Explaining the gap in educational achievement between second-generation immigrants and natives: The Italian case

(1)

This article was downloaded by: [EUI European University Institute] On: 17 February 2013, At: 13:05

Publisher: Routledge

Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Modern Italian Studies

Publication details, including instructions for authors and subscription information:

http://www.tandfonline.com/loi/rmis20

Explaining the gap in

educational achievement

between second-generation

immigrants and natives: the

Italian case

Anna Di Bartolomeo a

a

European University Institute

Version of record first published: 01 Sep 2011.

To cite this article: Anna Di Bartolomeo (2011): Explaining the gap in educational achievement between second-generation immigrants and natives: the Italian case, Journal of Modern Italian Studies, 16:4, 437-449

To link to this article: http://dx.doi.org/10.1080/1354571X.2011.593749

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages

(2)

(3)

Explaining the gap in educational achievement

between second-generation immigrants and natives:

the Italian case

Anna Di Bartolomeo

European University Institute

Abstract

This article studies the determinants of the achievement gap between second-generation immigrants and natives in Italy, investigating the role played by economic resources, cultural-capital background, pupils’ aspirations and ethnic school-segregation dynamics – the main factors emphasized in the literature. In addition, we verify to what extent second-generation status exerts an independent effect on early school performance, net of the abovementioned determinants. The Italian case is compared with those of France and Germany. As to the main results, we find that, in Italy, second-generation status per se is also a determinant; ethnic school-segregation dynamics and pupils’ socio-economic background represent the most critical barriers. Similarly in Germany, the traditional determinants are not the only constraints faced by second-generation immigrants in achieving at least the same educational performance level as their native peers, suggesting that further analysis is needed. By contrast, in France, second-generation status has no residual effect on educational performance.

Keywords

Educational achievement, second-generation immigrants, Italy, international comparison.

Introduction

The integration of migrants from all parts of the world into the host society has become one of the greatest challenges for highly developed countries. An imp-ortant role in the integration process is played by the educational system, as human capital accumulation is considered a fundamental precondition for the integration of migrants by representing the point of departure for labour market success.

Education in fact has long been considered a way of social advancement for immigrant families. For the majority of migrants who come to a new country with low skills, without an established family business, accumulated wealth or long-standing local social networks, the educational system represents a unique opportunity for social mobility with respect to the next generation. Success in

Journal of Modern Italian Studies 16(4) 2011: 437–449

Journal of Modern Italian Studies

ISSN 1354-571X print/ISSN 1469-9583 onlineª2011 Taylor & Francis http://www.informaworld.com DOI: 10.1080/1354571X.2011.593749

(4)

the education system would allow their children to obtain higher-paying and higher-status jobs with a contemporaneous rise in the family’s social standing. The literature uses educational attainment to capture progression in human capital accumulation up through national education systems. However, by comparing students’ performances it emerges that, notwithstanding the strong incentives, in industrialized countries immigrants’ children do not perform as well as other students (see among others: Brinbaum and Heath 2007; Marks 2005). This achievement gap and its policy implications are shown to be particularly relevant when considering the trajectories of second-generation immigrants who were born, socialized and educated in the host country and who are thus supposed to shareceteris paribusthe same opportunities as natives. The aim of this paper is to detect the role played by economic resources, cultural-capital background, pupils’ aspirations and school-segregation dy-namics – the main factors stressed by literature in determining the achievement gap between second-generation immigrants and natives in Italy. In particular, we want to verify whether there is a residual effect of the generation status, net of the mentioned determinants, that is to what extent second-generation status exerts an independent effect on early school performance.

As the integration process, as well as the related problems and policies, may be strongly affected by its nature – for example the problems of early immigration countries are rather different from those arising in countries with a consolidated story of immigration – we want to verify also whether different stages of immigration are likely to play any role in the educational gap, by comparing the cases of three European countries. We analyse the Italian case and compare it with those of France and Germany, which broadly correspond to three different stages of immigration of European economies: (a) Italy, a case typical of new countries of immigration; (b) France, a country with a long immigration history linked to its colonial past; and (c) Germany, which experienced large inflows of immigrants under its politics of post-war labour recruitment. In this respect, our study is particularly relevant for policy makers as our cases, together, fully represent the development of the immigration process and integration in Europe.

The rest of the paper is structured as follows. The next section focuses on the main literature on the determinants of the gap in educational achievement between immigrants’ children and their native peers; the paper then deals with the research objectives as well as data with the adopted data and methodology; followed by a section focusing on our main results; the final section presents conclusions and further directions of the research.

Explaining the gap in educational achievement: literature review

In order to assess educational attainment, two kinds of measures are normally adopted: (a) test scores, which deal directly with scholastic performances, and (b) educational outcomes, representing the continuation rates into upper

(5)

secondary and tertiary education when compulsory schooling is concluded and students can choose whether to continue or to enter the labour market. The determinants of the educational gap by origin may differ or play different roles depending on whether educational attainment is measured by test scores or continuation rates. In this paper, educational achievement is measured in terms of test scores and its peculiar determinants are presented; continuation rates are also discussed.

To explain the gap in educational achievement, sociologists of education have tended to focus on class inequalities by developing two broad families of explanations: the ‘structural’ and the ‘cultural’ ones respectively. Their common feature relies on Bourdieau and Passeron’s cultural and social reproduction theory (1970, 2000), which argues that the unequal distribution of power resources (social, cultural and economic capital) between classes is transmitted over generations.

‘Structural explanations’ stress that the weaker performance of immigrants’ children is largely due to socio-economic factors, usually indexed by parental occupation. While these factors are directly connected with continuation rates, since inequalities in material resources make continuing in education beyond the period of compulsory school more costly for children from working-class origin (Brinbaum and Heath 2007), it seems reasonable also to link parents’ economic conditions and the lack of material resources to test score performance, in terms of possibility to access additional resources for the acquisition of knowledge (e.g. extra scholastic courses, private lessons, cultural possessions, etc.).

‘Cultural explanations’ instead consider cultural capital as the most important form of capital for children at school. In its most general form, cultural capital consists of familiarity with the dominant culture in society, which corresponds to the culture found at school. Like many forms of capital, cultural capital is inherited by children from their parents. Thus, according to this theory, the weaker performance of immigrants’ children can be related to their weaker cultural-capital background. It can be indexed by several proxies, that is the educational level of parents together with their familiarity with the hosting culture, parental skills in helping children with their schoolwork (and thus the importance of language skills) and knowledge on how to surf with the educational system (Van de Werfhorst and Hofstede 2007). The negative effect of this cultural dissonance is also known as the primary effect of stratification on educational attainment measured by test scores.1

Also, other explanations focus on school characteristics and thus on the importance of the context. In most countries, ethnic minorities are often concentrated in particular areas, typically economically disadvantaged ones, so that in local schools there are high proportions of immigrants’ children. As a matter of fact, these deprived neighbourhoods are often associated with poorer schooling, characterized by difficulties in attracting and retaining suitable teachers (higher teacher turnover) or by parents who have less time and resources to contribute to the school. Many scholars (Portes and MacLeod

Explaining the gap in educational achievement

439

(6)

1996; Rumberger and Willms 1992; Wang and Goldschmidt 1999) found evidence of the strong association between school segregation and the weaker performance of ethnic minorities.

Finally, the decision-making process may also play a role in affecting scholastic performance. According to the relative risk aversion theory (Breen and Goldthorpe 1997; Goldthorpe 2010), the primary goal for each member of the society is to avoid downward social mobility with respect to their parents’ trajectories. Aspirations may thus play an important role in determining scholastic performance and enhance intergenerational social mobility.2

Research objectives, data and methods

As shown in the previous section, scholars have highlighted several factors in explaining the gap in educational achievement between immigrants’ children and their native peers. This paper addresses the following key question: to what extent does generation status exert an independent effect on early school performance, assuming equal economic resources, cultural-capital background, pupils’ aspirations and ethnic school segregation?

Furthermore, we want to verify whether different stages of immigration play any role in the educational performance by comparing the case of Italy with the French and German ones.

Data had been taken from the Organisation for Economic Cooperation and Development (OECD)’s 2006 Program for International Students Assessment (PISA), a triennial survey of the educational attainment of 15-year-old pupils, which consists of a sample of more than 400,000 pupils from fifty-seven countries. The focus of PISA 2006 was science.

Our sample is composed of 30,095 observations (69.8% in Italy; 15.2% in Germany; 15.0% in France). For our purposes, we divided immigrants’ children into: first-generation pupils, born outside the receiving country and whose parents were also born in a different country; and second-generation pupils, born in the receiving country but whose parents were born in a different country. In this second subgroup we also include pupils who were born outside the recei-ving country and who emigrated at preschool age, the so-called ‘1.5 generation’. The inclusion of these pupils is justified by the fact that both the second and the 1.5 generation of pupils share the unique position of having experienced the whole period of ‘secondary socialization’ in the receiving country.3

As mentioned before, comparing the performance differences between first- and second-generation pupils with those of natives may give some insight into the effectiveness of countries’ school systems in developing immigrants’ children’s scholastic skills. First-generation pupils have typically had only part of their schooling in the receiving country and may have had very different schooling experiences before they arrived there. Thus the level of achievement they have reached at age fifteen can be only partly attributed to the school system of the receiving country. Their relative performance

(7)

may serve as a rough baseline for the potential that immigrants’ children bring with them when they enter the different receiving countries. By contrast, the achievement of second-generation pupils is largely determined by the receiving country’s school system, and thus it ought to be closer to the achievement of native children. The gap in performance between first- and second-generation pupils may thus indicate the extent to which the dif-ferent school systems succeed in supporting immigrant pupils’ learning (OECD 2006).

To measure scholastic achievement, we used the PISA classification of test scores which divides science education attainment into six levels. Level six denotes a ‘very strong engagement in explaining and applying related knowledge and ability to conceptualize and generalize’; level five corresponds to a ‘very strong engagement in explaining and applying related knowledge’; pupils that are in the fourth level are characterized by a ‘strong engagement in applying related knowledge’; in level three pupils are ‘able in describing issues related to the specific knowledge’; in level two ‘only literal interpretations are performed’; and, last, level one denotes only a ‘limited knowledge of the subject restricted to such familiar contexts’.

An evident gap in educational achievement can clearly be seen between generations (Figure 1). While natives perform better in each country, higher proportions of immigrants’ children are observed in the ‘low performance’ levels. A decreasing performance trend is observed according to generation status, in which natives perform better than the second generation and the latter better than the first. An exception is observed in the German context, in which first- and second-generation immigrants show similar distributions and a relatively high number of first-generation pupils are observed in the sixth level. Since a descriptive analysis may be misleading, we ran several nested logistic regressions in order to detect the probability of achieving a high school performance. Here, a high school performance is denoted by having achieved a proficiency level from 3 to 6, while a bad school performance coincides with levels 2, 1 and below 1. Our response variable thus equals 1 for ‘high school performance’ and 0 otherwise.

Figure 1Pupils by proficiency level and generation status: France, Germany and Italy, 2006 (source: author calculation using PISA database).

441

(8)

To understand where this gap derives from, and detect the role of generation status once controlled for traditional determinants, we insert the following covariates in the models, which had been performed separately for each country:

. In the baseline model, the key covariate ‘generation status’ is inserted. . Further, we add pupils’ socio-economic background indexed by the highest

job status of parents. Can immigrants’ children’s socio-economic back-ground alone explain the educational gap that immigrants face? This question is of great relevance: if immigrants’ achievement differs from that of natives only because of their poorer economic backgrounds, educational policies would not need to address immigrants’ special needs. Otherwise, migrants’ disadvantages could then be reduced by policies providing additional support to all children with disadvantaged family backgrounds. . The impact of cultural-capital background is inserted in the third model

aimed at observing whether the economic and cultural explanations alone are able to explain the gap. It is indexed by (a) language spoken at home; (b) highest educational level of parents, considered as a measure of ‘institutionalized’ cultural capital and measured in accordance with the International Standard Classification of Education (ISCED); and (c) index of cultural possessions (e.g. personal computer, scientific books) as a proxy of the possibility to access cultural resources.

. We also control for pupils’ aspirations to test the risk-aversion theory by building an index composed of the following measures: (a) interest in scientific topics; (b) future-oriented motivation to learn science; and (c) science-related activities.

. The last covariate is represented by the proportion of immigrants’ children in each school to verify the mechanisms of ethnic school segregation.

3. Main results

The analysis confirms the importance of the traditional determinants in explaining the gap in educational achievement between immigrants’ children and natives. However, substantial differences are found among countries.

In Italy, even after taking account of pupils’ socio-economic and cultural backgrounds, pupils’ aspirations and ethnic school segregation dynamics, the educational dissonance of both the first and the second generation persists (Table 1). Socio-economic determinants, but especially ethnic school segregation dynamics, seem to act as the main constraints for a good scholastic performance of immigrants’ children. Actually, other studies have already found socio-economic background and differential access to opportunities to be the most relevant drivers in the social mobility process in Italy (Sabbadini 2008.) In this respect, the educational gap according to parents’ origin adds a new element of dissonance, which triggers the peculiar vulnerability of

(9)

second-generation status. Furthermore, given that the aspirations of immi-grants’ children tend to be higher with respect to both their parents’ and peers’, frustrations are also expected to be larger, ultimately leading to a sense of social inadequacy for them. Additional constraints are promptly linked to ethnic school segregation dynamics, too. In this respect, the capacity of Italian schools to integrate instead of reproducing social inequalities is again called into question. Surprisingly, the language spoken at home does not show any significant effect on pupils’ performance. However, it is worth mentioning that Table 1Logistic regression results: odds ratio of achieving high performance in test scores, Italy{

Italy Model 1 Model 2 Model 3 Model 4 Model 5

Generation status

Natives (ref.) 1 1 1 1 1

Second generation 0.47*** 0.55*** 0.62* 0.62** 0.70*

First generation 0.31*** 0.39*** 0.47*** 0.45*** 0.53***

þSocio-economic status:

Highest parental job status White collar high-skilled

(ref.)

1 1 1 1

White collar low-skilled 0.68*** 0.80*** 0.80*** 0.80***

Blue collar high-skilled 0.36*** 0.54*** 0.54*** 0.54***

Blue collar low-skilled 0.37*** 0.55*** 0.55*** 0.55***

þCultural capital background

Language spoken at home Language of receiving

country (ref.)

1 1 1

Other languages NS NS NS

Highest parental education level

ISCED 5 or 6 (ref.) 1 1 1 ISCED 3 or 4 NS NS NS Below ISCED 3 0.64*** 0.66*** 0.66*** Cultural possessions 410 (ref.) 1 1 1 7–10 0.49*** 0.51*** 0.51*** 57 0.21*** 0.23*** 0.24* þRisk-aversion theory

Index of pupils’ aspirations

High aspirations (ref.) 1 1

Low aspirations 0.67*** 0.68***

þEthnic school segregation

% of immigrants’ children per school

the mean value (ref.) 1

4the mean value 0.49***

{

Estimates are controlled for sex. ***p50.001; **p50.01; *p50.05.

443

(10)

the questionnaire did not provide the possibility to give a double answer to the question we used as a proxy ‘language ability’: ‘Generally, what language do you speak at home?’ So, we were not able to capture those pupils who use both Italian and the language of origin, and this would probably have led ultimately to a bias in the estimate. As argued by many scholars, in fact, a bilingual environment may bring substantial benefits in educational achievement if compared with a monolingual one (see among others Bialystok 2001).

In Germany, similar to the Italian case, the gap in educational achievement between second-generation immigrants and natives does not disappear once controlled for the traditional determinants. Here, cultural-capital background plays the most relevant role (Table 2). In this respect, it is worth mentioning that, as in certain German federal states, parents are institutionally entitled to select from among primary schools. Several studies (see, among others, Kristen and Granato 2007) have demonstrated that many immigrant parents are simply unaware of this regulation and their children probably enter schools that offer less favourable environments with respect to their pupils’ backgrounds. According to these findings, ethnic school segregation dynamics represent an additional strong challenge faced by the German educational system, as also confirmed by this analysis.

Interestingly, the raw disadvantages of the first generation of pupils are instead wholly explained by socio-economic and cultural background. Here, first-generation status does not show any residual effect on educational performance. A potential explanation of this pattern may probably be found in their ethnic background. Indeed, the majority of first-generation pupils originated in the former Soviet states and are thus likely to be linked to the Spa¨t-Aussiedlerimmigration flows, which reached a total of three million people entering Germany between 1988 and 2005.Spa¨t-Aussiedlerare ethnic Germans, so-called repatriates, who are characterized by good levels of education and language skills, and this probably makes their children experience fewer dissonances at school (O¨ zcan 2007).

The French context is a case in point. Here, after controlling for the socio-economic status of parents, cultural-capital background, pupils’ aspirations and ethnic school segregation, the educational disadvantage disappears for the second generation, while it tends to persist with respect to the first one (Table 3). So, in France, second generation status does not affect independently educational performance.

Here, the main constraint faced by immigrants’ children at school seems to be linked to ethnic school segregation dynamics. These results support the findings of numerous studies that have continually emphasized the importance of considering ethnic spatial segregation as an important determinant in negatively affecting the integration trajectories of immigrants and their children and have also pointed out the inadequacy of current public policies addressed at avoiding these dynamics (see, among others, Felouziset al. 2005; Simon 1998). Similarly, in the Italian case, the language spoken at home does not have any

(11)

significant effect. A potential explanation may be linked to French colonial history and assimilation policies, which have made the linguistic issues less constraining for immigrants’ children. However, data limitations are valid also in the French case.

Contrary to expectations, after adding the index of pupils’ aspirations and motivations, the gap in educational achievement tends to increase instead of diminish. In other words, we found that, despite immigrants’ children having higher aspirations than their native counterparts, this seems to lead to a sort of Table 2Logistic regression results: odds ratio of achieving high performance in test scores, Germany{

Germany Model 1 Model 2 Model 3 Model 4 Model 5

Generation status

Second generation 0.31*** 0.43*** 0.61** 0.62** 0.71**

First generation 0.34*** 0.40*** n.s. n.s. n.s.

þSocio-economic status

(ref.)

1 1 1 1

country (ref.)

1 1 1

Other languages 0.51*** 0.51*** 0.53***

ISCED 5 or 6 (ref.) 1 1 1 ISCED 3 or 4 NS NS NS Below ISCED 3 0.29*** 0.31** 0.32** Cultural possessions 410 (ref.) 1 1 1 7–10 0.46*** 0.50*** 0.50*** 57 0.28*** 0.29* 0.29* þRisk-aversion theory

{

445

(12)

anxiety and ultimately cause an additional effect of differentiation between immigrants’ children and their native peers.

Concluding remarks

Notwithstanding the strong incentives aimed at improving outcomes for immigrants’ children in the educational context, in most industrialized countries they do not perform as well as others. This achievement gap and Table 3Logistic regression results: odds ratio of achieving high performance in test scores, France{

France Model 1 Model 2 Model 3 Model 4 Model 5

Generation status

Second generation 0.53*** 0.69*** 0.79* 0.70** n.s.

First generation 0.34*** 0.37*** 0.39** 0.30*** 0.34***

þSocio-economic status

(ref.)

1 1 1 1

country (ref.)

1 1 1

Other languages n.s. n.s. n.s.

ISCED 5 or 6 (ref.) 1 1 1 ISCED 3 or 4 0.84** n.s. n.s. Below ISCED 3 0.67*** 0.72** 0.72** Cultural possessions 410 (ref.) 1 1 1 7–10 0.33*** 0.37*** 0.37*** 57 0.14*** 0.15*** 0.15*** þRisk-aversion theory

{

(13)

its policy implications are shown to be particularly relevant when considering the trajectories of second-generation immigrants who were born, socialized and educated in the host country and who are thus supposed to share ceteris paribusthe same opportunities as natives. By using the OECD’s 2006 Program for International Students Assessment dataset, this paper aimed to investigate the extent to which generation status exerts an independent effect on early school performance both conditional and unconditional upon socio-economic resources, cultural-capital background, pupils’ aspirations and ethnic school segregation dynamics by comparing the Italian case with the French and German ones, which broadly correspond to three different stages of immigration of European economies. In this respect, our study proved to be extremely relevant for policy makers as our cases, together, fully represent the development of the immigration process and integration in Europe.

Interesting findings had been indicated by empirical analysis. As a whole, the importance of the traditional determinants as the main drivers of the apprehension process had been confirmed, although substantial differences are observed between countries.

In Italy, socio-economic determinants and ethnic school segregation dynamics are the main constraints for second-generation immigrants in the achievement of a good performance at school, underlying the importance of social class and scholastic environment in the Italian educational system. Here, generation status has an important residual effect – after controlling for the classical factors of educational dissonance – and leads to substantial concerns about the Italy’s capacity to carry out equal opportunities in the social mobility process. In the German case, too, educational performance is affected, ceteris paribus, by second-generation status. Here, language mastery and cultural background represent the main constraints in the educational performance. In addition, first-generation pupils represent a case in point by achieving similar scholastic performances to natives, once controlled for socio-economic status, and this is probably related to their ethnic background. Finally, in France, after controlling for socio-economic status, cultural background, pupils’ aspirations and ethnic school segregation dynamics, belonging to the second generation does not have any significant effect on educational achievement: among others, ethnic school segregation dynamics have the largest impact in reducing the ethnic gap.

Indeed, the main point emphasized by this analysis is the complexity of the phenomenon that presents different characteristics and problems between countries. Except for the French case – and the German one for first-generation pupils – the status ‘immigrants’ children’ is more likely associated with negative scholastic performance compared with native peers, an alarming finding especially for second generation-immigrants. In addition, this research raises more questions about what kind of additional barriers are preventing immigrants’ children in achieving at least the same educational performance as their native counterparts. In this respect, further analyses will be oriented

447

(14)

towards detecting the impact of macro-variables related to the different national educational systems in which pupils are inserted, for example early selectivity processes, the implementation of scholastic policies directly addressed to multicultural issues, institutional mechanisms of discrimination to investigate in more depth the differences between countries and to understand the phenomenon better. So, a multilevel analysis probably represents the most suitable tool for unravelling further this phenomenon.

The reception of children of different origins thus poses new (and old) challenges to educational institutions in terms of adaptation of teaching and learning methods to different cultural connotations. The issue is to search for the best policies which may avoid trajectories of ‘segmented assimilation’ (Portes 1996; Portes and MacLeod 1996; Portes and Rumbaut 2001; Zhou 2001), according to which just a fraction of the immigrants’ children have access to mainstream society, while the majority of them enter and remain at the bottom of society, in the underclass.

Notes

1 According to this theory, stratification turns out instead to be an advantage when educational achievement is measured by continuation rates. Since immigrants are positively selected for their ambitions and aspirations, their children are more likely to show higher expectations than their native counterparts in continuing school once the compulsory period is ended.

2 Because of data limitations, it was not possible to investigate further the effect of the country of origin, which has been continuously found as a main driver in determining the scholastic trajectories of the integration of immigrants’ children (see among others: Glick and Hohmann-Marriot 2007; Portes and Rumbaut 2001). In France, in fact, owing to legislative constraints, the PISA staff could not ask for it.

3 Sociologists identify the ‘primary socialization’ as the socialization children experience in the first years of their life, during which the family context assumes a fundamental role. In addition, in the ‘secondary socialization’, other social agents assume relevance that often impose different values from those elaborated during the primary socialization. In particular, in the secondary socialization, school has the predominant role in driving children’s choices and attitudes.

References

Bialystok, Ellen (2001) Bilingualism in Development. Language, Literacy & Cognition, Cambridge: Cambridge University Press.

Bourdieau, Pierre and Passeron, Jean-Claude (1970) La reproduction. Ele´ments pour une the´orie de syste`me d’enseignement, Paris: Editions de Minuit.

——— (2000)Reproduction in Education, Society, and Culture, London: Sage.

Breen, R. and Goldthorpe, J. H. (1997) ‘Explaining educational differentials: towards a formal rational action theory’,Rationality & Society9(3): 275–305.

Brinbaum, Y. and Heath, A. (2007) ‘Explaining ethnic inequalities in educational attainment’,Ethnicities7(3): 291–304.

Felouzis, Georges, Liot, Franc¸oise and Perroton, Joe¨lle (2005) L’apartheid scolaire, Enqueˆte sur la se´gre´gation ethnique dans les colle`ges, Paris: Editions du Seuil.

(15)

Goldthorpe, J. H. (2010) ‘Class analysis and the reorientation of class theory: the case of persisting differentials in educational attainment’,British Journal of Sociology 61(1): 311–35.

Kristen, C. and Granato, N. (2007) ‘The educational attainment of the second gen-eration in Germany. Social origins and ethnic inequality’, IAB Discussion Paper, 4. Marks, G. N. (2005) ‘Accounting for immigrant non-immigrant differences in reading

and mathematics in twenty countries’,Ethnic and Racial Studies28(5): 925–46. OECD (2006) Where immigrants succeed – a comparative review of performance and

engagement in PISA 2003, OECD.

O¨ zcan, V. (2007) ‘Focus Migration. Country profile: Germany’, available at: http:// www.focus-migration.de (last accessed 1 April 2011).

Portes, Alejandro (1996) The New Second Generation, New York: Russell Sage Foundation.

Portes, Alejandro and Macleod, Dag (1996) Educational progress of children of immigrants: the role of class, ethnicity and school context, Sociology of Education

69(4): 255–75.

Portes, Alejandro and Rumbaut Rube´n G. (2001)Legacies: the Story of the Immigrant Second Generation, Berkeley, CA: University of California Press: Russell Sage Foundation.

Rumberger, R. W. and Willms, J. D. (1992) ‘The impact of racial and ethnic segregation on the achievement gap in Californian high schools’, Educational Evaluation and Policy Analysis14(4): 377–96.

Sabbadini, L. L. (2008) ‘Mobilita` sociale e traiettorie di vita: il percorso della statistica ufficiale’, Conference Proceedings: IX Conferenza di Statistica, Rome: ISTAT. Simon, P. (1998) ‘Ghettos, immigrants, and integration. The French dilemma’,

Netherlands Journal of Housing and the Built Environment13(1): 41–61.

Van de Werfhorst, H. G. and Hofstede, S. (2007) ‘Cultural capital or relative risk aversion? Two mechanisms for educational inequality compared’,British Journal of Sociology58(3): 391–415.

Wang, J. and Goldschmidt, P. (1999) Opportunity to learn, language proficiency, and immigrant status effects on mathematics achievement’,Journal of Educational Research

93(2): 101–11.

Zhou, Min (2001) ‘Progress, decline stagnation? The new second generation comes of age’, in Roger Waldinger (ed.) Strangers at the Gates: New Immigrants in Urban America, Berkeley, CA: University of California Press, pp. 272–307.

449