Children of labour immigrants and their schooling in Luxembourg and Switzerland: similarities and differences between Portuguese-, Italian-, and ex-Yugoslavian-origin students
paper to be presented at the IMPALLA-ESPANET Conference: “The European Social Model in a Global Perspective” (6-7 March 2009, Luxembourg)
- please, do not cite without permission -
Aigul Alieva IMPALLA programme CEPS/INSTEAD (Luxembourg) & KU Leuven (Belgium) E-mail: firstname.lastname@example.org
In European realm Luxembourg and Switzerland is home to the largest number of immigrants, as a proportion to their total population (42% and 20.6% respectively (OECD 2008)). For both countries, immigrants from Portugal, Italy and former Yugoslavian republics represent the largest foreign-born population segment thus forming the “visible” minority in both cases. For that reason, it is evident that the issues related to their socio-economic and political integration are discussed with certain frequency. Studies in migration research most commonly look at the differences in occupational and educational achievement between groups of various origins within same country, often comparing them to the native population. This paper looks at the schooling performance of students in mathematics and aims to show that there are: a) significant differences between these three groups within the same country, as well as b) significant differences for same-origin groups between different countries. In case of between-group differences, future studies might benefit from testing models such as “segmented assimilation” or “downward assimilation” hypotheses developed in migration research. In the second case, studies could further test institutional hypothesis, when the differences between structural elements of the society produce variations in a particular outcome. The study is based on data from OECD PISA 2003 study and applies SEM techniques.
International migration nowadays is bringing together a varied student population into schools across the globe. Hence, the central questions that arise are: “Are the schools prepared to teach this diverse body of children?” and “How efficient are educational systems in competitive economies?” These challenges will continue affecting educational systems to a great extent in many European countries. Recently a new body of international and cross-country comparative studies has become available, mostly due to large datasets like PIRLS (2001, 2006), TIMSS (1995, 1999, 2003, 2007) and PISA (2000, 2003, and 2006). The most recent study with a particular interest in performance of
immigrant students in OECD countries (OECD 2006) has concluded that social origin and language skills do not fully explain the differences in performance between them and native students.1 Majority of theoretical frameworks for understanding life course outcomes of immigrants are coming from the USA. Educational attainment of certain ethnic groups did not improve over time, as is expected according to the classical assimilation model (Gordon 1964). Instead, the empirical evidence gave enough grounds to formulate the downward assimilation model (Portes and Rumbaut 2001), where second generation immigrants perform worse in school than first generation students. American studies portray Asian and European immigrants as better-doers, in high contrast to immigrants originating from Latin America and Africa. Several theories were employed for explaining these phenomena and particularly dominant are the theories of social and human capital. Borjas (1994, 1999) has suggested that, in the USA, the higher level of education and social capital of European immigrants continues conveying its influence on third and later generations. Alba, Lutz and Vesselinov (2001) have found evidence of the contrary: no correspondence is established between the first generation of immigrants from Europe and educational attainment of the third generation. This branch of research is continuously developing, but most researchers tend to support Portes and his associates with the model of segmented assimilation (1993, 1995) with the core argument about different immigrant groups adopting different patterns of assimilation/integration. In contrast to the experience of immigrants in the USA, educational attainment of those in Canada does not decrease but rather improves over the generations (Boyd 2000, 2002). These results go in line with what can be labelled as classical assimilation, success orientation (Boyd and Grieco 1998), or the immigrant optimism hypothesis (Kao and Tienda 1995). In other words, the downward assimilation model, where the second and following generations perform worse than the first generation, does not hold true for Canada. The strong educational mobility is argued to be the result of different history and nature of institutions and inter-racial and interethnic relations: Canada did not experience slavery or civil war as did the USA. Migration into Canada also took place later, mostly after the World War II, and the profile and origins of the immigrant population is different from the USA immigrant population (Boyd 2002:1043-1044). Moreover, racially and ethnically visible immigrants of the second and later generations in Canada attain higher levels of education, when compared to nonvisible immigrants (Boyd 2002).
Regression estimates of PISA score on immigrant status, education of parents, their jobs and language spoken at home have produced on average R-squared equalling to 0.12 (OECD 2006:201-202), ranging from highest R-squared for Germany (0.20) and Belgium (0.19) and lowest for Canada (0.08).
The majority of immigrants in Europe were coming from their Southern neighbours, and can to a certain degree be compared to economic immigrants. The main difference, however, was that these immigrants arrived by a Guest Workers programme, which was institutionalised by bilateral agreements on a state-level. Today, the integration of Italian, Spanish, Greek, and Portuguese descendants is regarded as the silent success, in contrast to the situation of Moroccan- and Turkishorigin immigrants (Lindo 1995, 2000). These latter groups had starts similar to the immigrants from the South European countries: they arrived as guest workers in the 1960s as low skilled or unskilled labour with little education, aged somewhere between 20 and 30 years old. Nowadays, the situation of Southern Europeans is described as a quick success in, for example, the Netherlands, as the educational level of 1st generation was overall low, their children (2nd generation) have achieved more advanced levels of education, nearly equal to their Dutch peers. This is the evidence of strong intergenerational advancement, combined with overall successful labour market performance, higher income and better housing conditions, more contacts with native population and more frequent mixed marriages (Lindo 2000). Only recently have researchers included the institutional context into explanations of why immigrant students do not perform equally well in schools as native ones are (Buchmann and Parrado, 2006; Crul and Vermeulen, 2006; Crul, 2004, 2005; Dronkers and Levels, 2006; Levels and Dronkers, 2005; Levels, Dronkers, and Kraaykamp, 2006; Marks, 2005; Schnepf, 2006). They come more often to an agreement that institutional differences play a significant role in shaping educational and economic outcomes of non-native groups, even though this role is not as large as their socioeconomic background. This becomes increasingly clear in cross-country comparisons. I bring here one study as telling example that was also among pioneering works. M. Crul and H. Vermeulen (2006) analysed the position of second generation Turks in European countries, namely Germany, Austria, Switzerland, the Netherlands, Belgium, and France. They have compared school attendance rates, educational performance of children, highest educational attainment, and dropout rate. With respect to the employment position, they have compared rates of unemployment and its duration, current job level and income. They have concluded that considerable differences exist among these countries. First of all, there are differences in the age at which children start school. For example, in Belgium and France semi-obligatory crèche and kindergarten groups welcome children from the age 2.5 and 2 years old, respectively. Therefore, the authors argue, having 3.5 years of socialisation in the majority language is a significant advantage to just 1 year, as is the case in German-speaking
countries. Second, the selection into different educational programmes takes place at different ages in different educational systems. For German-speaking countries, it occurs at 10 years old, while for Belgium and France at 15, and in the Netherlands between 12-14 years old. In the secondary education, up to 75% of students with Turkish background are channelled into lower vocational tracks in German-speaking countries, while 50-70% of them go to higher academic tracks in France, Belgium and the Netherlands, and as expected, the university enrolment rate is higher in later countries. A third important factor is the number of hours spent face-to-face with a teacher. In German-speaking countries this indicator is significantly lower than in, for example, the Netherlands. This is again due to the differences in schooling systems, such as half-day schooling in Germany versus full-day schooling in the Netherlands. The next aspect which helps to explain the differences is how much help and support is provided within and outside of school. Again, there are differences between the German-speaking states and others. In the former group, especially in Germany, children receive the least support with homework and language learning, while in France, Belgium and the Netherlands, the amount of assistance is the highest. Language programmes, for instance, were introduced at different years. France did this in the 1970s, Belgium in the late 1980s, and Germany had fewer projects even in the 1990s. In the next step, the authors have analysed the performance of Turkish students in different types of schools. Although compared to other countries, more children in France and the Netherlands enter tracks for higher education, more drop out from the vocational education, and thus almost 50% of school children have no diploma. The dropout indicators are lower for the Netherlands (approx. 35%) and far less in Germany, Austria, and Switzerland. In countries that have early selection, or channelling, more children are directed into vocational schools. And in countries where channelling is later or nonexistent, such as France and Belgium, the education systems allow more young Turks to pursue higher educational degrees, although the dropout rates remain high. This study shows that an immigrant group with many characteristics in common, like Turkish students in this case, attain different positions depending on the institutional setting.
Institutional effects and performance of Turkish-origin students in Europe Belgium France 2 15 Germany 6 10 661 60-75 1990s Low 7 11 Netherlands 4 12-14 1019 25-35 1990s High 21 23
Education starts (y.o.) Selection at age (y.o.) Contact hours with teachers (per year) Tracking (% in lower vocational school) Language learning help (introduced since) Help and support outside school Drop-out (%) Higher education enrolment (%) Source: Vermeulen and Crul (2006)
30-50 1980s High 50
25-35 1970s Highest 46 21
This paper aims to formulate some tentative ideas on whether either assimilation/integration models or the institutional theory can be tested in the future. The study concentrates on three immigrant groups represented in two European countries, namely Portuguese, Italian, and ex-Yugoslavian offspring residing in Luxembourg and Switzerland. The study tests whether there are differences between groups within the same country, and between the countries within the same-origin group. It might well be that a number of various processes is taking place. For instance, there might be certain affinities within the same ethnic group living in different countries. If this is the case, the future studies can benefit from testing the continuity effects of the country of origin. If immigrant groups behave differently within the same country, then testing models like segmented assimilation might prove to be useful. And finally, if immigrant groups within one country significantly differ from comparable immigrant groups in another country, then indeed the institutional effects should be studied more vigorously.
Immigration and Integration processes in Luxembourg and Switzerland
Luxembourg is largely known today as an immigration country, with a history of labour immigration similar to the neighbouring Germany, Switzerland, and, to some extent, Belgium. But back in the 19th century it was largely an emigration country with 40% of its population leaving it between 1841 and 1891 (Cordeiro 1976 in Waringo 2004). The steel industry, which was set up with the help of German engineers and workers in the second half of the 19th century, made Germans the largest and
the fastest growing group of immigrants in Luxembourg and by 1910 they made up 15.3% of its entire population. The growth of the steel industry led to an increased demand for low-skilled labour and the beginning of the 20th century witnessed a large influx of workers from Italy. These two groups were the largest immigrant groups in Luxembourg, making up 49.3% of its industrial labour force (Hoffmann 1995 in Waringo 2004). The number of foreign workers during the period between WWI and WWII fluctuated constantly, and reached its lowest limit by the end of WWII. Luxembourg had to compete with other countries for attracting unskilled workers, and in 1957 the Grand Duchy signed an agreement with Italy that permitted family unification. There was a certain contradiction in the immigration policy at that period: despite the urgent need for labour force, the policy remained rather restrictive, which led to a low number of re-united families and low number of Italian workers altogether. During the first half of the 1960s, Luxembourg signed a bilateral agreement with Portugal and the number of Portuguese workers soon outgrew the Italian community. Despite the higher turn-out of labour immigrants, the outflow remained very high as well, particularly among Italian workers. In an attempt to meet the needs of its industrial economy, two additional contracts with Portugal and Yugoslavia were signed in 1970. Various authors described the recruitment policy as limited to the Romanophone countries of Mediterranean region for “...the nationals of these countries cause fewer problems with regard to their integration” (Waringo 2004:8). In 1974 a new government was formed in Luxembourg, led by Social Democrats, and an integration policy was for the first time introduced at the national level. However, the subsequent economic crisis forced the government to almost halt the recruitment of new workers. While Luxembourg could introduce more restrictions towards immigrants from non-European countries, little could be done with regard to those coming from within the European Community. When Portugal, Greece, and Spain joined the EC Luxembourgian government, fearing the massive inflow of Portuguese, requested special arrangements to limit the freedom of movement of the labour force. In the 1980s the economy continued to grow, and this time these were the frontaliers –cross-border commuters from neighbouring regions of Germany, France, and Belgium who rapidly became one of the main sources of labour for Luxembourg. From 1984 to 1994 their share in total employed force grew from 10.9% to 26.5% (SESOPI 1994 in Waringo 2004), and today they make up 38.4% of the entire labour force2. The frontaliers also serve as a cushion against unemployment in Luxembourg, i.e. they are the first to lose their jobs when there is economic stagnation. Current data
The empirical analysis in this project does not include children of cross-border commuters because only in very rare cases do their children attend schools in Luxembourg.
show that in addition some groups have lower salaries when compared to native Luxembourgian employees (Brosius 2005). In 2003 there were 38.2% of foreign-born nationals in Luxembourg, of which the largest group is Portuguese, with 14.3%, followed by French and then Italian immigrants. It is evident that Luxembourg has a very low number of immigrants from the developing countries.
Table 2. Immigrant population in Luxembourg in 2003 Nationality Luxembourgian Portuguese Belgian French German Italian Other EU-15 New EU-25 Other non-EU Total Missing Total with missing Number 272086 62986 13212 27266 10952 19080 15211 1094 18341 440226 264 440490 Percent 61.8 14.3 3.0 6.2 2.5 4.3 3.5 0.2 4.2 99.9 0.1 100
Data: PSELL 2003, weighted by individual weights
More recent debates in Luxembourg, similar to in other European countries, are concerned with demographic challenges. On the one hand, Luxembourg is gradually turning into an ageing society, and the sustainability of its very generous welfare benefits and pensions will at some point be challenged. In addition to that, all the transfrontaliers once retired also receive their pensions from Luxembourg. Logically, as their number is steadily growing, the larger share of social security contribution will be sent abroad in coming decades. Although, the birth rate in Luxembourg is also low, it is much lower among the native Luxembourgian population than among the non-native population. An outcry against “the suicide of the nation” (Waringo 2004, Kollwelter 2005) makes the immigration debate a more heated subject. The naturalisation of immigrants in Luxembourg is steadily increasing in recent years. The law on Luxembourgian nationality from 24th of July 2001 states that in order to become a citizen, an individual should prove residence in the country during the last five years, give up his or her previous citizenship, and demonstrate knowledge of one of the national languages, as well as basic knowledge of the Luxembourgish language. This procedure is applicable for those who were born in
Luxembourg to parents of foreign nationality, who have received their obligatory education in Luxembourg, and those foreign citizens married to Luxembourgian nationals3. Similar to Luxembourg, the massive immigration to Switzerland started in the second half of the 19th century, with the start of industrialisation, reaching 14.7% by 1910 (Mahnig and Wimmer 2003). During the WWI and WWII this number dropped and reached 5.2% in 1941. The Swiss economy and industry have suffered much less from the wars, and shortly after the end of WWII, the second wave of immigration started. The peak was reached in 1970, when the foreign population reached 17.2%, or more than 1mln people. The international economic crisis in 1973-1974 hit immigrant workers most severely: 67% of them lost their jobs and 35% of them returned home. This happened not only due to unemployment, but also because of the expiration of temporary residence permits, which were not prolonged by Swiss authorities. The precarious working and legal conditions forced governments of the sending countries, in this case Italy, to act. With the help of their government the Italian immigrants created their own local organisations and unions to help to improve their working and living conditions. Since the Swiss authorities and municipalities had no plans of developing any support measures for immigrants at that period, the number of these associations grew very rapidly. Other immigrant groups which arrived after Italians did not succeed in creating such local support systems; it then became the responsibility of local municipalities and cantons. The economy's quick revival stimulated a new influx of immigrants during the 1980s, mostly as a process of family reunification, with the immigrant population reaching 18.1% in 1990. The latest immigration wave took place in 1980s and 1990s from the republics of ex-Yugoslavia. It had also started as labour migration, and in 1990s was intensified by the political and military conflict. In 1995 the new policy in Switzerland restricted the access to work permits, and the flow of Yugoslavian-origin immigrants turned into highly-skilled and highly-educated flow (Gross 2006a:20). If earlier immigrant workers were used as an “economic buffer”, today this policy is no longer successful for two main reasons: Firstly, as the figure 1.1 below shows, majority of immigrant families have by now obtained the status of legal residents. In other words, the enjoy to a large extent same legal rights as the Swiss citizens;
See the link for more details: http://www.gouvernement.lu/tout_savoir/population_langues/natiolux.html
Secondly, in 2002 Switzerland started implementing the agreement of workers’ free mobility in accordance with EU/EFTA, and that has decreased the legal authority of the Swiss government to control the immigrant flows into the country (Gross 2006b).
Today the largest groups are Italians (19.3%), nationals of Serbia and Montenegro (12.8%), Portuguese (11%) and Germans (10.3%).
Table 3. Immigrant population in Switzerland in 2005 Number 158 700 70 900 297 900 167 900 72 200 196 800 75 900 501 600 1 541 900 Percent 10.3 4.6 19.3 11 4.7 12.8 4.9 32.5 100
Nationality Germany France Italy Portugal Spain Serbia and Montenegro Turkey Other Total number of foreigners
Source: BFS 2007
A large proportion of immigrants do not have Swiss citizenship, and the children of immigrants who were born there are referred to as “immigrants of the second generation” (Mahnig and Wimmer 2003:139). Obtaining citizenship is not an easy matter in Switzerland; the reasons for such a strict policy were formed in 1930s with the introduction of the Bundesgesetz über Aufenthalt und Niederlassung der Ausländer – ANAG (the Federal Law of Residence and Settlement of Foreigners). To a large extent this law was a reaction to xenophobia against Überfremdung (“overforeignisation”) and a feared loss of identity in the event that the immigrant population would outgrow the native population. Although immigrants do not enjoy the same rights as Swiss citizens, those from the EU countries are eligible for permanent residence permit after 5 years of living in Switzerland, while citizens of other countries are eligible after 10 years. The graph below shows that 66% of immigrants (or 1 087 500 individuals) have already received permanent residence permits. Nearly a quarter of the nonnative population has temporary permits, and other groups do not exceed 1-2% in each category.
Immigrants by the type of residence permit
The process of naturalisation, as mentioned earlier is not simple in Switzerland and has three main stages. First of all, an individual must become a citizen of a municipality, and then of a canton. In order to do that, he or she applies for a federal authorisation of naturalisation through the Federal Department of Police. After receiving this document, the person applies for the right of citizenship in the municipality. Municipalities have the rights of creating additional criteria for naturalisation. These rules are built on “an ethno-cultural “logic” (Mahnig and Wimmer 2003:147). In addition to being a complex procedure, it also entails very high costs. Again, on a regional level, there are differences between French- and German-speaking cantons. French-speaking cantons apply more formalised rules, while in German-speaking cantons, the principle of citizens’ participation is preferred. The federal structure of Switzerland gives lots of decision-making freedom to its cantons, which has both positive and negative aspects. As a positive aspect, each canton can better adjust its measures and policies depending on needs, cultural diversity, and specificities of its population. However, the negative side lies in the large diversity of institutional organizations. It is well exemplified with the presence of 26 educational systems, equal to the number of cantons. Although there is a centralized body Schweizerische Konferenz der kantonalen Erziehungsdirektoren – EDK, all the recommendations, including those aiming at homogenisation of curricula and structure are non-obligatory on a cantonal level. On a regional level one can see the general differences in pursuing integration policy at school. German-speaking cantons place newcomers into separate educational intuitions, while Italian- and French-speaking cantons allocate them into the mainstream schools.
Similar to the education policy, there is a formal centralised institution for immigration and integration – Eidgenössische Ausländerkommission – EKA (“Federal commission for foreigners”), but again its authority has a limited power. The commission is mostly a consultative body with members coming from various segments like non-governmental associations, education institutions, trade unions, immigrants’ organisations, governmental bodies and so forth.
Data and model
The main data used in this project is from the OECD Programme for International Student Assessment (PISA) 2003 study on students aged 15 years old. Both Luxembourg and Switzerland took part in waves 2000, 2003, and 2006. The main focus of the study in 2003 was mathematical skills, with other skills such as reading, science, and problem-solving also tested. Students took part in a two-hour test that includes open and multiple-choice questions, plus spent approximately 30 minutes filling in individual questionnaires for gathering information on their background, learning and motivation. The school principals also complete questionnaires about their schools, teachers, and learning environments. The main dependent variable in this study is the mathematics score which is less biased towards immigrant students than, for instance, reading. However, maths knowledge tends to be more gender-biased, i.e. boys perform better in maths and science tests than girls, at least among the native-born students (OECD 2003). Mathematics scores are grouped into six levels of competences, and the OECD mean is set at the level of 500 points with the standard deviation of 100 points4. 8420 student took part in Switzerland in the PISA study of 2003. 48.4% (4079) are females and 51.6% (4341) males. Out of the 20% who are of migratory background 55.3% are first-generation students, whose parents and they themselves were born abroad. Nearly 60% of all these migrant children came to Switzerland before the age of 6, with nearly 20% of them being infants between 0 and 2 years old. 3923 students have participated in PISA 2003 in Luxembourg. 50.8% (1992) are female students and49.2% (1931) are male students. 33.2% of all students have migratory background, 52% of which are first-generation generation and, similar to Switzerland, around 60% have arrived in Luxembourg
before turning 6. Both countries have similar distributions of age at migration among firstgeneration immigrants. Altogether there are seven latent variables to measure the effect of individual characteristics on performance in mathematics, reading and sciences. The interpretation of results will be given with regard to only mathematical performance. Apart from the socio-economic and cultural capital of families, other effects such as generation, gender, spoken language and type of educational programme attended by student are measured5. Theoretical and empirical literature demonstrates that some of the individual background factors often correlate between each other. This is the case for economic, cultural, and educational assets. Also language skills of immigrant population improve from the first to the second generation. The advantage of the structural equation modelling is the possibility to adjust the model assuming the relation between these variables.
Model for immigrant population
See Annex for more information about variables
The basic equation is written the following way:
Achievement = w1 * (Socio-economic status) + w2 * (Single parent family) + w3 * (Gender) + w4 * (Cultural capital) + w5 *(Study programme) + w6 *(Generation) + w7 *(Foreign language) + Error
In this case, the equation will be completed by additional specifications:
Cultural capital = z1*(Socio-economic status) + z2*(Single parent family) + Error Foreign language = d1*(Generation) + d2*(Single parent) + d3*(Socio-economic background) + d4*(Cultural capital) + Error Study programme = k1*(Socio-economic status) + k2*(Cultural capital) + k3*(Foreign language) + k4*(Single parent) + Error
where z1-z2, d1-d4, k1-k4 are parameters to be estimated.
Empirical results 1.1.Descriptive analysis
In both countries there are significant proportions of students of Portuguese background. Nearly 600 students in Luxembourg in PISA 2003 have Portuguese background: 322 of them were born in Portugal; around 595 have Portuguese-origin parents. In Switzerland this number is lower, these are approximately 200 students: 133 were born in Portugal, and approximately 200 have Portugueseorigin parents. Portugal has also participated in the PISA study and it is therefore possible to look for the patterns of affinity between Portuguese immigrants in Luxembourg and Switzerland and the native population in Portugal. Table below shows that there are many similarities between these three groups. The average result in mathematics is 466 in Portugal, which is also well below the OECD level (500). Students with Portuguese background in Luxembourg have performed lower (446), and those in Switzerland have slightly better results (478). With regard to the socio-economic and occupational status, there are some, but modest differences. The average occupational status of Portuguese in Luxembourg and Switzerland is identical, and 10 standard units lower than in Portugal. This is not surprising since emigration took place among low-skilled workers. On average, Portuguese parents in Switzerland and in Portugal have nine years of schooling, which corresponds
to obligatory schooling (ISCED2 ). Those in Luxembourg however, have on average one year less of education than in Switzerland and Portugal.
Table 4. Portuguese immigrants in Luxembourg and Switzerland compared to Portugal Portugal Average math score 466.14 (0.28) 43.10 (0.05) 9.24 (0.02) Portuguese in Luxembourg 445.97 (3.22) 33.74 (0.49) 8.30 (0.21) Portuguese in Switzerland 478.40 (1.99) 33.84 (0.31) 9.21 (0.10)
Average highest parental occupational status Average highest parental education in years of schooling Data: PISA2003 Weighted by individual weights; standard error in parentheses
In a similar fashion, we look at Italian-origin students. There were approximately five million Italians who migrated to Switzerland between the 1870s and the 1980s. Currently, there are approximately 300,000 people of Italian descent in Switzerland. Once they were the largest immigrant group, but nowadays they are outnumbered by Yugoslavians. The situation for Italian immigrants is similar to that for Portuguese immigrants: their average math score and socio-economic status are nearly same; the only difference is that Italian immigrants in Switzerland have 2 years less education than do Italians in Italy. The occupational status and education of Italian-origin parents in Luxembourg is high and comparable to those in Italy. Those in Switzerland have on average 1.5 years of education less, and their occupational status is somewhat lower. At the same time, students’ test scores are equal in all three countries.
Table 5. Italian immigrants in Luxembourg and Switzerland compared to Italy Italy Average math score 465.76 (0.14) 46.83 (0.02) Italians in Luxembourg 466.66 (8.69) 46.15 (1.53) 12.46 (0.56) Italians in Switzerland 468.89 (2.00) 39.00 (0.32) 10.95 (0.06)
Average highest parental occupational status Average highest 12.52 parental education in (0.051) years of schooling Data: PISA2003 Weighted by individual weights; standard error in parentheses
(Ex-)Yugoslavian immigrants have a less similar history in Luxembourg and Switzerland. While they were arriving to Luxembourg in the 1970s also as low-skilled labour, their number remained small until a new influx took place in the 1990s when many came as war-refugees. Yugoslavian-origin workers began migrating to Switzerland to work in the 1980s in steadily growing number. In 2000 they outnumbered the previously largest group of immigrants in Switzerland, Italians (337335 vs. 319675) (Gross 2006:33). This group shows the largest performance differences between Luxembourg and Switzerland. Immigrants of Yugoslavian origin in Luxembourg have educational and occupational characteristics similar to those in Switzerland. However, test scores of students in the latter country are 40 points higher. These results are also higher than achievements of students in Yugoslavia. In terms of occupational and educational level of the parents, those in Switzerland have two years less education and lower occupational position.
Table 6. (Ex)-Yugoslavian immigrants in Luxembourg and Switzerland compared to (Ex)-Yugoslavia Yugoslavia Average math score 436.53 (0.32) 48.07 (0.06) Ex-Yugoslavians in Luxembourg 420.98 (9.20) 36.03 (1.55) 11.40 (0.68) Ex-Yugoslavians in Switzerland 460.80 (1.39) 35.80 (0.19) 10.82 (0.06)
Average highest parental occupational status Average highest 12.75 parental education in (0.01) years of schooling Data: PISA2003 Weighted by individual weights; standard error in parentheses
1.2.SEM analysis – comparison within group between countries
1.2.1. Portuguese-origin students in Luxembourg and Switzerland Portuguese immigrants are the largest group and, therefore, the most “visible” group in Luxembourg and their integration causes controversial debate. Results below yield somewhat surprising low effect of socio-economic background. It means the difference between students coming from white-collar and blue-collar background of Portuguese origin in Luxembourg is 17 points, and in Switzerland 10 points. Educational track has the largest effect on differences among students. Students in vocational programmes score on average 18 points lower than those in general programmes in Switzerland, and 29 points lower than in Luxembourg. Language has a negative significant effect on performance in Luxembourg, but not in Switzerland. There is a positive significant effect of generation: those who were born and grew up in Luxembourg perform 11 points higher than those who were born abroad, while the effect is smaller in Switzerland with 6 points. Students with one parent perform 16 points lower in Luxembourg than those from twoparent families. There seems to be no such effect in Switzerland. The higher amount of cultural capital has a positive effect also only in Luxembourg, and not in Switzerland. The model explains 36% of variance in Switzerland and 22% in Luxembourg.
Table 7.Total unstandardised effects of latent variable s on mathematics score, Portuguese immigrants
PORTUGUESE IMMIGRANTS IN SWITZERLAND
Socio-economic status Cultural capital Single parent family Educational track Foreign language Second generation Male R-square Socio-economic status Cultural capital Single parent family Educational track Foreign language Second generation Male R-square Unstandardised coefficients -16.70 14.80 -4.01 -18.53 -6.65 5.98 3.75 Standard error 8.06 9.06 3.21 3.13 2.90 3.06 3.07 0.36 t-value -2.07 1.63 -1.25 -5.92 -2.30 1.96 1.22 p-value 0.04 0.11 0.21 0.00 0.02 0.05 0.23
PORTUGUESE IMMIGRANTS IN LUXEMBOURG
-10.19 8.34 -15.82 -28.66 -1.84 11.42 0.75 3.64 2.57 3.06 2.82 3.61 1.66 2.66 0.22 -2.80 3.25 -5.17 -10.15 -0.51 6.87 0.28 0.00 0.00 0.00 0.00 0.61 0.00 0.78
1.2.2. Results: Italian-origin students in Luxembourg and Switzerland
Immigrants of Italian origin in these countries are not different from the native population in Italy with regard to their social background characteristics or their achievement level. Low educational and social background of parents results in a lower score by 15 points in Switzerland and 30 points in Luxembourg compared to peers from their ethnic group with higher background characteristics. At the same time, different variables seem to be responsible for producing the outcomes for these two groups. For example, cultural capital, which was significant for Swiss students appears to less insignificant in the case of Luxembourg (p=0.06). For the effect of a single parent family, its effect is negative and significant in Luxembourg (-8.4), although twice smaller than for Portuguese offspring (-16), but not significant in Switzerland. Speaking other than the national or test language has no impact on test performance output in either case. The second generation of students have results on average 7 points higher than the first generation in Switzerland, while there seem to be no such effect in Luxembourg. Educational track does not have a statistically significant effect on mathematics for either group, which makes this ethnic group different from two others. Amount of the explained variance is rather modest (R2=0.13 in Swiss case and 0.23 in Luxembourgian case), compared to ex-Yugoslavians and Portuguese group.
Table 8.Total unstandardised effects of latent variable s on mathematics score, Italian immigrants
ITALIAN IMMIGRANTS IN SWITZERLAND
Socio-economic status Cultural capital Single parent family Educational track Foreign language Second generation Male R-square Socio-economic status Cultural capital Single parent family Educational track Foreign language Second generation Male R-square Unstandardised coefficients -15.20 15.98 -2.51 -6.63 6.30 7.02 1.74 Standard error 4.42 7.02 3.33 3.57 3.31 3.43 3.57 0.13 t-value -3.44 2.28 -0.76 -1.86 1.91 2.05 0.49 p-value 0.00 0.02 0.45 0.06 0.06 0.04 0.63
ITALIAN IMMIGRANTS IN LUXEMBOURG
-31.91 7.89 -8.38 3.72 -9.63 -0.66 -7.31 5.46 4.20 4.27 4.26 6.21 2.45 4.07 0.23 -5.84 1.88 -1.96 0.87 -1.55 -0.27 -1.80 0.00 0.06 0.05 0.39 0.13 0.79 0.07
1.2.3. Results: Ex-Yugoslavian-origin students in Luxembourg and Switzerland Parents of Yugoslavian students in Luxembourg and Switzerland have similar socio-economic and educational status. However, their average performance differs by approximately 40 points. For immigrant students in Switzerland, the major determinants of performance appear to be the socioeconomic status of parents and single parent families. Those students who come from lower-income background perform nearly 50 points lower than those from higher-income background. These estimates are very close to both native and non-native Swiss samples. These two variables alone explain 43% of the variance. However, all the results related to the performance of specific immigrant groups should be cautiously considered because of the small sample size in many cases. Immigrant students of ex-Yugoslavian origin in Luxembourg appear to be slightly less affected by lower socio-economic status of their parents, when compared to their peers in Switzerland. But in contrast to Switzerland, their mathematical performance is 15 points lower if they study in a vocational programme. All other variables are not statistically significant – there is no effect on performance of cultural capital, language used at home, gender, or differences between different generations.
Table 9.Total unstandardised effects of latent variable s on mathematics score, Yugoslavian immigrants
YUGOSLAVIAN IMMIGRANTS IN SWITZERLAND
Socio-economic status Cultural capital Single parent family Educational track Foreign language Second generation Male R-square Socio-economic status Cultural capital Single parent family Educational track Foreign language Second generation Male R-square Unstandardised coefficients -49.36 4.33 -17.09 4.40 0.90 3.80 5.59 Standard error 5.57 4.74 3.89 8.29 3.97 4.07 3.22 0.43 t-value -8.85 0.91 -4.40 0.53 0.23 0.93 1.74 p-value 0.00 0.37 0.00 0.60 0.75 0.82 0.09
YUGOSLAVIAN IMMIGRANTS IN LUXEMBOURG
-39.13 2.57 -4.53 -14.46 -0.34 -0.02 0.22 6.47 4.84 3.95 3.68 4.05 3.63 3.17 0.26 -6.05 0.53 -1.15 -3.92 -0.08 0.01 0.07 0.00 0.59 0.25 0.00 0.94 0.99 0.94
1.3.SEM analysis – comparison between groups within country
1.3.1. Luxembourg Estimates for particular immigrant groups should be accepted with caution, because the sample size is very small, especially for Italian and Yugoslavian immigrants. The differences between the three ethnic groups in Luxembourg are strongly pronounced. Also the reference group in the analysis below is the peers of the same ethnic group. For example, Italians of blue-collar parents are compared to Italians of white-collar parents. The effect of lower socio-economic background is strong for Italian and Yugoslavian students, and less so for Portuguese. In other words, the achievement difference between blue-collar and white-collar Portuguese-origin students is smaller, compared to the same categories of Yugoslavian-origin and Italian-origin ones. Cultural capital is, by contrast, not significant for the Yugoslavian group, but is especially significant for Portuguese students; it has also an effect on the performance of students of Italian origin. Coming from a single-parent family has both direct and indirect significant negative effects on maths score of Italian and Portuguese students, but not on Yugoslavian. The non-academic educational track has a particularly negative effect on Portuguese students, while the effect is absent for Italians. Speaking a foreign language at home has a statistically significant negative indirect effect for Portuguese students. Generation has a positive direct effect for Portuguese students and an indirect effect for Yugoslavian students. Portuguese students of the second generation score 11 points better than those of the first generation. Whether a student is male or female produces mixed results: while males perform worse than females among Italian students (p≤0.10), Portuguese male students perform better than their female peers. The same is true for Yugoslavian students, with the only difference being that female students appear to perform better than their male peers.
Table 10.Total and indirect effects on achievement scores of Italian, Portuguese and Yugoslavian immigrants in Luxembourg Italian total effect Socio-economic status Cultural capital Single parent family Educational track Foreign language Second generation Male indirect effect -0.12* (0.06) -0.07* (0.05) 0.05* (0.04) -----0.01 (0.03) 0.01 (0.01) Portuguese total effect indirect effect 0.13*** (0.04) --Yugoslavian indirect effect -0.48*** (0.08) 0.03 (0.06) -0.06 (0.05) -0.18*** (0.05) --indirect effect -0.05 (0.10) ---0.01 (0.02) ---
Chi-square Degrees of freedom p-value 0.00 RMSEA 0.06 ECVI 1.04 CFI 0.97 AGFI 0.87 R2 Structural 0.23 equations R2 Reduced 0.21 form equations +p≤0.10 *p≤0.05 **p≤0.01 ***p≤0.001 Standard errors in parentheses
-0.43*** (0.07) 0.11* (0.06) -0.11* (0.06) 0.05 (0.06) -0.13 (0.08) -0.01 (0.03) -0.10+ (0.06) 125.58 53
-0.15** (0.05) 0.12*** (0.04) -0.23*** -0.07*** (0.04) (0.02) -0.42*** --(0.04) -0.03 -0.07** (0.05) (0.05) 0.17*** 0.17*** (0.02) (0.02) 0.01 0.04*** (0.04) (0.01) 130.64 58 0.00 0.04 0.39 0.99 0.95 0.22 0.09
0.08*** (0.02) --0.11** (0.04) ---0.05*** (0.01) 109.14 64 0.00 0.04 0.52 0.96 0.94 0.26 0.21
1.3.2. Switzerland These three (Portuguese, Italian and ex-Yugoslavian) groups show large differences among each other. First of all, the low socio-economic background results in largest differences in maths scores among Yugoslavian students, and far less so among Portuguese and Italian students. Cultural capital does not seem to have an impact for results among Portuguese and Yugoslavians, while there is a positive effect for Italian students. Having one or both parents at home makes no difference for immigrants of Italian and Portuguese origin, but it matters for students of Yugoslavian origin. Educational track influences the achievement results of Portuguese students the most. While this impact is lower in the case of Italians, it is completely absent in the case of Yugoslavians. Speaking a foreign language leads to lower performance in mathematics only among Portuguese immigrant students, not others. And the second generation seems to be performing better than the first one in the cases of Italian and Portuguese students. Model fit parameters show that results are acceptable for all three groups, although the same caution should be used for generalising these results over an entire ethnic group student population because of the small sample size. The model explains 13% of variance for Italian-origin students, 36% for Portuguese-origin students, and 43% for Yugoslav-origin students.
Table 11.Total and indirect effects on achievement scores of Italian, Portuguese and Yugoslavian immigrants in Switzerland Italian total effect -0.20*** (0.06) 0.21* (0.02) -0.03 (0.02) -0.09* (0.05) 0.08 (0.04) 0.09* (0.04) 0.02 (0.05) 54.05 74 0.96 0.001 0.44 1.00 0.97 0.13 0.06 indirect effect -0.04* (0.02) ---------0.01 (0.01) --Portuguese total effect -0.27** (0.13) 0.24 (0.15) -0.07 (0.25) -0.30*** (0.05) -0.11* (0.05) 0.10* (0.05) 0.06 (0.05) 95.03 71 0.03 0.03 0.56 0.96 0.94 0.36 0.11 indirect effect -0.14* (0.08) 0.01 (0.02) 0.01 (0.01) ----0.02 (0.05) --Yugoslavian total indirect effect effect -0.70*** -0.03 (0.09) (0.05) 0.06 --(0.07) -0.24*** --(0.05) 0.06 --(0.12) 0.01 --(0.06) 0.05 -0.02 (0.06) (0.02) 0.08+ 0.00 (0.05) (0.02) 129.56 56 0.00 0.05 0.59 0.97 0.92 0.43 0.43
Socio-economic status Cultural capital Single parent family Educational track Foreign language Second generation Male Chi-square Degrees of freedom p-value RMSEA ECVI CFI AGFI R2 Structural equations R2 Reduced form equations *p≤0.05 **p≤0.01 ***p≤0.001 Standard errors in parentheses
Summary and conclusion
Previous empirical studies on performance of immigrant students’ show that the educational inequalities are more persistent among immigrant children than among natives and the primary sources of inequality are the social background and educational level of parents. Language deficiencies are another cause of low performance, but these are being gradually cured with length of living in the host country. The second generation shows significantly better language knowledge over all immigrant populations. However, the second generation does not necessarily outperform the first one in education and labour market integration. The reasons are partially lying in community and peer effects, and also in the limited social and human capital of a particular ethnic group. Nonetheless, the patterns of integration yield differing result, making the results of comparison between countries and between different ethnic groups less consistent. Empirical analysis in this paper adds some important findings, even if altogether they do not lead to more clarity in our knowledge. To summarise: Portuguese students perform on average better in Switzerland, compared to those in Portugal and Luxembourg. Also, within Switzerland, this group performs better than Italian and Yugoslavian-origin students, despite the fact that their parents have on average 1.5 years less of education and have lower occupational status. Portuguese students in Luxembourg perform least well compared to those in Portugal and Switzerland, but they perform better than students of Yugoslav origin, given that their parents have on average three years of less education than ex-Yugoslavian parents. Also, Portuguese students in Luxembourg tend to improve their performance from the 1st to the 2nd generation. Italian students show strong similarities in their performances. Mathematics scores are on average the same in Italy, Luxembourg and Switzerland. Immigrant families in Luxembourg have higher SES and more years of education than in Switzerland and in that regard they are more comparable to Italian sample. While the 2nd generation of Italian-origin students show progress in their performance, there seem to be no such process in Luxembourg. Ex-Yugoslavian students, similar to Portuguese show variation in their performance. Those in Luxembourg have the lowest results, and those in Switzerland the highest, while the individual characteristics are very similar between two countries. Those in former Yugoslavia
have on average two years more education and higher occupational status. However, the performance of their children remains lower. Of all immigrant groups, this group is affected most severely by low occupational status of their parents: coming from such a family results in 50 points lower test score. In terms of within-country between-groups comparison, the following tendencies are observed: Different factors affect the schooling results of three immigrant groups in Luxembourg. Certainly, the socio-economic status of parents remains the decisive factor, but more so for Italian- and ex-Yugoslavian-origin students with the effect being three times higher than among Portuguese students. But Portuguese students are the most negatively affected group if they are coming from single-parent households. Also, the largest and the most important reason for low achievement is the type of the educational track. Those in vocational programmes have on average 30 points lower results than those in academic. This does not seem to concern Italian-origin students. For Yugoslavian-origin students the vocational track entails the disadvantage of -14 points, although this does not seem to be the case for the same group in Switzerland. Among the 2nd generation, students with Portuguese background show bigger progress, followed by ex-Yugoslavian students. Italian students, on the contrary, do not show such tendency. In the Swiss case, the effect of socio-economic status has lower effect for Italian and Portuguese students, but clearly not for Yugoslavians. Those coming from low background families and in addition to that, from one-parent families perform nearly 70 points lower than those from highly skilled two-parent families. The cultural capital seems to have a rather marginal importance for Italian-origin students and no effect for two others. Vocational track affects Portuguese students the strongest, its effect is modest among Italian, and absent among Yugoslavian students. There seem to be some improvement for the 2nd generation Portuguese and Italian students, but not within ex-Yugoslavian group. And finally, boys do not master mathematics better than girls, at least not as far as the immigrant population is concerned. The empirical results suggest that there is an interplay between individual- and institutional-level factors in integration processes. The analysis treats the effect of school tracking as an institutional feature and its effect has large negative impact in Luxembourg, for two out of three groups. The puzzle here is to understand the difference between effect in Luxembourg and Switzerland, which
also has tracking system. One of the answers might be the quality of educational programmes. Recent studies show the importance of the teacher’s education and experience both in academic and vocational schools. This might be one of the directions to go in future investigation. With regard to the integration hypotheses, it will be necessary to test various models (i.e. “segmented”, “downward”, “classical” models), for there seem to be a mixture of results and the future studies should analyse other aspects of structural integration, such as labour market participation and social mobility. Given the results outlined above, one might slightly lean towards the segmented assimilation model: the degree of influence of socio-economic variables differs between groups and between countries, and also while some groups show tendency of advancing with duration of residence in the host society, such as Portuguese in this case, other groups do not show such development. However, the following technical note has to be taken into consideration: due to the sample size for immigrant students the results for particular ethnic groups in both countries should be treated cautiously.
Alba, R. and Nee, V. (1999): Rethinking Assimilation Theory for a New Era of Immigration. In The Handbook of International Migration, edited by C. Hirschmann, P. Kasinitz and J. DeWind. New York: Sage, 137-160. Alba, R. D., Handl, J. and Müller, W. (1994): Ethnic Inequalities in the German School System. Mannheim, MZES. Alba, R. D., Lutz, A. and Vesselinov, E. (2001): How Enduring Were the Inequalities Between European Immigrant Groups in the United States? Demography. 38 (3), 349-356. Arnett, S. (2007): Influences of National Education Policies on the Academic Achievement of High- and LowSocial Status Students. Florence, European University Institute, RSCAS Working Papers, 1-35. Baganha, M. I. B., Gois, P. and Pereira, P. (2005): International Migration from and to Portugal: What Do We Know and Where Are We Going? . In: Zimmermann, K. F. (ed.): European Migration: What Do We Know? Oxford, Oxford University Press. Baganha, M. I. B. (2003): From Closed to Open Doors: Portuguese Emigration under the Corporatist Regime. Electronic Journal of Portuguese History. 1 (1), 1-16. Bauer, P. and Riphahn, R. T. (2006): Education and Its Intergenerational Transmission: Country of Origin-Specific Evidence for Natives and Immigrants from Switzerland. Portuguese Economic Journal. 5 (2), 89-110. Benavot, A. (1997): Institutional approach to the study of education. In: Saha, L. J. (Ed.): International encyclopaedia of the sociology of education, Saha, 340-344. Betts, J. R. and Lofstrom, M. (2000): The Educational Attainment of Immigrants: Trends and Implications. In: Borjas, G. J. (ed.): Issues in the Economics of Immigration. Chicago, The University of Chicago Press, 51-117. Boado, H. C. (2007): Immigrant Concentration in Schools: Peer Pressures in Place? European Sociological Review. 23 (3), 341-356. Boyd, M. (2000): Ethnicity and Immigrant Offspring. In: Kalbach, M. A. and Kalbach, W. E. (eds.): Perspectives on Ethnicity in Canada. Toronto, 137-154. Boyd, M. (2002): Educational Attainments of Immigrant Offspring: Success or Segmented Assimilation. International Migration Review. 36 (4), 1037-1060. Boyd, M. and Grieco, E. M. (1998): Triumphant Transitions: Socioeconomic Achievements of the Second Generation in Canada. International Migration Review. 32 (4), 853-876. Brohy, C. (2005): Trilingual Education in Switzerland. International Journal of the Sociology of Language. 171, 133-148. Brosius, J. (2005): La Différence de Demandes Salariales Comme Source des Ecarts de Taux de Salaire Entre Frontaliers et Résidents Luxembourgeois. Differdange, CEPS/INSTEAD, Cahier PSELL. 145, 45. Buchmann, C. and Parrado, E. A. (2006): Educational Achievement of Immigrant-Origin and Native Students: a Comparative Analysis Informed by Institutional Theory. International Perspectives on Education and Society. 7, 345-377.
Buchmann, M., Charles, M. and Sacchi, S. (1993): The Lifelong Shadow: Social Origins and Educational Opportunity in Switzerland. In: Shavit, Y. and Blossfeld, H.-P. (eds.): Persistent Inequality: Changing Educational Attainment in Thirteen Countries. Boulder, Westview Press, 177-192. Chiswick, B. R. (1991): Speaking, reading, and earning among low-skilled immigrants. Journal of Labour Economics. 9 (2), 149-170. Crul, M. (2004, October 29-30): How Do Educational Systems Integrate? Integration of Second Generation Turks in Four Different Institutional Settings. Paper Presented at the Conference: The Next Generation: Immigrant Youth and Families in Comparative Perspective, Radcliffe Institute for Advanced Study, Harvard University. Crul, M. (2005, February 11-12): Integration of Turkish Second-Generation Men and Women in Germany and the Netherlands. The Impact of Differences in Academic Tracking Systems. Paper Presented at the Nuffield/SSRC Workshop: Education and Migration in International Perspective: Pathways to Mobility and Citizenship, London. Crul, M. and Vermeulen, H. (2004): Immigration and Education. The Second Generation in Europe. www.lisproject.org/immigration/papers/crul.pdf Crul, M. and Vermeulen, H. (2006): Immigration, Education and the Turkish Second Generation in Five European Nations: A Comparative Study. In: Parsons, C. A. and Smeeding, T. (eds.): Immigration and the Transformation of Europe. Cambridge, Cambridge University Press, 235-250. Diamantopoulos, A. and Siguaw, J. A. (2000): Introducing LISREL. London, Sage Publications. Dronkers, J. and Levels, M. (2006, May 11–14). Social-Economic and Ethnic School-Segregation in Europe and Australia and Educational Achievement of Migrant-Pupils Coming From Various Regions of Origins. Paper Presented at the Spring meeting 2006 of the ISA Research Committee on Social Stratification and Mobility Intergenerational Transmissions: Cultural, Economic or Social Resources?, Nijmegen, Netherlands. Duran, B. and Weffer, R. (1992): Immigrants’ aspirations, high school process and academic outcomes. American Educational Research Journal. 29 (1), 163-181. Esser, H. (2004): Does the "New" Immigration Require a "New" Theory of Intergenerational Integration? International Migration Review. 38 (3), 1126-1159. Freeman, G. P. (2004): Immigrant Incorporation in Western Democracies. International Migration Review. 38 (3), 945-969. Gibson, M. A. (1988): Accommodation without Assimilation: Sikh Immigrants in an American High School. New York, Cornell University Press. Glenn, C. L. and Jong, d. E. (1996): Educating Immigrant Children: Schools and Language Minorities in 12 Nations. New York, Garland. Gordon, M. (1964): Assimilation in American Life: the Role of Race, Religion and National Origins. New York, Oxford University Press. Gross, D. M. (2006a): Immigration Policy and Foreign Population in Switzerland, World Bank, WPS3853. Gross, D. M. (2006b): Immigration to Switzerland: The Case of the Former Republic of Yugoslavia, World Bank, WPS 3880. Heath, A. F. and McMahon, D. (1997): Education and Occupational Attainment: the Impact of Ethnic Origins. In: Halsey, A. H., Lauder, H., Brown, P. and Wells, A. S. (eds.): Education: Culture, Economy and Society. Oxford, Oxford University Press.
Heckmann, F. and Schnapper, D. (eds.) (2003): The Integration of Immigrants in European Societies. National Differences and Trends of Convergence. Stuttgart, Lucius & Lucius. Kalter, F. and Nadia, G. (2007): Educational Hurdles on the Way to Structural Assimilation in Germany. In: Heath, A. F. and Cheung, S. Y. (eds.): Unequal Chances: Ethnic Minorities in Western Labour Markets. New York, Oxford University Press, 271-320. Kao, G. and Thompson, J. S. (2003): Racial and Ethnic Stratification in Educational Achievement and Attainment. Annual Review of Sociology. 29, 417-442. Kao, G. and Tienda, M. (1995): Optimism and Achievement: the Educational Performance of Immigrant Youth. Social Science Quarterly. 76, 1-19. Kao, G. and Tienda, M. (1998): Educational Aspirations of Minority Youth. American Journal of Education. 106, 349-384. Kogan, I. (2007a): Continuing Ethnic Segregation in Austria. In: Heath, A. F. and Cheung, S. Y. (eds.): Unequal Chances: Ethnic Minorities in Western Labour Markets. New York, Oxford University Press, 103-142. Kogan, I. (2007b): Working Through Barriers: Host Country Institutions and Immigrant Labour Market Performance in Europe. Dordrecht, Springer. Kollwelter, S. (2007): Immigration in Luxembourg: New Challenges for an Old Country. http://www.migrationinformation.org/Profiles/display.cfm?ID=587 [Accessed on: 10.12.2007]. Levels, M., Dronkers, J. and Kraaykamp, G. (1-2 September, 2006). Scholastic Achievement Differences of Immigrants in Western Countries: A Cross-National Study on Origin, Destination, and Community Effects. Paper Presented at the ECSR Conference: European Comparative Studies: Quality and Inequality in Education, Prague, Czech Republic. Lindo, F. (1995): Ethnic Myth or Ethnic Might? On the Divergence in Educational Attainment between Portuguese and Turkish Youth in the Netherlands. In: Baumann, G. and Thijl, S. (eds.): PostMigration Ethnicity. De-Essentializing Cohesion, Commitments and Comparison. Amsterdam, Het Spinhuis Publishers. Lindo, F. (1999): Immigrant Networks, Group-Specific Behaviour and School Attainment of Migrant Youth in the Netherlands. In: Crul, M., Lindo, F. and Pang, C. L. (eds.): Culture, Structure and Beyond: Changing Identities and Social Positions of Immigrants and their Children. Amsterdam, Het Spinhuis, 75-90. Lindo, F. (2000): Does Culture Explain? Understanding Differences in School Attainment between Iberian and Turkish Youth in the Netherlands. In: Vermeulen, H. and Perlmann, J. (eds.): Immigrants, Schooling and Social Mobility. Does Culture Make a Difference? Basingstoke, Macmillan, 206-224. Lindo, F. (2000): The Silent Success: The Social Advancement of Southern European Labour Migrants in the Netherlands In: Vermeulen, H. and Penninx, R. (eds.): Immigrant Integration: the Dutch Case. Amsterdam, Het Spinhuis, 123-152. Mahnig, H. and Wimmer, A. (2003) Integration without Immigration Policy: the Case of Switzerland, in Heckmann, F. and Schnapper, D. (eds), The Integration of Immigrants in European Societies, Stuttgart, Lucius & Lucius Verlagsgesellschaft mbH. Meyer, T. (2003): When being smart is not enough: institutional and social access barriers to upper secondary education and their consequences on successful labour market entry. The case of Switzerland. Competencies and careers. Madeira. Ministère de la Justice (2002): La nationalité luxembourgeoise. Commentaire pratique. Luxembourg, Service Information et Presse, 36.
Ministère de L'Education National et de la Formation Professionnelle (2007): PISA 2006: Rapport National. Luxembourg. OECD (2004): Learning for Tomorrow's World. First Results from PISA2003. Paris, OECD Publishing. OECD (2005): PISA 2003 Data Analysis Manual: SPSS Users. Paris, OECD Publishing. OECD (2006a): Education at a Glance. Paris, OECD Publishing. OECD (2006b): Where Immigrant Students Succeed- a Comparative Review of Performance and Engagement in PISA 2003. Paris, OECD publishing. Parsons, C. and Smeeding, T. (eds.) (2006): Immigration and the Transformation of Europe. Cambridge, Cambridge University Press. Portes, A. (1995): Children of Immigrants: Segmented Assimilation and its Determinants. In: Portes, A. (ed.): The Economic Sociology of Immigration: Essays on Networks, Ethnicity, and Entrepreneurship. New York, Russel Sage Foundation. Portes, A. and Hao, L. (1998): E Pluribus Unum: Bilingualism and Loss of Language in the Second Generation. Sociology of Education. 71 (4), 269-294. Portes, A. and MacLeod, D. (1996): Educational Progress of Children of Immigrants: The Roles of Class, Ethnicity, and School Context. Sociology of Education. 69, 255-275. Portes, A. and MacLeod, D. (1999): Educating the Second Generation: Determinants of Academic Achievement Among Children of Immigrants in the United States. Journal of Ethnic and Migration Studies. 25, 373-396. Portes, A. and Rumbaut, R. G. (1996): Immigrant America: A Portrait. Berkeley, California Press. University of
Portes, A. and Zhou, M. (1993): The New Second Generation. Segmented Assimilation and its Variants among Post-1965 Immigrant Youth. Annals of the American Academy of Political and Social Science. 530, 74-98. Schmid, C. L. (2001): Educational achievement, language-minority students, and the new second generation. Sociology of Education. 74 (Current of thought: sociology of education at the dawn of the 21st century), 71-87. Schnepf, S. V. (2006): How Different Are Immigrants? A Cross-Country and Cross-Survey Analysis of Educational Achievement. In: Parsons, C. and Smeeding, T. (Ed.): Immigration and the Transformation of Europe. Cambridge, Cambridge University Press. Silberman, R. and Fournier, I. (2007): Is French Society Truly Assimilative? Immigrant Parents and Offspring on the French Labour Market. In: Heath, A. F. and Cheung, S. Y. (eds.): Unequal Chances: Ethnic Minorities in Western Labour Markets. New York, Oxford University Press, 221-270. Soysal, Y. N. (1994): Limits of Citizenship: Migrants and Postnational Membership in Europe. Chicago & London, University of Chicago Press. Unsen, M., Vallado, D. and Barthelemy, M. (2005): Etude sur les élèves quittant prématurément nos écoles. Luxembourg, Ministère de l'Education National et de la Formation Professionnelle. Vallado, D. (ed.) (2003): Les Chiffres Clés de L’Education Nationale. Statistique et Indicateurs 2001-2002. Luxembourg, Ministère de l’éducation nationale de la formation professionnelle et des sports, SCRIPT. Van Ours, J. C. and Veenman, J. (2001): The Educational Attainment of Second Generation Immigrants in The Netherlands. Bonn, The Institute for the Study of Labour (IZA), Discussion Paper No. 297. 29
Waringo, K. Migration Policies in Luxembourg. http://www.emzberlin.de/projekte_e/pj32_1pdf/MigPol/MigPol_Luxembourg.pdf [Accessed on: 25.10.2005]. Waringo, K. Minorities and Nation-State Building in Luxembourg. berlin.de/projekte_e/pj50_pdf/luxembourg.pdf [Accessed on: 25.10.2005]. http://www.emz-
Zhou, M. and Bankston, C. L. (1998): Growing Up American: How Vietnamese Children Adapt to Life in the United States. New York Russell Sage Foundation. Zimmermann, K. F. (ed.) (2005): European Migration: What Do We Know? Oxford, Oxford University Press.
Table 12. Competence levels in PISA 2003
Below Level 1 Level 1 Level 2 Level 3 Level 4 Level 5 Level 6
Below 358.3 358.3-420.4 420.4-482.4 482.4-544.4 544.4-606.6 606.6-668.7 668.7 and higher
Each level of proficiency is associated with various skills and abilities of students (OECD 2004:47).
At Level 1, students can answer questions involving familiar contexts where all relevant information is present and the questions are clearly defined. They are able to identify information and to carry out routine procedures according to direct instructions in explicit situations. They can perform actions that are obvious and that follow immediately from the given stimuli. At Level 2, students can interpret and recognise situations in contexts that require no more than direct inference. They can extract relevant information from a single source and make use of a single representational mode. Students at this level can employ basic algorithms, formulas, procedures or conventions. They are capable of direct reasoning and of making literal interpretations of the results. At Level 3, students can execute clearly described procedures, including those that require sequential decisions. They can select and apply simple problem-solving strategies. Students at this level can interpret and use representations based on different information sources and reason directly from them. They can develop short communications reporting their interpretations, results and reasoning. At Level 4, students can work effectively with explicit models for complex concrete situations that may involve constraints or call for making assumptions. They can select and integrate different representations, including symbolic ones, linking them directly to aspects of real-world situations. Students at this level can utilise well-developed skills and reason flexibly, with some insight, in these contexts. They can construct and communicate explanations and arguments based on their interpretations, arguments and actions. At Level 5, students can develop and work with models for complex situations, identifying constraints and specifying assumptions. They can select, compare, and evaluate appropriate problem31
solving strategies for dealing with complex problems related to these models. Students at this level can work strategically using broad, well-developed thinking and reasoning skills, appropriately linked representations, symbolic and formal characterisations, and insight pertaining to these situations. They can reflect on their actions and can formulate and communicate their interpretations and reasoning. At Level 6, students can conceptualise, generalise, and utilise information based on their investigations and modelling of complex problem situations. They can link different information sources and representations and flexibly translate among them. Students at this level are capable of advanced mathematical thinking and reasoning. These students can apply this insight and understanding, along with a mastery of symbolic and formal mathematical operations and relationships, to develop new approaches and strategies for attacking novel situations. Students at this level can formulate and precisely communicate their actions and reflections regarding their findings, interpretations, arguments, and the appropriateness of these to the original situations.
Table 13. Variables in the structural equation modelling Latent variable Socioeconomic, educational level of parents Observed variable Educational level of mother – low Educational level of mother - high Educational level of father - low Educational level of father - high Mother’s status – blue collar Mother’s status – white collar Father’s status – blue collar Father’s status – white collar Possession of literature Possession of poetry Possession of art Foreign language spoken at home Single parent family Immigrant’s generation General/academic track Vocational track Student’s gender Mathematics Reading Science Variable name m_ed_low m_ed_high f_ed_low f_ed_high m_blue m_white f_blue m_white literatu poetry art lang sing_par 2nd_gen gen_edu voc_edu male female math read scie Measurement Dummy Dummy Dummy Dummy Dummy Dummy Dummy Dummy Dummy Dummy Dummy Dummy Dummy Dummy Dummy Dummy Dummy Dummy Continuous Continuous Continuous Coding 0;1 Reference group 0;1 Reference group 0;1 Reference group 0;1 Reference group 0;1 0;1 0;1 0;1 0;1 0;1 0;1 Reference group 0;1 Reference group
Language Single parent 2nd generation Study programme Gender Achievement