Catching Up? Country Studies on Intergenerational Mobility and Children of Immigrants

Chapter 4. The Netherlands: Intergenerational mobility of native-born children of immigrants from Morocco and Turkey

Maurice Crul

Abstract

Using research reports from the Dutch Social Cultural Planning Bureau (SCP) and data from The Integration of the European Second Generation (TIES) Survey, this chapter compares the intergenerational social mobility of the offspring of immigrants and their parents for the two most disadvantaged ethnic groups in the Netherlands. It follows the school and labour market careers of the native-born children of Turkish and Moroccan descent, describing outcomes at various stages and noting differences with peers of Dutch descent. Attempting to ascertain what produces the stark polarisation within this group – whereby some enjoy exceptionally steep mobility while others stay behind – the chapter points to the role played by the complex policies and institutional arrangements of the country’s educational system. It goes on to discuss how educational outcomes translate into labour market outcomes, highlighting striking gender differences. Finally, it shows how the phenomenon of the “multiplier effect” can help children of less educated immigrants be successful against all odds.

Main findings

• The educational attainment of the native-born children of immigrants in the Netherlands varies across parental origin groups. Children of Turkish descent are three times more likely than the children of Antillean descent to reach only primary education (9% and 3% respectively), and 4 points more likely than children of Surinamese or Moroccan descent.

• The overall better educational results for the children of Antillean immigrants are for the most part explained by the much better socio-economic position of their parents, of whom two-thirds hold a secondary vocational certificate or tertiary degree.

• The most striking feature of the outcome of both the children of Turkish and Moroccan immigrants is the polarisation within each of these two groups. Among the children of Turkish immigrants, 27% are in higher education while an almost equal share (28%) are early school leavers. Children of Moroccan descent show a similar polarisation.

• A portion of the native-born children of Turkish and Moroccan immigrants have thus demonstrated a steep social upwards mobility compared to their parents. Especially women have often accomplished a remarkable social rise compared to their largely uneducated immigrant mothers. In fact, the daughters of Moroccan and Turkish immigrants have overtaken sons in almost all higher-level streams.

• Although preschool and its emphasis on second language learning are a key pillar in the Dutch policy framework, participation in it remains voluntary and about one quarter of children “at risk” are not reached. Furthermore, the separation of students who need language support from those who don’t may contribute to a persisting division.

• The streaming recommendation that is made by schools at the end of elementary school reveals particular hurdles for the children of immigrants: it relies heavily on Dutch language ability and involves a discussion with the pupil’s parents in which middle- and upper class families can use their social and cultural capital to secure the desired recommendation for their children. In contrast, immigrant parents are often less informed or less able to influence the school’s recommendation.

• A little more than half of the children of Turkish and Moroccan immigrants responding to the TIES Survey in universities of applied sciences took the “long route” into higher education – meaning that they moved up gradually through a system of vocational education. The long route is one of the most crucial success factors in the Dutch school system for the upward mobility of disadvantaged children.

• Unemployment is frequent among early school leavers with immigrant parents. One and a half years after drop out, the unemployment rate reaches 54% for the children of Moroccan immigrants and 31% for the children of Turkish immigrants. In both cases, there is a large difference with early school leavers of Dutch descent (18%).

• Among the highly educated offspring of immigrants, more than one quarter works in a job below their skill level. Yet at the same time, more than 40% of early school leavers with Turkish of Moroccan parents manage to work in a white collar or executive profession. This finding points towards considerable intra-generational upward mobility and the lower end of the occupational spectrum but a ceiling at the higher end.

Introduction

This chapter discusses the intergenerational mobility of Turkish and Moroccan immigrants’ offspring in comparison to their parents, who came to the Netherlands through labour migration and marriage migration in the 1970s and 1980s. With 389 000 people of Moroccan descent and about 400 000 people of Turkish descent living in the Netherlands, the two groups are almost equal in size (SCP, 2016a). The two are also the most disadvantaged ethnic groups in the Netherlands, showing relatively low levels of education and high levels of unemployment. These outcomes result mainly from the immigrant parents’ socio-economic situation, arising from the fact that the men recruited for work in the Netherlands came from the poorest and least developed regions in both Morocco and Turkey (Crul, 2000). In the case of Morocco, about two-thirds of the “guest workers” came from the Rif region in the north of the country. Education beyond elementary school was still largely absent in the Rif area. In central Anatolia, the region where most labour migrants from Turkey originate, the schooling situation was slightly better; a somewhat larger share of Turkish parents attended primary school, and some of the men also had a few years of secondary education. In the Netherlands, the initial situation of the Moroccan immigrants was comparatively the most difficult (Crul, 2000).

The three other large immigrant groups in the Netherlands came from former Dutch colonies: Indonesia, Surinam and the Netherlands Antilles. The people from Indonesia, mainly of mixed Dutch-Indonesian heritage, are considered a success among immigrant groups, and as a result are not targeted in policies or in research. The two other groups, however, have been subjects of research for the past four decades. The fact that in Surinam and the Antilles the curriculum followed was Dutch, and also taught in Dutch, has been an advantage for the integration of those who migrated into Dutch society. The parents of those of Antillean descent often came to the Netherlands to study and then stayed. More recent Antillean immigrants often come from the lowest socio-economic segments of the island population. Younger cohorts of Antillean descent consequently experience more negative outcomes, also scoring below the school outcomes of students of Moroccan or Turkish descent. As a result, the attention paid that group has been growing over the past decade, both in research and in policies.

The chapter makes use of research reports from the Social Cultural Planning Bureau (SCP) because these, on a biannual basis, provide detailed overviews of the school and labour market positions of the two topic groups. Apart from the SCP reports, the paper utilises research from scholars who are experts either on education or on the labour market in the Netherlands. Much of that research, like the SCP reports, is commissioned by the Dutch government because of political concerns about the disadvantaged position of the two groups. Lastly, the paper uses the author’s research on the children of immigrants in Europe, which includes the largest survey on those of Turkish and Moroccan descent carried out in the Netherlands. This survey also contains information on the parents, and is therefore best suited to assess intergenerational mobility.

The chapter follows the school and labour market careers of the Turkish and Moroccan Dutch, describing outcomes at different stages. In doing so it reveals how the differences between this group and the Dutch of native descent come about, and whether gaps in outcomes become smaller over time, or not. The discussion will point to a key characteristic of intergenerational mobility: polarisation within these offspring of immigrants – with one portion, of considerable size, demonstrating exceptional steep mobility and another, almost equal in size, staying behind. This chapter will try to ascertain what produces such stark polarisation in this generation and, also and even more so, in the generation that follows. This places the focus on policies and institutional arrangements in education and (the transition to) the labour market that seem to have benefited some more than others.

Box 4.1. The Integration of the European Second Generation (TIES) Survey

The TIES Survey (Crul, Schneider and Lelie, 2012) is the first systematic collection of data on the children of immigrants from Turkey and Morocco (as well as the former Yugoslavia) in 15 European cities inside 8 countries. In total, almost 10 000 people were interviewed. The participating countries are Austria, Belgium, France, Germany, Spain, Sweden, Switzerland and the Netherlands. This chapter uses data from the Dutch portion of the TIES Survey for the native-born children of Turkish and Moroccan immigrant parents from Rotterdam and Amsterdam. In the survey, the term “second generation” refers to children of immigrants who have at least one parent born in Turkey or Morocco; who were themselves born in the survey country; and who have had their entire education there. (Almost all respondents had two parents with the same national background.) At the time of the interviews, the respondents were between 18 and 35 years old. The total sample size was N=1000, with 500 respondents having Turkish and 500 respondents having Moroccan immigrant parents. Both groups had 250 respondents per city.

The TIES Survey is a sound source of information for studying intergenerational mobility, because of its focus both on the children of immigrants and on their parents. With about a thousand young adults in the former category, the Dutch portion of the TIES Survey is the largest study on this group in the Netherlands. Because the survey is conducted in two cities only – Amsterdam and Rotterdam – the study is not representative of the entire Turkish and Moroccan Dutch population in the Netherlands. Nonetheless, a comparison between the outcomes of the TIES survey with those of national surveys shows no substantial differences (Crul and Heering, 2008; Groenewold, 2008).

The immigrant parents

The Turkish and Moroccan parents of offspring born in the Netherlands are difficult to identify in most of the national survey data and reports. Because migration from both Morocco and Turkey is ongoing, a new cohort of immigrants is constantly merging with the earlier one whose children are the subject here. Today’s immigrants had a different educational career than the wave that came in 1970s and 1980s, because since then Turkey and Morocco have opened their educational system to people in rural areas. Those who arrived in the 1970s and 1980s did not have these same possibilities, and were in general very little educated (Crul and Heering, 2008).

The TIES survey shows that most immigrant fathers came to the Netherlands in the 1970s, and the mothers followed on average five to six years later. According to the survey, about 95% of the fathers and mothers of the respondents were both born in either Turkey or Morocco, showing that there were hardly any inter-ethnic marriages. About two-thirds of the parents have dual citizenship and about a third of them hold only citizenship of either Turkey or Morocco (Beets, ter Bekke and Schoorl, 2008). At the moment of the survey, in 2008, the majority of the parents had already lived thirty years or more in the Netherlands. This explains why the TIES respondents (all born in the Netherlands) over thirty were still a rather small group. Two-thirds of the Moroccan respondents lived in households with more than five persons. The Turkish households were somewhat smaller on average; only one-third of the Turkish respondents lived in a household with five or more persons.

The educational level of the immigrant parents was generally very low (see Table 4.1). For instance, over one-third of the mothers did not go to school at all. About half of all parents only went to elementary school at the most. The number of parents that attended tertiary education is very small indeed. The information about the education of the parents is derived from the interviews with the offspring respondents. As is clear from the category “Do not know” in the table, not all respondents were aware of the education their parents had in their home country. This is a limitation of the data within TIES.

Table 4.1. Educational level of the immigrant parents of native Dutch young adults between the ages of 18 and 35, percentages

	Turkish immigrant fathers	Turkish immigrant mothers	Moroccan immigrant fathers	Moroccan immigrant mothers
No schooling	7	17	21	40
Only Quran school	1	0	9	2
Elementary school	41	47	23	21
Lower secondary school	22	19	11	14
Upper secondary school	13	5	12	9
Tertiary education	6	3	4	1
Do not know	11	9	20	14

Source: TIES survey 2008.

For the majority of Turkish and Moroccan parents, education did not continue beyond elementary school. In contrast, in the comparison group of Dutch descent, only 7% of the fathers and 6% of the mothers did not continue beyond elementary school. In the Netherlands, this is a very small, atypical group; having no education beyond elementary school most likely indicates severe problems experienced during schooling. Looking at these three groups, i.e. Turkish, Moroccan and Dutch parents with very little education signals a need to treat comparative outcomes based on the education level of parents with great care. Low levels indicate different things for the three groups.

The TIES Survey also provides data on the labour market position of the parents before their migration, as well as when the offspring respondent was fifteen and at the time of the survey. Those data show that about 80-85% of the mothers did not have a paid job prior to their emigration. The share of Turkish working women was slightly higher than that of Moroccan working women. More than half of the men, on the other hand, were working in a paid job before their migration to the Netherlands. Three-quarters of those active were working in the lowest levels of the labour market in Turkey or Morocco. At age fifteen of the offspring respondent, about two-thirds of the Turkish fathers had paid jobs or were self-employed, as were a little over half of the Moroccan fathers. About 80% of the working fathers were doing unskilled work. One of the remarkable findings is that about one in five male immigrant men were in disability schemes in the Netherlands when the respondents were fifteen years old. This practice was common in the Netherlands: in case factories had to close or reduce the number of employees, as a political solution people were often not fired but admitted into disability schemes. For many, this would mean the end of working life. In comparison, only about 5% of the men were registered as being unemployed at that point. The majority of immigrant women were not working when their child was fifteen. Only one in five immigrant Turkish mothers and one in six immigrant Moroccan mothers were working.

Most immigrant males were not mobile during their working life, but stuck at the bottom of the labour market in unskilled jobs. At the time of the survey most fathers had left the labour market, either because they were retired or because they lived from other benefits they received. Only a third of the Turkish immigrant men and about a quarter of the Moroccan immigrant men were still working.

This situation, with only one parent working in a low-level job or on benefits – in combination with large households – usually created a precarious income situation for the children of immigrants to grow up in. In educational, labour market and income terms, immigrant parents are at the bottom of the socio-economic hierarchy in the Netherlands. As a result, it is hard for their children to be anything but upwardly mobile by comparison.

Education and the children of immigrants

School outcomes of the children of immigrants

Every other year, the National Social Cultural Planning Bureau (SCP) publishes an update of school and labour market outcomes for the four largest immigrant groups in the Netherlands: people of Turkish, Moroccan, Surinamese and Antillean descent (SCP, 2014a, 2016a). Other incidental reports of the SCP add important information, like reports addressing the emancipation of women, or separate reports addressing labour market outcomes (CBS/SCP, 2016; SCP, 2014b). The reports of the SCP usually distinguish between immigrants and persons born in the Netherlands with one or both parents born abroad, and between men and women, and compare outcomes with pupils of Dutch descent (that is, born in the Netherlands with both parents also born in the Netherlands).

The school outcomes documented by SCP show a rather consistent picture for the four migrant groups over time. As shown in Table 4.2, the school outcomes for the children of Turkish and Moroccan immigrants are very similar, and the same is true for the children whose parents came from the two former Dutch colonies. The latter two groups leave school early less often, and are more frequently found in higher education.

Table 4.2. School level of the children of immigrants between the ages of 15 and 65 in the Netherlands, percentages

	Elementary school only	Lower vocational education	Secondary vocational education	Tertiary education
Turkish descent	9	19	45	27
Moroccan descent	5	19	46	30
Surinamese descent	5	13	46	36
Antillean descent	3	12	45	41

Source: Centraal Bureau voor de Statistiek/Social Cultural Planning Bureau (Survey integratie migranten 2015).

The relatively better educational results for the Antillean children of immigrants are for the most part explained by the much better socio-economic position of their parents. Two-thirds of the immigrant Antillean parents hold a secondary vocational certificate or tertiary degree. Many early Antillean immigrants came to the Netherlands for study reasons (van Niekerk, 2004). Since there are no universities on the Netherlands Antilles, people came and continue to come to the Netherlands to study. As a result, the first wave was something of an elite migration. However, since the 1990s, the migration has become increasingly diverse and more representative of the general population of the islands. The children of the more socio-economically diverse group now begin to appear in the migration statistics as the younger cohort of the Antillean children of immigrants.

Surinamese migration has always been more representative of the general population. For both groups, an important factor is language. Although many immigrants from both Surinam and the Antilles speak another local language at home, they also speak Dutch. For most of those immigrating, Dutch has been the instruction language in school, which gives them a huge advantage over the native Dutch with parents from Turkey and Morocco. Differences in outcome are thus indeed substantial between the two subgroups (labour migrants versus those from former colonies), but both class and language help to explain these differences. The school outcomes for the children of Surinamese and Antillean immigrants are almost on par with people of Dutch descent. A proper comparison is difficult to make however, because immigrant offspring are still concentrated in the younger age cohorts. In the group of Dutch descent aged between 15 and 64, many more people are in the older cohorts and were educated just after the Second World War, when educational opportunities where far fewer. This brings down the average education level of the whole group.

The most striking feature of the outcome of both the children of Turkish and Moroccan immigrants is the polarisation within each of these two groups. Among the children of Turkish immigrants, 27% is in higher education while an almost equal share (28%) has a lower vocational education certificate at the most, officially making them early school leavers. The former group has made a huge upward mobility leap compared with their parents, while the latter has more or less reproduced the low status of their parents. A similar trend is evident with the offspring of Moroccan descent. The polarised outcome here is even more striking, because the Moroccan parents are almost unanimously very low educated. As shown in Tables 4.3 through 4.8, socio-economic status does not forecast school outcomes in the same way for the children of immigrants as it does for people of Dutch descent (see also Holdaway, Crul and Roberts, 2009).

The SCP integration report 2016 from which these figures stem (Herweijer, Iedema and Andriessen, 2016) does not show school outcomes for men and women separately, but the report on women’s emancipation from the Dutch Central Bureau of Statistics (CBS) and the Social Cultural Planning Bureau (CBS/SCP, 2016) does provide information on gender differences at different school levels. The results show how the daughters of Moroccan and Turkish immigrants have overtaken the sons in almost all higher-level streams. The educational success of women has important consequences for intergenerational mobility, which will be discussed in more detail further on. In short, if women are successful in school, their success is a reason to be allowed to postpone marriage. If they can postpone marriage until after finishing higher education, they will then be old enough to make their own choices. This involves choosing their own partner, who is then more often also born or raised in the Netherlands and is also more highly educated. With more highly educated couples, women more often enter the labour market.

The turning point for the women overtaking the men in educational success came a bit earlier in time for women of Moroccan descent than for those of Turkish descent. This is usually attributed to the fact that the greater social control in the Turkish community initially resulted in pressure on young girls to marry early, often resulting in them leaving school early (Crul, 2000; Crul, 2010; Crul and Doomernik, 2003; de Vries, 1987; Lindo, 1996). Nowadays it is the reverse, with the younger men ceasing their studies early and the young women continuing with theirs (CBS/SCP, 2016). The trend of women doing better in education found nowadays partly stems from the experiences of the older cohort of daughters of immigrants. For some of these women, marrying young did not work out and then, because of their lack of schooling, they were unable to find employment and support themselves. This made people in the communities aware that education is also important for women. Secondly, the smaller group of women in the older cohorts who did continue to study proved that one could be successful in school and be a respectable woman at the same time. These women became role models for the younger women, and opinions on the importance of education for women gradually changed as a result (Crul, 2010).

The differences in school outcomes for the four biggest ethnic groups compared to people of Dutch descent can therefore partly be explained by age, gender or the educational level of the parents. When corrected for those three variables only half of the gap remains for the children of Turkish immigrants and only a quarter for the Moroccan offspring. Over a ten-year period, the gap has been reduced by two-thirds for the children of Turkish and five-sixths for the children of Moroccan immigrants. This shows that if we take background characteristics into consideration the gaps are no longer sizeable.

The educational pipeline

The figures from the SCP report 2016 (Herweijer, Iedema and Andriessen, 2016) only show final school outcomes. By focusing on entire school careers – starting from preschool onwards – the next paragraphs will show in more detail how these final results have actually come about.

Box 4.2. The Dutch educational system

The vast majority of Dutch natives with immigrant parents attend school in the Dutch system, one of the most complicated school systems in Europe (EP & Nuffic, 2015).

Pupils enter primary school in the Netherlands at the age of four. In the Dutch educational system, one is free to choose the school of one’s religious or ideological preference at no extra cost. Schools all teach the same curriculum, and school fees are very low or non-existent.

Primary school consists of eight grades. During the final two years, pupils take national examinations crucial in shaping their further school career. Based on the test results and the recommendation or “advice” of the teacher, they will be assigned to different streams in the secondary school system. As children usually leave primary school at age 12, streaming in the Netherlands happens relatively early on.

In secondary school, there are six different potential streams: two prepare for higher education, and four prepare for different levels of vocational education. The two pre-academic streams, HAVO (higher general secondary education, five years) and VWO (preparatory scholarly education, six years) prepare for a university of applied sciences (HBO) and a research university (Universiteit) respectively. The four vocational streams all take four years and cover different cognitive and skill levels. Praktijk onderwijs, the lowest level, caters to students with severe learning and/or behavioural problems. Many of the students in this stream receive special assistance. VMBO-basis is also a very basic level of vocational education. Both levels only give access to a two-year MBO-2 level (middle-level applied education) of secondary vocational education at age sixteen. MBO-2 prepares students for unskilled labour and is considered the very minimum level with which one can enter the labour market (a startkwalificatie). VMBO-KB (preparatory middle-level vocational education, middle management-oriented learning path) and VMBO TL (theoretical learning path) are somewhat more prestigious and prepare for the middle-level secondary vocational streams MBO-3 (three years) and MBO-4 (four years). After completing MBO-4, students can continue studying higher vocational training in HBO. An important characteristic of the Dutch school system is that pupils can move from one stream to the other with relative ease. They can, for instance, start at VMBO and move up step by step via MBO to HBO and a Universiteit, taking what is called the “long route” through the educational system. Doing so takes up to three years longer, but it is the route chosen by many children of immigrants.

Preschool

One of the main Dutch policy instruments combating educational inequalities is directed at preschool. With the aim to reduce Dutch language deficits from 2000 onwards, the government launched policy arrangements for VVE (preschool and early school education, Voor- en vroegschoolse educatie). VVE policies especially target children of immigrants from disadvantaged families (Driessen, 2012; Inspectie van het Onderwijs, 2010; Jepma, Kooiman and van der Vegt, 2007; Onderwijsraad, 2014; van Tuijl and Siebes, 2006; Veen, Roeleveld and Leseman, 2000; Veen, van der Veen and Driessen, 2012). The emphasis on second language learning at a young age has been embraced by all Dutch governments in power over the past two decades (Onderwijsraad, 2014). It is often put forward as the solution to closing the educational gap. However, the Dutch school arrangements for preschool make it difficult to deliver on the aim of closing that gap. In the Netherlands, preschool is not part of compulsory school; participation is voluntary and comes at a cost. Therefore, in recent years much of the policy attention has been directed at raising the percentage of children of immigrants attending VVE preschool facilities. Efforts have been largely successful, but a considerable group (about a quarter) of potential “at risk” pupils are not reached (Veen, van der Veen and Driessen, 2012).

A further impediment to closing the gap is the very way preschool is designed in the Netherlands. The general policy is that children of immigrants and children of disadvantaged families attend a separate provision called voorschool three half-days a week (between 10 and 12 hours in total). With few exceptions, in practice this means that in the cities, children of immigrants are placed together in preschool, separately from children of Dutch descent whose first language is Dutch. The children of Dutch descent go to the crèche or peuterspeelzalen three to five full days a week (between 24 and 40 hours). In the voorschool, the preschool especially designed for children from disadvantaged families, much attention is paid to second language learning using specialised methods (Leseman and Veen, 2016).1

Elementary school

The largest portion of the extra budget for children of immigrants in education is spent in elementary school. In the past, schools received almost twice as much funding for a child of less educated immigrant parents as for a child of middle or upper class native Dutch parents. In 2007 this funding system was replaced by a weighting system based solely on the educational level of the parents (Claassen and Mulder, 2011). Still, schools with a high number of immigrant children receive considerable extra funding. The apportioning of this extra money is largely determined by the schools themselves, or by the school’s umbrella organisations. Most schools use the money to hire extra teaching staff and to reduce the number of pupils per class. Many schools also have invested in second language teaching. Although it is difficult to prove a direct relation between interventions and results, in general the school results show that gaps between the children of immigrants and pupils of native descent are indeed reduced during the elementary school period (Roeleveld et al., 2011).

Selection and streaming in secondary school

School advice at the end of elementary school which is partly based on a national test (the Cito test) is crucial to a pupil’s opportunities after elementary school (see Box 4.2). A point often raised is that the national test relies heavily on Dutch language capability. Since the language gap is not entirely closed during elementary school, this is an important issue in validating the national test score as a proper instrument to assess a pupil’s cognitive skills (Driessen, 2012; Fettelaar, Mulder and Driessen, 2014; Roeleveld et al., 2011). Dutch language capabilities thus have an important impact on streaming. During grade 8, all pupils – and their parents – receive what is called “school advice”, which is based on the national Cito test score and on the opinion of the teacher. In practice this is not just advice: it has important consequences for the type of school the pupil can be admitted to. The advice takes the form of an official document for the secondary school; it is given during a parent-teacher meeting to discuss the Cito test score and, more generally, the development and attitude of the pupil over the years. This introduces a subjective element into the pupil’s progression at the end of primary school. Parents can try to secure admission to a higher stream in secondary school than initially advised. The current advising process and the subsequent choice of and negotiation with secondary schools offer an opening especially for middle- and upper class families to use their social and cultural capital to secure a better outcome than most working class and immigrant families are able to do. In a report of the Onderwijsinspectie (Inspectorate of Education) (Inspectie van het Onderwijs, 2016), children with the same middle-level Cito test scores were followed over time in their trajectories in secondary school and beyond. Half of the children with highly educated parents would start in a pre-academic track in secondary school, while only a quarter of the children with less educated parents did. Further down the road, 55% of the children with highly educated parents had obtained a degree from tertiary education, while this was true for only 26% of the children with less educated parents, even with the same test results.

The long route into higher education

A special feature of the Dutch school system is that a student can climb the educational ladder by using different levels of vocational education as steppingstones. They can start at the lowest vocational level at age twelve, but in principle they can move from lower vocational education to secondary vocational education to higher vocational education to acquire a bachelor’s degree. Of course in practice this it is not at all easy, and students need a lot of stamina because it takes three years longer than the direct route from the pre-academic HAVO stream (higher general continued education) to higher vocational education at HBO (universities of applied sciences). A bit more than half of the children of Turkish and Moroccan immigrants responding to the TIES Survey in universities of applied sciences have indeed taken this “long route” (Crul, Schneider and Lelie, 2012), proving that this option is a crucial success factor of the Dutch educational system. The long route repairs much of the damage that early streaming and the convention of school advice cause for children of immigrants and children of less educated parents. It offers an important contribution to intergenerational upward mobility.

These four aspects – special arrangements for preschool, early selection, school advice and the possibility to move up in the school system through the long route – are the main characteristics of the school system that shape school outcomes for children of immigrants in the Netherlands. In the past two decades, all four of these key features underwent crucial changes that have impacted school results, both positively and negatively. Policies to include more children of immigrants in preschool had a slight positive effect, and over the years elementary schools have become better equipped to work with this group. Eliminating intermediary classes while retaining the school advice process has widened gaps in secondary school. The long route fortunately offers the possiblility of repairing a lot of the inequalities built into the first part of the school system. However, this long route is under pressure because of a change in the funding system for higher education, and because higher education institutions are erecting more and more barriers for students entering through the long route. The fact that the long route is one of the most crucial success factors in the Dutch school system for the upward mobility of disadvantaged children asks for careful monitoring and a more fundamental rethinking of the school system.

Intergenerational mobility through education

The information from the SCP reports used here is usually based on cohort studies reporting on national outcomes between ethnic groups and national outcomes over time. As shown, these studies do control for parental education, but they do not show patterns of intergenerational mobility for different ethnic groups. To study intergenerational mobility, the chapter makes use of the TIES Study. In this study respondents were asked a number of questions about their parents’ education and socio-economic position. This allows examination of patterns of intergenerational mobility through education.

The educational level of the Turkish fathers and mothers correlates significantly with that of their sons (p<0.01) and daughters (p<0.05). The significant relations are however much less strong than is the case among the respondents of Dutch descent. The most important explanation for a lack of transferability of parental capital to the children of immigrants is that parents, even when they studied in Turkey or Morocco, cannot transfer their knowledge to a Dutch context. They do not have necessary information about the complicated Dutch school system, and because of a lack of Dutch language capability, helping their children with homework and speaking with teachers prove difficult. In the case of the intergenerational educational mobility of the children of Moroccan immigrants compared to the immigrant parents, there are not even any significant outcomes. That the Moroccan immigrants have more homogeneously lower levels of education than the Turkish immigrants results in less variation in education levels, which makes it also more difficult to find statistically significant differences.

Table 4.3. The educational level of Turkish immigrant fathers and their native-born sons, percentages

	Sons
Fathers	Early school leavers	Middle level	Higher education
Primary or less	36	38	26
Secondary	18	43	39
Higher education	23	12	65

Source: TIES survey 2008.

Table 4.4. The educational level of Turkish immigrant mothers and their native-born daughters, percentages

	Daughters
Mothers	Early school leavers	Middle level	Higher education
Primary or less	25	54	21
Secondary	15	52	33
Higher education	0	20	80

Source: TIES survey 2008.

Table 4.5. The educational level of Moroccan immigrant fathers and their native-born sons, percentages

	Sons
Fathers	Early school leavers	Middle level	Higher education
Primary or less	21	51	28
Secondary	13	52	35
Higher education	11	67	22

Source: TIES survey 2008.

Table 4.6. The educational level of Moroccan immigrant mothers and their native-born daughters, percentages

	Daughters
Mothers	Early school leavers	Middle level	Higher education
Primary or less	18	48	34
Secondary	11	52	37
Higher education	0	100	0

Source: TIES survey 2008.

In the case of respondents of Dutch descent, we see a much stronger correlation between the education of both the mothers and their daughters and the fathers and their sons.

Table 4.7. The educational level of native Dutch fathers and their sons, percentages

	Sons
Fathers	Early school leavers	Middle level	Higher education
Primary or less	20	20	60
Secondary	10	34	56
Higher education	2	18	80

Source: TIES survey 2008.

Table 4.8. The educational level of native Dutch mothers and their daughters, percentage

	Daughters
Mothers	Early school leavers	Middle level	Higher education
Primary or less	21	46	33
Secondary	9	25	67
Higher education	2	15	83

Source: TIES survey 2008.

Labour market outcomes of the native-born children of immigrants

Labour market outcomes: Participation, unemployment, income

How do the differences in educational outcomes translate into labour market outcomes? The chapter now addresses the outcomes for three different school-level groups: early school leavers; those with a secondary vocational certificate; and those with a higher education degree. Data are based on the TIES Survey and additional information from SCP reports; the latter are focused on the situation following the impact of the financial crisis on the labour market position of the children of immigrants.

Beginning with the data from the TIES Survey, Table 4.9 shows the participation rate of the two target groups in the labour market. These data were collected just before the financial crisis of 2008.

Table 4.9. Participation in the labour market of children of Moroccan and Turkish immigrants in relation to final educational outcomes, ages 18-35, percentages

	Early school leavers	Secondary vocational education	Higher educated
Children of Moroccan immigrants	70	73	79
Children of Turkish immigrants	70	87	98

Source: TIES survey 2008.

In general, the table indicates that the more highly people are educated, the more likely they are to participate in the labour market. If the outcomes are split according to gender, the results especially show differences for the early school leavers. Of the female early school leavers among the children of Turkish immigrants, only 52% participate in the labour market, compared to 89% of the male early school leavers. For the children of Moroccan immigrants, the corresponding figures are 43% for daughters and 94% for sons. So, about half of the daughters of immigrants who left school early do not participate on the labour market, reproducing the traditional gender pattern of their mothers. Among their more highly educated peers however, comparison does not show a significant difference between men and women. For the middle group (secondary vocational education, community colleges) there is a sizeable difference between the female and male offspring of Turkish immigrants (73% of the women participate versus 100% of the men) and a smaller difference among the children of Moroccan immigrants (80% versus 94%, respectively). Also in the labour market, gender roles seem to change more slowly in the Turkish community. The results also indicate that gender differences among the children of immigrants become considerably larger when people are less educated.

The next important indicator is unemployment, for which once again the figures are split for the three final school outcomes. For those educated at the lowest and middle levels there is little difference in unemployment rates, but for the highly educated the pattern differs between the Moroccan and Turkish children of immigrants. The numbers of those more highly educated who are already participating in the labour market are, however, small; thus caution must be exercised when interpreting the differences between the two groups. For the same reason, outcomes are not split according to gender.

Table 4.10. Unemployment rates of children of Moroccan and Turkish immigrants in relation to final educational outcomes, ages 18-35, percentages

	Early school leavers	Secondary vocational education	Higher educated
Children of Moroccan immigrants	14	13	17
Children of Turkish immigrants	18	17	3

Source: TIES survey 2008.

The SCP report of 2013 offers a more up-to-date picture of the situation of the children of immigrants in the labour market, including the effects of the financial crisis (Meng, Verhagen and Hijgen, 2013). That report also splits ethnic groups into three educational levels: early school leavers, people with a secondary vocational certificate (MBO) and those who hold a bachelor’s or master’s degree. The findings of the SCP report show a more negative picture than the figures of the TIES survey from before the financial crises. The children of Moroccan immigrants who are early school leavers score very high, with an unemployment rate of 54% 18 months after they left education. Among the early school leavers who are the children of Turkish immigrants, the corresponding unemployment rate was 31%. In both cases, there is a huge difference with early school leavers of Dutch descent (18%). The high unemployment rates among early school leavers 18 months after school-leaving are alarming.

The young adult offspring of Turkish and Moroccan immigrants in possession of a secondary vocational certificate (MBO) show lower unemployment rates than the early school leavers, but still one in five (Turkish) and one in four (Moroccan) is unemployed. Among the children of Moroccan immigrants, unemployment is five times higher compared to peers of Dutch descent holding the same secondary vocational educational certificate (Meng, Verhagen and Hijgen, 2013). Part of the difference can be attributed to the type of secondary vocational educational certificate (Mbo-2, -3 or -4), but the report indicates (without giving further details) that the largest part of the gap remains unexplained (Meng, Verhagen and Hijgen, 2013). The financial crisis has hit the immigrant offspring especially hard: the gaps have grown notably between 2007 and 2012.

Many holding a secondary vocational certificate find their first job through an apprenticeship. Asked about problems finding an apprenticeship, 22% of the children of Moroccan descent indeed mention problems, compared to 15% of their peers of Dutch descent. The respondents of Turkish descent are doing slightly better, with 19% (Meng, Verhagen and Hijgen, 2013). It seems that difficulties finding an apprenticeship may play a role in explaining unemployment differences afterwards.

The TIES Survey revealed a major difference in unemployment rates between the highly educated children of Turkish and Moroccan immigrants. The SCP findings are based on a larger sample. Looking at the people with a BA degree, the situation of the highly educated children of Moroccan immigrants looks again a bit less favourable, but at the MA level it is the other way round. The gap with students of Dutch descent with the same diploma is very wide. The usual individual characteristics that explain such gaps do not greatly figure here, because the groups of recent graduates with BAs or MAs are very similar in terms of age and experience. This only leaves sector differences, networks, the way one searches for jobs and discrimination as the most important explanatory factors. Sectoral differences should actually give the children of immigrants a better position, because they are overrepresented in the prospering sectors: business, law, medicine and IT.

The SCP report shows that during the economic crisis the chances of unemployment grew substantially, indicating that if there are fewer jobs, the effect of discrimination and other factors, such as having the right network contacts, become more salient.

Table 4.11. Unemployment rates of recently graduated students with a bachelor’s or master’s degree, percentages

Recently graduated	Of Dutch descent	Children of Turkish immigrants	Children of Moroccan immigrants
Bachelor’s degree	6	14	17
Master’s degree	5	13	8

Source: SIS 2010-12 and VSNU WO monitor 2009-11.

In 2014, Andriessen, Ferhee and Wittebrood (2014) carried out a major research project on discrimination in different societal contexts, one being the labour market. For this research, people were selected based on a sample of register data with a response rate of 26%. Of the respondents of Moroccan descent, 41% mentioned experiencing discrimination while looking for work, as did 35% of the respondents of Turkish descent. Also, the discrimination reported during the search for an internship was relatively high, with 24% and 29% respectively for people of Moroccan and Turkish descent. At the workplace, slightly more than one in three persons in both groups reported instances of discrimination. In all three of these contexts, people of Surinamese and Antillean descent mentioned fewer instances of discrimination. The report created considerable public and political debate in the Netherlands over racism and discrimination in the labour market.

Andriessen et al. (2015) did a follow-up study in the region of The Hague the following year. The researchers worked with correspondence tests: they would submit applications for the same job openings with the same employers, submitting the same CV but under different names. They sent 504 applications for mid- and low-level jobs. The names used were typical Dutch, Moroccan or Hindustani Surinamese. The applicants with a Dutch name were invited almost twice (1.8) as often for an interview as those with a Moroccan name. When two extra years of experience were added to the CV of the applicant with a Moroccan name, the difference vanished. According to the researchers, this indicates that employers see a potential risk in hiring people from this group, but the risk can be reduced by extra work experience.

A further important question is whether or not the three groups (early school leavers, the senior vocational educated and the higher educated) find work at their own skill level. Again the TIES Survey data come into play.

Table 4.12. Job level of early school leavers among the children of Moroccan and Turkish immigrants, ages 18-35, percentages

Early school leavers	Unskilled	Skilled	Self-employed	White collar	Executive professionals
Children of Moroccan immigrants	42	11	3	35	9
Children of Turkish immigrants	45	10	0	40	6

Source: TIES survey 2008.

The findings show that about half of the early leavers still manage to get a position higher than unskilled work, which in turn reveals that for this group there is considerable room for mobility on the labour market.

Table 4.13. Job level of respondents with a secondary vocational certificate among children of Moroccan and Turkish immigrants, ages 18-35, percentages

Secondary vocational education	Unskilled	Skilled	Self-employed	White collar	Executive professionals
Children of Moroccan immigrants	19	9	1	56	15
Children of Turkish immigrants	14	12	5	56	13

Source: TIES survey 2008.

Depending on the type of diploma, senior vocational education can lead to skilled or white collar mid-level jobs. A considerable portion of this group even makes it into executive and professional jobs.

The last group to examine involves the people with higher education diplomas. Here a reversed trend can be seen: about a quarter of the people are working below their skill level. The comparison group of more highly educated people of Dutch descent in the TIES Survey works much more often (81%) in an executive or professional function.

Table 4.14. Job level of respondents with a higher education diploma among children with Moroccan and Turkish immigrant parents, ages 18-35, percentages

Higher education	Unskilled	Skilled	Self employed	White collar	Executive professionals
Children of Moroccan immigrants	3	0	0	25	72
Children of Turkish immigrants	2	0	2	27	69

Source: TIES survey 2008.

To sum up the findings: for less educated daughters of immigrants, traditional gender roles play an important role in (not) entering the labour market. For those who are less educated and educated at the mid-level, both male and female, the labour market offers considerable opportunities to move up beyond their educational level. Regrettably, the children of immigrants who hold a higher education diploma still face considerable challenges, both when entering the labour market and in finding a job at their skill level.

Gender differences in labour market outcomes

In the SCP integration report of 2013, there is a separate chapter on immigrant women in the labour market (van der Vliet, Gijsberts and Dagevos, 2013). Here the authors look at women aged 20 to 50 and focus on life events that happened between 2007 and 2010. The report shows that 70% of the women of Turkish descent were working, as were 75% of the women of Moroccan descent. This is 10-15% less than in the group of women of Dutch descent. The report does not give separate figures for the children of immigrants. The report also gives figures for women quitting their paid work upon having their first child, and here there are separate figures for the children of immigrants. About a quarter of the daughters of Turkish and Moroccan immigrants stop working in a paid job after having their firstborn. That is two and half times more than women of Dutch descent (10%). An important explanatory factor for this difference is how child care is arranged in the Netherlands. For working parents, the financial contribution to child care is substantial. The tipping point of that cost becoming higher than the second income in the family is easily reached, especially if one has a low-wage job. For women with high-paying jobs, the balance is much more positive. In fact, child care policy discourages women with low-level jobs from staying in the labour market. These “choices” in their turn create an important difference in family incomes. Much of the income difference among households depends on whether the woman is also working a paid job. That variable often decides whether people can enter the middle class or not.

The mothers of the preceding generation of immigrants often stopped working when they had children, and did not return to the labour market. This was always one of the big differences with women of Dutch descent. Among the children of immigrants, the group of women who re-enter the labour market is much larger: 38% of those of Turkish descent, and 32% of those of Moroccan descent. For those of Turkish descent, this is only a fraction lower than among women of Dutch descent returning to work (40%) (van der Vliet, Gijsberts and Dagevos, 2013).

Intergenerational mobility in the labour market

This paragraph will look at intergenerational mobility based on the labour market position of the sons and daughters of immigrants compared with their fathers and mothers respectively. The outcomes in general show that the offspring are doing much better in the labour market than their parents.

Beginning with sons and their fathers, the focus is on the position the parents held in their career when the respondent was fifteen. The two groups examined are the largest: fathers working in low-level jobs, and fathers that did not have paid work at that time.

The largest group of fathers was working in low-level jobs when their son was fifteen. Of the Moroccan sons of this group, about a quarter (24%) does not have paid work now. Another quarter (21%) of the sons works at the same low level as their fathers. Slightly under half (43%) of the sons whose father worked in low-level unskilled jobs made a considerable mobility leap compared to their fathers.

The other big group in the generation of Moroccan fathers consists of men who were unemployed or in disability schemes when their son was fifteen. About a third of the sons (32%) are also not in paid work; 41% work at an unskilled level; and a mere 18% work at a considerably higher job level than their father. It seems that the sons of Moroccan fathers who were not active in the labour market when they were fifteen are doing worse now than those with a father working a low-level job when they were fifteen. The fact that Moroccan immigrant families were larger, combined with the family living on benefits, probably has contributed to a more precarious situation for these sons to grow up in. One can imagine that this has, for instance, affected where the family was able to afford housing, the availability of a room to study, and the pressure on the son to already begin working during school age.

Table 4.15. Comparison of labour market outcomes for sons of two groups of Moroccan immigrant fathers, percentages

Moroccan	No paid work	Unskilled work	Skilled work	White collar and executive work
Father with low-paid work	24	21	12	43
Father with no paid work	32	41	9	18

Source: TIES survey 2008.

In the Turkish group, differences between sons of working fathers and those of fathers who were not in paid work are not as pronounced. In both groups, a considerable number are doing much better in the labour market than their father.

Table 4.16. Comparison of labour market outcomes for sons of two groups of Turkish immigrant fathers, percentages

Turkish	No paid work	Unskilled work	Skilled work	White collar and executive work
Father with low-paid work	17	19	24	40
Father with no paid work	19	27	11	43

Source: TIES survey 2008.

For both groups, there is no significant relation between the job level of the father and the job level of the son. There is also no significant relation between the educational level of father and the occupation of the son.

Now moving to the mothers and their daughters, the largest group among immigrant women is formed by those who did not have paid work when their daughter was fifteen. Looking first at the children of Moroccan immigrants, there are no major differences between the daughters of mothers working in paid jobs and those of mothers not working in paid jobs. Of the daughters, 57% work in white collar jobs or in executive or professional jobs. These women made a large intergenerational mobility leap compared to their mothers, who often had very little education.

Table 4.17. Comparison of labour market outcomes for daughters of two groups of Moroccan immigrant mothers, percentages

Moroccan descent	No paid work	Unskilled work	Skilled work	White collar and executive work
Mother with no paid work	41	1	1	57
Mother with paid work	43	0	0	57

Source: TIES survey 2008.

With the children of Turkish immigrants, one does see differences between the daughters of mothers who were in paid jobs and those whose mothers were not. But here too, about half of the daughters are in white collar jobs, in executive or professional jobs, or self-employed. They also made a huge intergenerational mobility leap compared to their less educated mothers.

Table 4.18. Comparison of labour market outcomes for daughters of two groups of Turkish immigrant mothers, percentages

Turkish descent	No paid work	Unskilled work	Skilled work	White collar and executive work
Mother with no paid work	53	5	0	42
Mother with paid work	41	0	0	59

Source: TIES survey 2008.

The outcomes for the daughters of immigrants compared to their mothers appear to be much more polarised than the outcomes for the men. Those who enter the labour market usually do so in mid-level or higher-level jobs. However, it seems that those who must choose between working a low-level job and being a full-time housewife choose the latter. The high costs of child care might be the deciding factor in this choice.

The same analysis can be performed for the female respondents of Dutch descent. A comparison with male respondents of Dutch descent is not possible, because there is no comparably large group of fathers of Dutch descent with the same indicators as the Turkish and Moroccan immigrant fathers. The rare occasions of fathers of Dutch descent having no education at all, or no education beyond primary school, involve severe learning or behavioural problems and thus make comparison problematic.

For the female respondents of Dutch descent, comparison is possible. There are not a lot of differences between daughters whose mothers had paid work when they were fifteen, and those whose mothers did not.

Table 4.19. Comparison of labour market outcomes for daughters of two groups of Dutch native mothers, percentages

Dutch descent	No paid work	Unskilled work	Skilled work	White collar and executive work
Mother with no paid work	11	6	1	82
Mother with paid work	15	5	4	76

Source: TIES survey 2008.

Intergenerational mobility at the household level

Some of the TIES respondents were already married or living together with someone at the time of the survey, and information was gathered about their partner. The TIES data allow a glimpse at the number of incomes in these households. Data indicate that in almost half (44%) of the Moroccan households of immigrant offspring, both partners have a paid job; in 42% one partner works; and in 14% of the cases neither partner has paid work (note that this last group also includes students). The situation in the Turkish households is very similar: 42% have two household incomes; 50% have one income; and in 8% neither partner has paid work (again, this group includes students).

Comparing their situation with that of their parents when the respondents were fifteen reveals a radical change. In the Turkish immigrant households, 18% had two incomes, against 42% in those of the offspring of immigrants. There is a similar pattern for the group of Moroccan descent; considerably fewer households of the offspring are without income from work.

Table 4.20. Number of incomes in the households of immigrant parents and those of immigrants’ offspring, percentages

	Turkish immigrants	Children of Turkish immigrants	Moroccan immigrants	Children of Moroccan immigrants
Two incomes	18	42	15	44
One income	45	50	53	42
No paid work	37	8	32	14

Source: TIES survey 2008.

This difference in the number of household incomes from paid work is the most dramatic change the offspring made compared with their immigrant parents. Two to three times more households live from two incomes, and two to three times fewer households must make ends meet without an income from paid work. More than half of the households with two incomes from paid work take home more than EUR 3 000 after taxation every month. Compared with the immigrant parents, they enjoyed a very steep rise in income. The women are the crucial factor behind this upward mobility; their educational level is decisive in this trend.

Table 4.21. Percentage of daughters of Moroccan and Turkish immigrants who have a paid job in relation to final educational outcomes

	Early school leavers	Secondary vocational education	Higher education
Daughters of Moroccan immigrants	32	83	96
Daughters of Turkish immigrants	35	60	68

Source: TIES survey 2008.

As seen before, when the women work they usually do so in well-paying white collar and executive or professional jobs. One can conclude that it is the women who are the drivers behind the most spectacular intergenerational mobility of the children of immigrants.

Looking at the people of Dutch descent a similar trend is evident, with the children much more often living in two-income households than their parents.

Table 4.22. Number of incomes in the households of parents of Dutch descent and those of their children, percentages

	Parents of Dutch descent	Children of Dutch descent
Two incomes	50	80
One income	45	17
No paid work	5	3

Source: TIES survey 2008.

Conclusion

Looking at education and labour market outcomes, there has been considerable intergenerational mobility for the native-born Dutch children of Turkish and Moroccan immigrants in the Netherlands. Especially for the daughters of the Moroccan immigrant mothers, the upward mobility has been remarkable. About a quarter of the daughters of Moroccan immigrants made it into higher education, while their mothers only went to elementary school for a few years or were illiterate. This is truly remarkable. The mobility of the children of Turkish immigrants is also considerable, but somewhat less pronounced. While the Turkish parents were slightly better educated than the Moroccan parents, the educational outcomes for the Turkish offspring are either at the same level as the Moroccan offspring or lower. The daughters of Turkish immigrants took longer than their peers of Moroccan parents to overtake the men in educational achievement. This can be attributed to more traditional gender roles and a tighter Turkish community that resisted change.

This chapter shows that educational success is no guarantee for a well-paying professional job. Although they do better than their peers with poor educational credentials, the more highly educated children of immigrants still have difficulties accessing the labour market, and many do not find a job at their skill level. This is a major concern, and not just for the people involved. It is a potential waste of talent, and it can send a negative signal to the next generation that there is not much merit in education.

The steepest mobility is evident at the household level. Comparing immigrants with their native-born children, the steepest mobility is found with respect to the number of incomes in the household. It is usual that whenever there are two incomes from paid work that people can move into the (lower) middle class and provide a substantially different starting position for their children, the next generation. These children will grow up in a middle class environment and will go to the better-performing and more mixed schools. These families will, for instance, also have the financial means to send their children to full-day day care.

The large gains that the children of immigrants made on average compared to their parents, however, also conceal a group for whom social mobility is almost absent. This is the group that left school early, the group that in the case of the women often did not enter the labour market, and in the case of the men often include a partner that has no paid work. Traditional gender roles continue in this group, which hampers their social mobility. At best, it can be said that this group stagnates in a lower working class position. But one can also argue that stagnation could in fact create downward mobility for their children. Being an urban poor family means living in neighbourhoods that are unsafe, are unattractive, and have a high concentration of urban poor. In these neighbourhoods the schools are often weak. It is possible that the grandchildren growing up in these circumstances could be worse off than the generation before them, because their parents, being early school leavers, probably will not (be able to) transfer a positive message to their children about the advantages of good schooling. However, empirical evidence on the integration of persons whose grandparents have immigrated is currently lacking in the Netherlands.

One of the key findings of this chapter is the polarisation that exists among the children of immigrants. A large group of them have been remarkably successful given the background characteristics of their parents. However, there is also a group similar in size that stays behind. And their situation is one of the most vulnerable of all groups living in the Netherlands. This polarisation is why, in the Dutch debate about the position of the children of immigrants, according to some the glass is half full, and according to others it is half empty. The reality, however, is that both beliefs are true at the same time. The polarised outcomes are to a large extent the result of the Dutch educational system. The main determining factor is what is happening in secondary vocational education, MBO. Both success and failure are decided at that juncture in the educational pipeline. About three-quarters of young people with Turkish and Moroccan parents follow secondary vocational education. Those who obtain their certificate can move on to higher vocational education and get their BA. However, those who drop out of secondary vocational education only have a lower vocational secondary education diploma, making them early school leavers. The line between success and failure is therefore sometimes very thin. It is not only the intellectual capacities that determine success or failure at this point, but also, for example, determination and pull factors from the labour market.

In a study of successful offspring of immigrants the author, together with E. Keskiner, J. Schneider and F. Lelie, developed the theory of the multiplier effect to explain steep mobility (Crul et al., 2017). According to the theory, initial small differences between youngsters and their families and their peers multiply over time. This is probably most clearly visible among the girls. If they are successful in school, that is associated with later marriage. One would expect this to involve a higher likelihood of choosing their partner in full independency, who then more often is born or raised in the Netherlands and is also more highly educated. That situation would allow them to enter the labour market, with the result that there are two incomes in these households. With each consecutive step, they move further away from their less successful peers. In contrast, as TIES data show, a woman who drops out of secondary vocational education is more likely to marry young and to marry a partner who is also less educated and more often a (recent) immigrant from Turkey or Morocco. The young age and the low education level make it all the more unlikely to enter the labour market. Here, institutional arrangements and traditional gender patterns make for an unfavourable situation. For those who can only apply for low-level jobs, the cost for child care tends to be almost equal to their wage. In these households, as the discussion has shown, there is often only one income out of paid work.

Choices early on in the school career can lead to big and consequential differences over time. The resulting “multiplier effect” (Crul et al. 2017) implies that the successful groups gain more social and cultural capital with each step upward that they make in education and on the labour market ladder. This explains how children of less educated immigrants can indeed be successful against all odds. They do not have the social and cultural capital in their family regarded as crucial to become successful, but they gain these forms of capital bit by bit along their way up. The people who, for instance, succeeded in getting into a pre-academic stream not only secured a position enabling them to continue to higher education directly after secondary school, but in doing so they typically also entered a different social environment, with students that mostly came from middle class families of Dutch descent. In the interaction with these middle class peers, they gain that cultural capital unavailable in their families. Once in university, they gain social capital through their fellow students, which later helps them access networks useful in the labour market. Thus, they acquire skills along the way.

The multiplier effect explains some of the large within-group differences. A second explanation may arise from how educational institutional arrangements affect outcomes. The Dutch educational system involves many choices, and unequal opportunities can result from these choices. This already starts at preschool: parents are the ones to decide whether to send their children to preschool or not. This also depends on the preschool places available in the neighbourhood where the parents live. Similar highly consequential choices must be made around the end of elementary school. As the chapter showed, pupils receive advice from their elementary school suggesting the type of secondary education best for them, but it is up to the parents whether to follow that advice or not. And children have different opportunities depending on that decision. Some secondary schools still offer the possibility of moving up to pre-academic streams after one- or two-year intermediary classes, while others do not. Limited knowledge of these differences can severely limit the chances of the children. The school “choice” is key in the Dutch system because of the early streaming at age twelve. This results in streaming a lot of pupils of immigrant descent into vocational tracks, often merely because of Dutch language deficiencies they still have at that age. That forces many onto the long route through senior vocational education. Since the route entails three additional years to get into higher education, deliberate choices must be made to continue studying and to pursue a higher education degree. This takes strong personal determination and family support, but it also involves financial consequences: postponing paid work but instead taking on a study loan. Looking at the entire educational pipeline reveals how, from the very beginning, inequalities persist that disfavour children with low-educated and poor parents, especially the many in this group who have immigrant parents.

References

Andriessen, A., H. Ferhee and K. Wittebrood (2014), Ervaren Discriminatie in Nederland, SCP, The Hague.

Andriessen, A., B. van der Ent, M. van der Linden and G. Dekker (2015), Op Afkomst Afgewezen: Onderzoek naar Discriminatie op de Haagsche Arbeidsmarkt, SCP, The Hague.

Beets, G., S ter Bekke and J. Schoorl (2008), “Migration history and demographic characteristics of the two second-generation groups”, in M. Crul and L. Heering (eds.), The Position of the Turkish and Moroccan Second Generation in Amsterdam and Rotterdam: The TIES Study in the Netherlands, Amsterdam University Press.

Bruggers, I., G. Driessen and M. Gesthuizen (2014), “Voor- en vroegschoolse voorzieningen, effectief of niet? Effecten van voor- en vroegschoolse voorzieningen op de taal- en rekenprestaties van leerlingen op de korte en langere termijn”, Mens & Maatschappij, Vol. 89, No. 2, pp. 117-50.

CBS/SCP (2016), Emancipatie Monitor 2016, CBS/SCP, The Hague.

Claassen, A, and L. Mulder (2011), Een Afgewogen Weging? De Effecten van de Gewijzigde Gewichtenregeling in het Basisonderwijs, ITS (Stichting Instituut voor Toegepaste Sociologie), Nijmegen.

Crul, M. (2010), “The progress in the educational position of children of Turkish descent in the Netherlands: A historic overview”, International Journal on Multicultural Societies, Vol. 11, No. 2, pp. 175-95.

Crul, M. (2000), De Sleutel Tot Succes: Over Hulp Keuzes en Kansen in de Schoolloopbanen van de Turkse en Marokkaanse Jongeren van de Tweede Generatie, Het Spinhuis.

Crul, M. and J. Doomernik (2003), “The second generation in the Netherlands: Divergent trends between and polarization within the two groups”, International Migration Review, Vol. 37, No. 4, pp. 1039-1064.

Crul, M. and L. Heering (2008), The Position of the Turkish and Moroccan Second Generation in Amsterdam and Rotterdam: The TIES Study in the Netherlands, Amsterdam University Press.

Crul, M., J. Schneider and F. Lelie (2012), The European Second Generation: Does the Integration Context Matter, Amsterdam University Press.

Crul, M., E. Keskiner, J. Schneider and F. Lelie (2017), “The multiplier effect: How the accumulation of cultural and social capital explains steep upward mobility of children of low educated immigrants”, Ethnic and Racial Studies, Vol. 40, No. 1-2, pp. 321-38.

de Vries, M. (1987), Ogen in je Rug: Turkse Jonge Meisjes en Jonge Vrouwen in Nederland, Wolters Samson, Alphen aan de Rijn.

Driessen, G. (2015), “De wankele empirische basis van het onderwijsachterstandenbeleid: De afnemende validiteit van indicatoren voor de toewijzing van extra middelen”, Mens & Maatschappij, Vol. 90, No. 3, pp. 221-43.

Driessen, G. (2013), De Bestrijding van Onderwijsachterstanden: Een Review van Opbrengsten en Effectieve Aanpakken, ITS, Nijmegen.

Driessen (2012), De Doelgroepen van het Onderwijsachterstandenbeleid: Ontwikkelingen in Prestaties en het Advies Voortgezet Onderwijs, ITS, Nijmegen.

EP and Nuffic (2015), Education system, The Netherlands: The Dutch Education System Described, EP and Nuffic, The Hague.

Fettelaar, D., L. Mulder and G. Driessen (2014), Ouderlijk Opleidingsniveau en Onderwijsachterstanden van Kinderen Veranderingen in de Periode 1995-2011, Radboud Universiteit, ITS, Nijmegen.

Fukkink, R.G., L. Jilink and R.J. Oostdam (2017), “A meta-analysis of the impact of early childhood interventions on the development of children in the Netherlands: An inconvenient truth?”, European Early Childhood Education Research Journal, http://dx.doi.org/10.1080/1350293X.2017.1356579

Groenewold, G. and L. Lessard-Philips (2012), “Research methodology”, in M. Crul, J. Schneider and F. Lelie (eds.), The European Second Generation: Does the Integration Context Matter?, Amsterdam University Press.

Groenewold, G. (2008), “Appendix: Sample design, TIES Survey implementation and evaluation”, in Eds. Crul and Heering, TIES Country Report: The Netherlands, Amsterdam University Press.

Herweijer, L., J. Iedema and I. Andriessen (2016), “Prestaties en schoolloopbanen in het onderwijs”, in SCP, Integratie in Zicht, SCP, The Hague.

Holdaway, J., M. Crul and C. Roberts (2009), “Cross-national comparison and outcomes for education of the second generation”, Teachers College Records, No. 111, pp. 1381-1403.

Huijnk, W. (2013) “De arbeidsmarktpositie ontleed”, in Jaarrapport integratie 2013”, SCP, The Hague.

Inspectie van het Onderwijs (2016), De Staat van het Onderwijs: Onderwijsverslag 2014/2015, Inspectie van het Onderwijs, Utrecht.

Inspectie van het Onderwijs (2010), Resultaatbepaling van VVE door Scholen: Een Verkenning, Inspectie van het Onderwijs, Utrecht.

Jepma, I.J., P. Kooiman and A. van der Vegt (2007), Landelijke Monitor VVE: Eerste Meting, Voorjaar 2007, Sardes, Utrecht.

Leseman, P., H. Mulder, J. Verhagen, M. Broekhuizen, P. Slot and S. van Schaik (2017), “Effectiveness of Dutch targeted preschool education policy for disadvantaged children”, in H.P. Blossfeld, N. Kulic, J. Skopek and M. Triventi (eds.), Childcare, Early Education, and Social Inequality – An International Perspective, Edward Elgar, Cheltenham, England.

Leseman, P. and A. Veen (2016), Ontwikkeling van Kinderen en Relatie met Kwalititeit van Voorschools Instellingen, Kohnstamm Instituut, Amsterdam.

Lindo, F. (1996), Maakt Cultuur Verschil? De Invloed van Groep Specifieke Gedragspatronen op de Schoolloopbanen van Turkse en Iberische Migrantenjongeren, Het Spinhuis, Amsterdam.

Meng, C., A. Verhagen and T. Hijgen (2013), “Van opleiding naar arbeidsmarkt”, in Jaarrapport Integratie 2013, SCP, The Hague.

Onderwijsraad (2014), Onderwijspolitiek na de Commissie Dijselbloem, Onderwijsraad, The Hague.

Roeleveld, J., et al (2011), Doelgroepleerlingen in het Basisonderwijs, Kohnstamm Instituut and ITS, Amsterdam.

SCP (2016a), Integratie in Zicht, SCP, The Hague.

SCP (2016b), Wikken en Wegen in het Hoger Onderwijs, SCP, The Hague.

SCP (2014a), Jaarrapport Integratie 2013, SCP, The Hague.

SCP (2014b), Huwelijksmigratie in Nederland, SCP, The Hague.

van der Vliet, R., M. Gijsberts and J. Dagevos (2013), “Veranderingen in de gezinssituatie en de arbeidsmarktpositie van vrouwen”, in Jaarrapport Integratie 2013, SCP, The Hague.

van Niekerk, M. (2004), “Afro Caribbeans and Indo-Caribbeans in the Netherlands: Pre-migration legacies and social mobility”, International Migration Review, Vol. 38, No. 1, pp. 158-83.

van Tuijl, C. and R. Siebes (2006), Het Rendement van Opstap in de Basisschoolperiode, Nederlands Jeugd Instituut, Utrecht.

Veen, A., J. Roeleveld and P. Leseman (2000), Evaluatie van Kaleidoscoop en Piramide, Kohnstamm Instituut, Amsterdam.

Veen, A., I. van der Veen and G. Driessen (2012), Het Bereik van Allochtone Kinderen met voor- en Vroegschoolse Educatie, Kohnstamm Instituut, Amsterdam.

Note

← 1. There is fierce debate among scholars about the effects of the voorschool. Driessen, in a blog in a national education magazine, talks about a denial by other scholars of facts suggesting that pre-school arrangements have hardly any effect (see also Bruggers, Driessen and Gesthuizen, 2014). Fukkink, Jilink and Oostdam (2017) also conclude, based on a meta-analysis of outcomes, that the effects of preschool arrangements are marginal or non-existent. Some reports, however, show that there are significant effects (van Tuijl and Siebes, 2006; Leseman and Veen, 2016).