than on the right hand side.

Figure 8.5 further shows that the female advantage in completion of upper secondary education is larger for a number of minority groups where girls outperform majority girls, while boys have lower completion rates than the majority (e.g. Turks in the Netherlands and Sweden; North Africans in the Netherlands and Sweden; West Asians, Iraqis, ex-Yugoslavs, Africans from Africa's Horn and the Sub-Sahara in Sweden; as well as Caribbeans in the Netherlands and in France; Filipinos and South Asians in the USA). In these cases, minority boys face a double disadvantage. The prevalence of this pattern clearly differs from the preceding stages in the school career and is the most notable exception to the general rule of assimilation to gendered patterns of educational achievement and attainment. Premature drop out or early exit from full-time education occurs at a crucial stage of the educational career and seems to be a major hurdle for a considerable number of second-generation males. However, this finding does not apply to all minorities alike. For instance, the East Asian and black second generation in the USA shows the reversed pattern of greater likelihood of completion among minority boys compared to majority boys, but this ethnic premium is cancelled out by a negative interaction with female gender among girls of the same ethnic background.

Another common pattern is the combination of an ethnic penalty with a gender gap that is similar to the gap in the majority population (e.g. among all minority groups in Belgium, West Europeans and Chileans in Sweden). A smaller number of groups display an ethnic premium and a similar gender gap as in the majority population (e.g. South Americans and Southeast Asians in Sweden or Indians, Pakistanis and Chinese in England and Wales). (p.213)

Gender and Ethnic Inequalities across the Educational Career

Figure 8.5. Gender differences in completion of upper secondary education

(probit regression).

(p.214) Finally, we also observe double advantage for girls for a few groups due to a combination of an ethnic premium and a particularly large female advantage in the completion of upper secondary education (e.g. North Africans in France, East Asians in Sweden and other Asians in the USA).

Completion of tertiary education

The last educational outcome under study is the completion of tertiary education; results are shown in Figure 8.6. Substantial female advantages are found among the majority populations in Belgium, Sweden and the USA, whereas in England and Wales there is no gender gap, and in the Netherlands it is in favour of males.2 Regarding the second generation, the general trend of female advantage in the completion of tertiary education is reversed among non-Mexican Hispanics in the USA, Chinese and Southeast Asians (Bangladeshis, Indians and Pakistanis) in England and Wales, as well as Turks and Moroccans in the Netherlands. Among all other minority groups, second-generation girls have higher rates of completing tertiary education than boys, thus showing prevalent female advantage. In the case of second-generation women in Sweden we even observe double advantages—yet these results refer to enrolment, not completion of tertiary education. Many groups in Sweden are located in the upper right quadrant of Figure 8.6, implying prevalent ethnic premia, with the notable exception of the Danish second generation, which combines a female advantage with an ethnic penalty. In Belgium, as well as among Caribbean minorities from Suriname and the Dutch Antilles in the Netherlands and Puerto Ricans and Mexicans in the USA, the combination of female advantages with ethnic penalties is more prevalent.

Discussion and conclusion

The key question of this chapter was whether the second generation has assimilated to the pattern of female advantage in educational achievement and attainment that is prevalent in the destination countries under study, despite widespread and persistent female disadvantage in many origin countries. Overall, our comparative findings across up to nine destination countries and (p.215)

Gender and Ethnic Inequalities across the Educational Career

Figure 8.6. Gender differences in completion of tertiary education

(probit regression).

(p.216) a large number of second-generation groups provide an affirmative answer to that question.

The five educational outcomes under study provide an overview of gender and ethnic disparities throughout the educational career. From the first outcome on (test scores or grades at age 15 or 16), we observed female advantages in education among both the majority population and most minority groups. Regarding the second outcome under study, continuation in full-time education after compulsory school, our results indicate a more modest female advantage. It must be noted, however, that the analyses of continuation do not control for prior achievement. Assuming that higher grades or better test scores increase the likelihood to continue and taking into account that girls are consistently found to have higher grades and test scores, this suggests that the gender gap in continuation may be even more modest or reversed when taking prior achievement into account.3 Among later educational outcomes (i.e., enrolment in academic versus vocational tracks, completion of upper secondary and of tertiary education), however, female advantages of similar size as those in school achievement are again found in most countries. These findings suggest that the gender gap is not widening throughout the educational career or after certain transitions. Rather, with the exception of continuation, significant and stable female advantages are found throughout the educational career. A possible explanation for the relatively modest size of the gender gap in continuation rates may lie in the high rates of continuation, with more than 80 per cent or even 90 per cent (in Finland) continuing after compulsory school—where continuation is nearly universal, there is little room for gender disparities.

With very few exceptions, we found no significant interactions between gender and ethnicity, which implies that the female advantage of the majority populations generalises to most ethnic minorities in the second generation in most countries and for most educational outcomes. The only educational outcome where we observed an exception to this general pattern of findings was upper secondary school completion. For this crucial outcome, ethnic minority boys among a number of second-generation groups turn out to be more disadvantaged due to a larger female advantage combined with ethnic penalties. Hence, where we encountered deviations from assimilation to the gendered patterns of educational achievement and attainment, these are to the disadvantage of second-generation boys rather than girls. Nevertheless, we found no evidence for a consistent double disadvantage for girls or boys such (p.217) that they are systematically penalised on or benefit from both their ethnicity and their gender throughout their educational career. However, in contrast to the prevalent assimilation to gendered patterns of educational success, the manifold ethnic differences—both penalties and premia—in educational outcomes that we found highlight that the second generation does not yet perform on a par with their peers from the majority population.

The expectation of double disadvantages was derived from the gender role socialisation argument as an explanation for differential gender gaps in educational achievement and attainment among children of immigrants from countries with more traditional gender values. To the extent that parental norms about gendered task distributions encourage boys while discouraging girls to succeed in education, we would expect a smaller female advantage or even male advantage among minorities from more traditional countries, particularly with regard to attainment. On the other hand, traditional gender values may also benefit girls more than boys in terms of their educational achievement, to the extent that stricter parental monitoring and more time spent at home contribute to better schooling outcomes. The overall null-finding, i.e. the absence of interactions between gender and ethnicity, may imply that gender role socialisation in immigrant families has no influence on the educational outcomes of the second generation; however, it may also be the case that opposing influences are at work that cancel each other out. This should be tested in future research that empirically assesses the role of gender role values in the education of the second generation.

With regard to educational achievement, some results are in line with the notion that gender role socialisation may explain the differential performance of second-generation boys and girls. For instance, a smaller female advantage or even female disadvantage was found among Africans in Finland and France or Turks in Belgium and Germany. Given the persistent female disadvantages in education in Turkey and Africa (see Table 8.1), these findings could suggest that gendered norms about educational performance and differential stimulation of boys and girls are maintained after migration among immigrant parents. However, many other findings are at odds with this interpretation. On the one hand, female disadvantages are not limited to minority groups who originate in societies with large gender disparities in education, but are also found among the second generation from ex-Yugoslavia in Finland and Switzerland and from the Nordic countries in Sweden—origin countries that are characterised by gender parity or female advantage in education. Moreover, groups originating in countries with persistent female disadvantages in education also exhibit female advantages in some destination countries, for instance North African girls in Sweden and Indian and black African girls in England and Wales. The latter findings also imply that there is little support for the assumption that traditional gendered patterns of (p.218) educational achievement and attainment are more easily supported in more open school systems, such as in Sweden and England and Wales, as against systems that more strongly enforce assimilation as in Germany or France.

The fact that second-generation girls whose parents migrated from more traditional countries are not consistently lagging behind boys in terms of academic achievement may imply a change in parental norms and behaviour. However, gender role socialisation is not limited to the parental home—teachers, schools and the wider societal context may also influence gendered patterns of behaviour and related educational outcomes. In cases where the prevalent norm of gender equality in institutions of the majority society stands in contrast to traditional gender norms among immigrant parents, we would expect a stronger influence of parental gender norms on educational outcomes that are open to decisions taken by students and their parents. A good example of such a decision is whether or not to continue education after compulsory schooling, at least in educational systems where this decision is not heavily restricted by prior achievement (e.g. in the comprehensive systems in Finland and Sweden). The finding of a female disadvantage in continuation among girls from North Africa and ex-Yugoslavia in Finland would be in line with this reasoning; however, other findings speak against such an explanation due to the greater female advantage in continuation (e.g. among North Africans in Sweden).

The second major explanation for gender gaps in education refers to returns to education. Increasing returns to education for women in industrialised societies are described as an important explanation for the reversal of the gender gap in these societies over the last few decades (Breen & Goldthorpe 1997; DiPrete & Buchmann 2006). Extending this argument to ethnic minorities, the stark contrast in returns to education for women from more traditional societies should work as a powerful motivator for educational attainment among second-generation girls (Abada & Tenkorang 2009). Indeed, for some groups in some countries, we find that women outperform men from their origin group as well as majority women with regard to the completion of tertiary education. This applies, for instance, to the Asian and Mexican second generation in the USA, Caribbeans in the Netherlands as well as most minority groups in Sweden. However, a less optimistic explanation for these high attainment levels may be lower opportunity costs for higher education due to gender-specific labour market discrimination. Women's motivation to pursue a tertiary degree may be higher if they do not succeed to translate their educational success in secondary education into vocational training and occupational attainment to the same extent as men from their ethnic minority group. So far, however, we have little cross-nationally comparative evidence about gender-specific discrimination of ethnic minorities on the labour market—yet the existing work suggests that minority men experience (p.219) discrimination more often than minority women (Heath et al. 2013). So if gender-specific discrimination was driving the results regarding completion of tertiary education, we should see more of a male advantage in groups that are commonly targeted for discrimination, such as the Bangladeshis, Indians and Pakistanis in England and Wales and Hispanics in the USA. However, this pattern is at odds with the argument that the relatively higher returns to education as compared to migrants' origin country act as a particularly strong motivator for educational attainment among second-generation girls.

Since we could not directly assess the influence of gender role socialisation and returns to education and because some of these expectations can go in opposite directions, we cannot conclude at this point to what extent the various explanations can account for gender gaps in the education of the second generation. Neither could we test whether they affect the second generation in a different way than the majority population. These questions are open for future research that includes information about parental norms and practices as indicators of gender role socialisation as well as measures of educational aspirations to assess the importance of expected returns to education. Moreover, future research may address gender differences in experienced and perceived discrimination at school and in the labour market as a possible explanation for lower opportunity costs among ethnic minority girls or boys to stay on in education. Due to stereotyping of minority boys or to visible markers of minority status, such as a headscarf, male and female members of the second generation might be differentially affected by labour market discrimination.

To conclude, the findings of this chapter show that gender gaps in educational achievement and attainment are by and large of the same size among the second generation as the majority population. This means that girls are getting higher grades and performing better on standardised reading and language tests, that they continue more often in full-time education after compulsory schooling, that they attend academic tracks more often and complete both upper secondary and tertiary education more often than boys, all else being equal. This female advantage in education seems to be particularly marked regarding avoidance of drop out after compulsory schooling. The finding that this pattern of prevalent female advantage is replicated among the second generation implies a substantial change in gender gaps in education among minorities coming from countries where male advantage in education is still the rule. While girls are catching up with boys even in the countries and world regions where they are most disadvantaged, our results indicate that the reversal of the gender gap occurs much faster, namely in the course of one generation, in the context of migration to destination countries where female advantage in education is the rule.

(p.220) References

Bibliography references:

Abada, T. & Tenkorang, E.Y. (2009), ‘Gender Differences in Educational Attainment among the Children of Canadian Immigrants’, International Sociology, 24: 580–608.

Allmendinger, J. (1989), ‘Educational Systems and Labor Market Outcomes’, European Sociological Review, 5: 231–50.

Babad, E. (1993), ‘Pygmalion—25 Years after Interpersonal Expectations in the Classroom’, in P.D. Blanck (ed.), Interpersonal Expectations: Theory, Research and Applications (New York, Cambridge University Press), 125–53.

Baker, D.P. & Jones, D.P. (1993), ‘Creating Gender Equality: Crossnational Gender Stratification and Mathematics Performance’, Sociology of Education, 66: 91–103.

Bedard, K. & Cho, I. (2010), ‘Early Gender Test Score Gaps across OECD Countries’, Economics of Education Review, 29: 348–63.

Birdsall, N., Levine, R. & Ibrahim, A. (2005). Toward Universal Primary Education: Investments, Incentives, and Institutions. UN Millennium Project, Task Force on Education and Gender Equality (London, Earthscan).

Boudon, R. (1974). Education, Opportunity, and Social Inequality (New York, Wiley).

Bradley, K. (2000), ‘The Incorporation of Women into Higher Education: Paradoxical Outcomes?’, Sociology of Education, 73: 1–18.

Brandon, P.R. (1991), ‘Gender Differences in Young Asian Americans' Educational Attainments’, Sex Roles: A Journal of Research, 25: 45–61.

Breen, R. & Buchmann, M. (2002), ‘Institutional Variation and the Position of Young People, A Comparative Perspective’, The Annals of the American Academy of Political and Social Science, 280: 288–305.

Breen, R. and Goldthorpe, J.H. (1997), ‘Explaining Educational Differentials: Towards a Formal Rational Action Theory’, Rationality and Society, 9: 275–305.

Breen, R, Luijkx, R., Müller, W. & Pollak, R. (2010), ‘Long-term Trends in Educational Inequality in Europe: Class Inequalities and Gender Differences’, European Sociological Review, 26: 31–48.

Buchmann, C., DiPrete, T.A. & McDaniel, A. (2008), ‘Gender Inequalities in Education’, Annual Review of Sociology, 34: 319–37.

Cammarota, J. (2004), ‘The Gendered and Racialized Pathways of Latina and Latino Youth: Different Struggles, Different Resistances in the Urban Context’, Anthropology and Education Quarterly, 35: 53–74.

Chiswick, B. & Miller, P.W. (2001), ‘A Model of Destination-language Acquisition: Application to Male Immigrants in Canada’, Demography, 38: 391–409.

Coley, R.J. (2001), Differences in the Gender Gap: Comparisons across Racial/Ethnic Groups in Education and Work (Princeton, NJ, Policy Information Center).

Diehl, C., Friedrich, M. & Hall, A. (2009), ‘Jugendliche ausländischer Herkunft beim Übergang in die Berufsausbildung: Vom Wollen, Können und Dürfen’, Zeitschrift für Soziologie, 38: 48–67.

Dion, K.K. & Dion, K.L. (2001), ‘Gender and Cultural Adaptation in Immigrant Families’, Journal of Social Issues, 57: 511–21.

DiPrete, T.A. & Buchmann, C. (2006), ‘Gender-specific Trends in the Value of Education and the Emerging Gender Gap in College Completion’, Demography, 43: 1–24.

(p.221) Downey, D.B. & Vogt Yuan, A.S. (2005), ‘Sex Differences in School Performance during High School: Puzzling Patterns and Possible Explanations’, The Sociological Quarterly, 46: 299–321.

Else-Quest, N.M., Hyde, J.S. & Linn, M.C. (2010), ‘Crossnational Patterns of Gender Differences in Mathematics: A Meta-analysis’, Psychological Bulletin, 136: 103–27.

Entwisle, D.R., Alexander, K.A. & Olson, L.S. (2007), ‘Early Schooling: The Handicap of being Poor and Male’, Sociology of Education, 80: 114–38.

Esser, H. (2006), Migration, Language, and Integration (Berlin, WZB).

Eurydice (2009), Key Data on Education in Europe 2009 (Brussels, Education, Audiovisual and Culture Executive Agency).

Evertsson, M., England, P., Mooi-Reci, I., Hermsen, J.M., De Bruijn, J. & Cotter, D.A. (2009), ‘Is Gender Inequality Greater at Lower or Higher Educational Levels? Common Patterns in the Netherlands, Sweden, and the United States’, Social Politics: International Studies in Gender, State and Society, 16: 210–41.

Feliciano, C. & Rumbaut, R.G. (2005), ‘Gendered Paths, Educational and Occupational Expectations and Outcomes among Adult Children of Immigrants’, Ethnic and Racial Studies, 28: 1087–118.

Fuligni, A.J., Tseng, V. & Lam, M. (1999), ‘Attitudes toward Family Obligations among American Adolescents with Asian, Latin American, and European Backgrounds’, Child Development, 70: 1030–44.

Gerber, T.P. & Cheung, S.-Y. (2008), ‘Horizontal Stratification in Postsecondary Education: Forms, Explanations, and Implications’, Annual Review of Sociology, 34: 299–318.

Grant, M.J. & Behrman, J.R. (2010), ‘Gender Gaps in Educational Attainment in Less Developed Countries’, Population and Development Review, 36: 71–89.

Harris, M.J. & Rosenthal, R. (1985), ‘Mediation of Interpersonal Expectancy Effects, 31 Meta-analyses’, Psychological Bulletin, 97: 363–86.

Heath, A.F. & Cheung, S.-Y. (2007), Unequal Chances: Ethnic Minorities in Western Labour Markets, Proceedings of the British Academy 137 (Oxford, Oxford University Press).

Heath, A.F., Rothon, C. & Kilpi, E. (2008), ‘The Second Generation in Western Europe: Education, Unemployment and Occupational Attainment’, Annual Review of Sociology, 34: 211–35.

Heath, A.F., Liebig, T. & Simon, P. (2013), ‘Discrimination against Immigrants—Measurement, Incidence and Policy Instruments’, in International Migration Outlook 2013 (Paris, OECD), 191–230.

Kerckhoff, A.C. (2001), ‘Education and Social Stratification Processes in Comparative Perspective’, Sociology of Education, 74: 3–18.

Koopmans, R., Michalowski, I. & Waibel, S. (2012), ‘Citizenship Rights for Immigrants, National Paths and Cross-national Convergence in Western Europe, 1980–2008’, American Journal of Sociology, 117: 1202–45.

Mandel, H. & Semyonov, M. (2006), ‘A Welfare State Paradox: State Interventions and Women's Employment Opportunities in 22 Countries’, American Journal of Sociology, 111: 1910–49.

Marks, G.N. (2007), ‘Accounting for the Gender Gaps in Student Performance in Reading and Mathematics: Evidence from 31 Countries’, Oxford Review of Education, 34: 89–109.

(p.222) Merton, R.K. (1948), ‘The Self-fulfilling Prophecy’, The Antioch Review, 8: 193–210.

Nosek, B.A., Smyth, F.L., Sriram, N., Lindner, N.M., Devos, T., Ayalac, A., Bar-Anan, Y., Bergh, R., Caie, H., Gonsalkorale, K., Kesebir, S., Maliszewski, N., Neto, F., Olli, E., Park, J., Schnabel, K., Shiomura, K., Tudor Tulbure, B., Wiers, R.W., Somogyi, M., Akrami, N., Ekehammar, B., Vianello, M., Banaji, M.R. & Greenwald, A.G. (2009), ‘National Differences in Gender—Science Stereotypes Predict National Sex Differences in Science and Math Achievement’, Proceedings of the National Academy of Sciences of the United States of America, 106: 10, 593–97.

OECD (2006), Where Immigrants Succeed: A Comparative Review of Performance and Engagement in PISA 2003 (Paris, OECD).

OECD (2009a), Equally Prepared for Life? How 15-year-old Boys and Girls Perform in School (Paris, OECD).

OECD (2009b), Education at a Glance 2009 (Paris, OECD).

Raijman, R. & Semyonov, H. (1997), ‘Gender, Ethnicity, and Immigration: Double Disadvantage and Triple Disadvantage among Recent Immigrant Women in the Israeli Labor Market’, Gender and Society, 11: 108–25.

Riphahn, R.T. & Serfling, O. (2002), ‘Neue Evidenz zum Schulerfolg von Zuwanderern der zweiten Generation in Deutschland’, Vierteljahreshefte zur Wirtschaftsforschung, 71: 230–48.

Rosenthal, R. & Jacobson, L. (1966), ‘Teachers' Expectancies: Determinants of Pupils' IQ Gains’, Psychological Reports, 19: 115–18.

Schofield, J.W. (2006), Migration Background, Minority-group Membership and Academic Achievement: Research Evidence from Social, Educational, and Developmental Psychology (Berlin, WZB).

Segeritz, M., Walter, O. & Stanat, P. (2010), ‘Muster des schulischen Erfolgs von jugendlichen Migranten in Deutschland: Evidenz für segmentierte Assimilation?’, Kölner Zeitschrift für Soziologie und Sozialpsychologie, 62: 113–38.

Shelton, B.A. & John, D. (1996), ‘The Division of Household Labor’, Annual Review of Sociology, 22: 299–322.

Steele, C.M. (1997), ‘A Threat in the Air: How Stereotypes Shape Intellectual Identity and Performance’, American Psychologist, 52: 613–29.

Tournaki, N. (2003), ‘Effect of Student Characteristics on Teachers' Predictions of Student Success’, The Journal of Educational Research, 96: 310–19.

UNDP (2010), Indicators, Human Development Report 2009 UN Development Programme, http://hdrstats.undp.org/en/indicators/.

Van der Lippe, T. & Van Dijk, L. (2002), ‘Comparative Research on Women's Employment’, Annual Review of Sociology, 28: 221–41.

Van Langen, A., Bosker, R. & Dekkers, H. (2006), ‘Exploring Cross-national Differences in Gender Gaps in Education’, Educational Research and Evaluation, 12: 155–77.

Van Tubergen, F., Maas, I. & Flap, H. (2004), ‘The Economic Incorporation of Immigrants in 18 Western Societies: Origin, Destination and Community Effects’, American Sociological Review, 69: 704–29.

Wils, A. & Goujon, A. (1998), ‘Diffusion of Education in Six World Regions, 1960–90’, Population and Development Review, 24: 357–68.

Zhou, M. & Bankston, C.L. (2002), ‘Family Pressure and the Educational Experience of the Daughters of Vietnamese Refugees’, International Migration, 39: 133–51.

Table A8.1. Gender differences in test scores or grades at age 15–16 (OLS regression).

Belgium

Finland

France

Germany

Netherlands

Sweden

Switzerland

England and Wales

USA

Competence measure

Test scores (reading)

Grades sum score

Test scores (French)

Test scores (reading)

Test scores (CITO)

Grades sum score

Test scores (reading)

Grades (GCSE)

Test scores (PVT)

N

8162

2242

9909

30,822

18,167

581,368

11,815

16,134

11,375

Majority

Female advantage

0.36

0.63

0.54

0.26

0.03

0.37

0.29

0.19

0.14

East Asian

Chinese

Female advantage

[0.19]

[0.04]

Penalty/premium

0.65

[0.15]

East Asian

Female advantage

0.30

[0.39]

Penalty/premium

0.60

0.57

Southeast/South Asian

Bangladeshi

Female advantage

[0.15]

Penalty/premium

0.41

Indian

Female advantage

[0.28]

Penalty/premium

0.24

Pakistani

Female advantage

[0.15]

Penalty/premium

[0.13]

Southeast Asian

Female advantage

[0.34]

Penalty/premium

0.47

West Asian

Iranian

Female advantage

[0.33]

Penalty/premium

0.29

Iraqi

Female advantage

[0.29]

Penalty/premium

[0.00]

Turkish

Female advantage

[0.18]

[0.15]

[-0.04]

[0.39]

[0.27]

Penalty/premium

−0.67

−0.52

−0.17

−0.07

−0.62

West Asian

Middle East

Female advantage

0.26

Penalty/premium

[0.02]

Other Asian

Other Asian

Female advantage

[0.42]

[-0.03]

Penalty/premium

0.31

[-0.18]

East European

Albanian/Kosovar

[0.12]

Female advantage

[0.12]

Penalty/premium

−1.42

East European

Female advantage

[0.37]

Penalty/premium

0.08

Ex-Yugoslav

Female advantage

0.12

[0.66]

[0.38]

[0.13]

Penalty/premium

−0.30

[-1.20]

[0.04]

−0.84

Former Soviet Union

Female advantage

[0.21]

Penalty/premium

[-0.30]

Polish

Female advantage

[0.34]

0.30

Penalty/premium

[-0.20]

0.24

Russian/Estonian

Female advantage

[0.53]

Penalty/premium

[-0.03]

South European

Iberian

Female advantage

−0.15

Penalty/premium

−0.27

Italian

Female advantage

[0.57]

[0.02]

Penalty/premium

−0.93

−0.41

Greek

Female advantage

[0.50]

Penalty/premium

[-0.48]

South European

European

Portuguese

Female advantage

[0.46]

[0.29]

[0.42]

Penalty/premium

[-0.13]

[0.13]

[0.05]

West European

Danish

Female advantage

0.28

Penalty/premium

−0.11

Finnish

Female advantage

[0.39]

Penalty/premium

−0.10

Norwegian

Female advantage

0.28

Penalty/premium

[-0.01]

West European

Swedish-sp.

white

Female advantage

0.74

[0.47]

Penalty/premium

−0.12

[-0.41]

African

From Africa's Horn

Female advantage

[0.30]

Penalty/premium

0.19

North African

Moroccan

Maghrebian

Moroccan

Female advantage

[0.33]

0.16

[0.60]

[0.14]

0.53

Penalty/premium

−0.52

[-0.15]

−0.15

−0.40

[-0.02]

Sub-Saharan African

black

Female advantage

0.08

[0.63]

[0.37]

[0.47]

[0.27]

Penalty/premium

[0.08]

[-0.32]

[0.00]

[-0.16]

−0.38

Caribbean

Caribbean

Female advantage

[0.18]

Penalty/premium

−0.22

Dom-Tom

Female advantage

[0.64]

Penalty/premium

−0.14

Surinamese/Antillean

Female advantage

[0.13]

Penalty/premium

−0.21

South American

Chilean

Female advantage

0.25

Penalty/premium

−0.14

Mexican

Female advantage

[0.01]

Penalty/premium

−0.48

South American

other Latino

Female advantage

[0.38]

[0.30]

Penalty/premium

[0.02]

−0.48

Notes. Cell values represent OLS regression coefficients of the effect of being female (majority) and the sum of the coefficients of being female and the interaction between being female and belonging to a particular second-generation group (minorities). For the second-generation groups, the second row refers to the ethnic coefficient. Non-significant main effects (for ethnic origin) and interactions (between gender and ethnic origin) are indicated by square brackets (p < 0.05).

(p.224) (p.225) (p.226) (p.227) (p.228)

Table A8.2. Gender differences in continuation (probit regression).

Finland

France

Germany

Netherlands

Sweden

England and Wales

USA

N

23,158

11,869

16,872

15,861

565,923

16,219

10,950

Majority

Female advantage

[-0.05]

0.23

0.07

0.10

0.16

0.10

0.23

East Asian

Chinese

Female advantage

[-0.11]

[0.44]

Penalty/premium

0.90

[-0.05]

East Asian

Female advantage

[-0.06]

[0.22]

Penalty/premium

[-0.03]

0.35

Southeast/South Asian

Bangladeshi

Female advantage

[-0.12]

Penalty/premium

0.82

Indian

Female advantage

[0.06]

Penalty/premium

0.71

Pakistani

Female advantage

[0.01]

Penalty/premium

0.38

Southeast Asian

Female advantage

[0.21]

Penalty/premium

0.37

West Asian

Iranian

Female advantage

[0.21]

Penalty/premium

0.30

Iraqi

Female advantage

[0.22]

Penalty/premium

[-0.05]

Turkish

Female advantage

[0.10]

[0.10]

[0.19]

Penalty/premium

−0.15

[-0.17]

−0.26

West Asian

Middle East

Female advantage

0.07

Penalty/premium

−0.07

Other Asian

Other Asian

Female advantage

[0.20]

[0.67]

Penalty/premium

0.21

[0.17]

East European

East European

Female advantage

[0.15]

Penalty/premium

[-0.03]

Ex-Yugoslav

Female advantage

[-0.56]

[0.16]

[0.17]

Penalty/premium

[-0.07]

[-0.06]

[0.00]

Former Soviet Union

Female advantage

[0.09]

Penalty/premium

[-0.29]

Polish

Female advantage

0.59

[0.19]

Penalty/premium

[-0.05]

0.08

Russian/Estonian

Female advantage

[0.29]

Penalty/premium

[-0.27]

South European

Italian

Female advantage

[0.18]

Penalty/premium

[0.02]

Greek

Female advantage

[-0.26]

Penalty/premium

[0.31]

South European

European

Portuguese

Female advantage

[0.35]

[0.32]

[0.15]

Penalty/premium

−0.61

[0.18]

[-0.05]

West European

Danish

Female advantage

[0.11]

Penalty/premium

−0.17

Finnish

Female advantage

[0.16]

Penalty/premium

−0.12

Norwegian

Female advantage

[0.11]

Penalty/premium

[0.02]

West European

Swedish-sp.

white

Female advantage

[-0.05]

[0.11]

Penalty/premium

[-0.01]

[-0.24]

African

From Africa's Horn

Female advantage

[0.22]

Penalty/premium

0.18

N African

Moroccan

Female advantage

[-0.38]

[0.39]

[0.24]

0.44

Penalty/premium

[-0.28]

0.31

−0.32

−0.22

Sub-Saharan African

black

Female advantage

[0.02]

1.20

[0.11]

[0.94]

−2.83

Penalty/premium

−0.42

[-0.29]

[-0.02]

0.49

3.86

Caribbean

Caribbean

Female advantage

−0.48

Penalty/premium

0.55

Dom-Tom

Female advantage

0.55

Penalty/premium

[-0.13]

Surinamese/Antillean

Female advantage

[0.19]

Penalty/premium

−0.38

South American

Chilean

Female advantage

[0.14]

Penalty/premium

−0.15

Mexican

Female advantage

[0.31]

Penalty/premium

[0.06]

South American

other Latino

Female advantage

[0.09]

[0.35]

Penalty/premium

[0.06]

[-0.08]

(p.229) (p.230) (p.231) (p.232)

Table A8.3. Gender differences in track followed in upper secondary education (probit regression).

Belgium

Finland

France

Germany

Netherlands

Sweden

England and Wales

N

22,184

21,596

11,253

11,923

12,688

511,103

13,662

Majority

Female advantage

0.22

0.55

0.47

0.20

0.10

0.31

0.08

East Asian

Chinese

Female advantage

[0.18]

Penalty/premium

0.81

East Asian

Female advantage

[0.35]

[0.40]

Penalty/premium

0.80

0.91

Southeast/South Asian

Bangladeshi

Female advantage

0.71

Penalty/premium

[0.03]

Indian

Female advantage

0.39

Penalty/premium

0.43

Pakistani

Female advantage

[0.10]

Penalty/premium

0.31

Southeast Asian

Female advantage

[0.27]

Penalty/premium

0.69

West Asian

Iranian

Female advantage

0.45

Penalty/premium

0.90

Iraqi

Female advantage

[0.39]

Penalty/premium

0.67

Turkish

Female advantage

−0.25

[0.37]

[0.00]

[0.39]

Penalty/premium

[0.04]

[0.12]

[0.11]

0.68

West Asian

Middle East

Female advantage

[0.35]

Penalty/premium

0.55

Other Asian

Other Asian

Female advantage

0.62

Penalty/premium

0.82

East European

East European

Female advantage

[0.33]

Penalty/premium

0.45

Ex-Yugoslav

Female advantage

[0.01]

[-0.27]

0.46

Penalty/premium

−0.47

[0.29]

0.39

Former Soviet Union

Female advantage

[0.45]

Penalty/premium

[-0.26]

Polish

Female advantage

[0.25]

[0.25]

Penalty/premium

[0.16]

0.56

Russian/Estonian

Female advantage

[0.41]

Penalty/premium

0.30

South European

Italian

Female advantage

[0.08]

[-0.44]

Penalty/premium

[0.20]

[0.04]

Greek

Female advantage

[0.36]

Penalty/premium

[0.16]

South European

European

Portuguese

Female advantage

[0.76]

[0.29]

0.52

Penalty/premium

[0.43]

0.36

0.42

West European

Danish

Female advantage

[0.19]

Penalty/premium

[0.01]

Finnish

Female advantage

[0.36]

Penalty/premium

[0.03]

Norwegian

Female advantage

0.17

Penalty/premium

[0.08]

West European

Swedish-sp.

Female advantage

[0.63]

Penalty/premium

−0.13

African

From Africa's Horn

Female advantage

[0.43]

Penalty/premium

1.01

North African

Moroccan

Moroccan

Female advantage

[0.05]

[0.27]

[0.68]

[0.31]

[0.39]

Penalty/premium

[0.14]

0.45

0.27

[-0.11]

0.78

Sub-Saharan African

Female advantage

[0.22]

[0.23]

[0.29]

[0.22]

Penalty/premium

1.11

[0.36]

0.60

[0.26]

Caribbean

Caribbean

Female advantage

[0.22]

Penally/premium

[0.02|

Dom-Tom

Female advantage

[0.40]

Penalty/premium

[0.12]

Surinamese/Antillean

Female advantage

[-0.14]

Penalty/premium

0.41

South American

Chilean

Female advantage

[0.26]

Penalty/premium

0.39

South American

Female advantage

[0.19]

Penalty/premium

0.57

Notes. Cell values represent probit regression coefficients of the effect of being female (majority) and the sum of the coefficients of being female and the interaction between being female and belonging to a particular second-generation group (minorities). For the second-generation groups, the second row refers to the ethnic coefficient. Non-significant main effects (for ethnic origin) and interactions (between gender and ethnic origin) are indicated by square brackets (p < 0.05).

(p.233) (p.234) (p.235) (p.236)

Table A8.4. Gender differences in completion of upper secondary education (probit regression).

Belgium

France

Netherlands

Sweden

England and Wales

USA

N

115,045

11,869

2037

272,520

7648

10,950

Majority

Female advantage

0.33

0.34

−0.32

0.21

0.22

0.19

East Asian

Chinese

Female advantage

[0.23]

−3.36

Penalty/premium

[0.50]

3.71

East Asian

Female advantage

0.60

Penalty/premium

[0.19]

Southeast/South Asian

Bangladeshi

Female advantage

[0.34]

Penalty/premium

[0.44]

Indian

Female advantage

[0.15]

Penalty/premium

0.40

Pakistani

Female advantage

[0.21]

Penalty/premium

[0.11]

Southeast Asian

Female advantage

[0.19]

Penalty/premium

0.15

West Asian

Iranian

Female advantage

[0.28]

Penalty/premium

[0.01]

Iraqi

Female advantage

0.48

Penalty/premium

−0.32

Turkish

Female advantage

[0.30]

0.03

0.48

Penalty/premium

−0.34

−0.61

−0.41

West Asian

Middle East

Female advantage

0.37

Penalty/premium

−0.32

Other Asian

Other Asian

Female advantage

[0.37]

0.74

Penalty/premium

[-0.10]

[-0.24]

East European

East European

Female advantage

[0.30]

Penalty/premium

[-0.08]

Ex-Yugoslav

Female advantage

0.38

Penalty/premium

−0.11

Polish

Female advantage

[0.23]

Penalty/premium

[0.06]

South European

Italian

Female advantage

[0.35]

Penalty/premium

−0.18

South European

Portuguese

Female advantage

[0.68]

[0.24]

Penalty/premium

[0.04]

−0.15

West European

Danish

Female advantage

[0.23]

Penalty/premium

−0.22

Finnish

Female advantage

[0.25]

Penalty/premium

−0.17

Norwegian

Female advantage

[0.27]

Penalty/premium

−0.22

West European

white

Female advantage

[-0.27]

Penalty/premium

[-0.09]

African

From Africa's Horn

Female advantage

0.60

Penalty/premium

[-0.15]

North African

Moroccan

Moroccan

Female advantage

[0.27]

0.83

[-0.07]

0.51

Penalty/premium

−0.25

[0.09]

[-0.29]

−0.45

Sub-Saharan African

black

Female advantage

[0.81]

0.52

[0.90]

−3.75

Penalty/premium

[-0.28]

−0.46

[0.68]

3.80

Caribbean

Caribbean

Female advantage

[0.10]

Penalty/premium

[-0.01]

Dom-Tom

Female advantage

0.62

Penalty/premium

−0.14

Surinamese/Antillean

Female advantage

0.35

Penalty/premium

−0.38

South American

Chilean

Female advantage

[0.29]

Penalty/premium

−0.37

Mexican

Female advantage

[0.41]

Penalty/premium

[-0.17]

South American

other Latino

Female advantage

[0.27]

[0.43]

Penalty/premium

0.24

[-0.32]

Notes: Cell values represent prohibit regression coefficients of the effect of being female (majority) and the sum of the coefficients of being female and the interaction between being female and belonging to a particular second-generation group (minorities). For the second-generation groups, the second row refers to the ethnic coefficient. Non-significant main effects (for ethnic origin) and interactions (between gender and ethnic origin) are indicated by square brackets (p < 0.05).

(p.237) (p.238) (p.239) (p.240)

Table A8.5. Gender differences in completion of tertiary education (probit regression).

Belgium

Netherlands

Sweden

England and Wales

USA

N

39,527

2037

272,520

23,007

10,094

Majority

Female advantage

0.39

−0.15

0.31

[0.07]

0.20

East Asian

Chinese

Female advantage

[-0.33]

Penalty/premium

1.21

East Asian

Female advantage

0.74

Penalty/premium

0.80

Southeast/South Asian

Bangladeshi

Female advantage

−0.56

Penalty/premium

[0.12]

Indian

Female advantage

[-0.17]

Penalty/premium

0.81

Pakistani

Female advantage

[-0.09]

Penalty/premium

[0.19]

Southeast Asian

Female advantage

0.10

Penalty/premium

0.80

West Asian

Iranian

Female advantage

0.16

Penalty/premium

0.74

Iraqi

Female advantage

[0.41]

Penalty/premium

0.37

Turkish

Female advantage

[0.18]

[-0.08]

[0.41]

Penalty/premium

−0.32

−0.38

0.34

West Asian

Female advantage

[0.41]

Penalty/premium

0.27

Other Asian

Other Asian

Female advantage

[0.18]

[0.44]

Penalty/premium

0.79

0.47

East European

East European

Female advantage

[0.28]

Penalty/premium

0.33

Ex-Yugoslav

Female advantage

[0.39]

Penalty/premium

0.25

Polish

Female advantage

0.18

Penalty/premium

0.56

South European

Italian

Female advantage

[0.38]

Penalty/premium

[-0.05]

South European

Female advantage

[0.30]

Penalty/premium

0.31

West European

Danish

Female advantage

[0.28]

Penalty/premium

[-0.10]

Finnish

Female advantage

[0.34]

Penalty/premium

[-0.01]

Norwegian

Female advantage

[0.17]

Penalty/premium

[0.10]

African

From Africa's Horn

Female advantage

[0.44]

Penalty/premium

0.43

North African

Moroccan

Moroccan

Female advantage

[0.23]

[-0.23]

[0.58]

Penalty/premium

−0.18

[-0.29]

0.28

Sub-Saharan African

Female advantage

[0.35]

[0.60]

Penalty/premium

[0.13]

[0.05]

Caribbean

Caribbean

Female advantage

[0.39]

Penalty/premium

[0.16]

Puerto Rican

Female advantage

[0.46]

Penalty/premium

[-0.55]

Surinamese/Antillean

Female advantage

[0.00]

Penalty/premium

[-0.15]

South American

Chilean

Female advantage

[0.23]

Penalty/premium

0.19

Mexican

Female advantage

[0.51]

Penalty/premium

[-0.10]

South American

Hispanic

Female advantage

[0.34]

[-0.30]

Penalty/premium

0.27

0.72

Notes: Cell values represent prohibit regression coefficients of the effect of being female (majority) and the sum of the coefficients of being female and the interaction between being female and belonging to a particular second-generation group (minorities). For the second-generation groups, the second row refers to the ethnic coefficient. Non-significant main effects (for ethnic origin) and interactions (between gender and ethnic origin) are indicated by square brackets (p < 0.05).

(p.241) (p.242) (p.243) (p.244)

Notes:

Proceedings of the British Academy, 196, 193–244. © The British Academy 2014

(1) Recent OECD data reveal consistent female advantages in the completion of upper secondary education, also in the Netherlands (OECD 2010). However, the female advantage in the Netherlands is limited to the youngest age cohort (25–34 year olds in 2010). This recent change may not yet show up among the relatively older cohorts included in this analysis.

(2) A comparison of these findings with recent data from the OECD (2009b) confirms the female advantage in completion of tertiary education in Belgium, Sweden and the USA and the male advantage in the Netherlands. For England and Wales, the OECD data indicate modest female advantage in the completion of tertiary education.

(3) Results from Finland, where continuation can be analysed contingent on prior achievement, are in line with this reasoning: the non-significant main effect of female gender (b = −0.05, S.E. = 0.04) shifts to a significant female disadvantage when grades are controlled (b = −0.38, S.E. = 0.04). Importantly, including prior achievement does not affect the magnitude and significance of the interactions between gender and ethnic background.