they will get there!prof. dr. d.c. van den boom ten overstaan van een door het college voor...
TRANSCRIPT
![Page 1: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/1.jpg)
UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)
UvA-DARE (Digital Academic Repository)
They will get there! Studies on educational performance of immigrant youth in theNetherlands
van Welie, E.A.A.M.
Link to publication
Citation for published version (APA):van Welie, E. A. A. M. (2013). They will get there! Studies on educational performance of immigrant youth in theNetherlands. TIER.
General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).
Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.
Download date: 16 Mar 2021
![Page 2: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/2.jpg)
Top Institute for Evidence
Based Education Research
TIER Research SeriesUniversity of Amsterdam
They will get there!
Studies on Educational Performance of
Immigrant Youth in the Netherlands
Liesbeth van Welie
The educational performance of youth from immigrant descent has been a key subject in
political and public debate over the past decades. Typically, these debates focus on
underrepresentation in academic tracks in secondary school, and drop-out. The steady
increase, however, in enrolments in academic secondary levels among youth from immigrant
descent, seems almost overshadowed by the emphasis on problematic school careers. This
thesis shows that the average access to higher secondary tracks is slightly lower in largely
stratified neighbourhoods. Further research into three zip code areas with persistent
socioeconomic challenges, however, shows that pupils from Moroccan descent have better
chances to complete secondary school, when they do not switch schools. Furthermore, we
found that up to 90% of pupils in the four major Dutch cities choose another school than
the school nearest to their home. While native Dutch pupils on average prefer a school with
a lower percentage of migrant pupils, pupils from immigrant descent prefer a school with a
higher percentage of migrant pupils than is the case at the nearest school. For pupils from
immigrant descent, chances for upward mobility to a higher track increase slightly, but
significantly, with a lower distance between home and school. Finally, we found that relevant
and manageable existing scientific research, can adequately be matched with key questions
of diverse schools, and support these schools in the closing of the achievement gap.
Liesbeth van Welie graduated in Biology from Radboud University (Nijmegen,
the Netherlands), with a specialization in Aquatic Ecology. She started her career as a teacher
in Biology and was a few years later appointed as Principal of an innovative newly-founded
school for secondary education. Her second appointment as Principal was at a diverse
inner-city school in Amsterdam. Next, she worked as advisor to the Board of the University of
Amsterdam, and participated in several international networks in the field of quality
assessment of higher education, internationalization and organizational change. Her next
step was advisor to the Board of Maastricht University, followed by two years as senior
consultant at a consultancy firm. For several years she held the position of Chief Inspector of
secondary and higher education, where after she was asked to develop a new directorate for
the enhancement of evidence/ information based policymaking at the Ministry of Education,
Culture and Science. She worked on this doctoral thesis at the University of Amsterdam. They w
ill get the
re! Stud
ies o
n Educa
tiona
l Perfo
rma
nce o
f Imm
igra
ntYouth in the
Nethe
rland
s - Liesb
eth
van
We
lie
![Page 3: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/3.jpg)
THEY WILL GET THERE!
Studies on Educational Performance of Immigrant Youth in the Netherlands
![Page 4: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/4.jpg)
© Liesbeth van Welie, Amsterdam 2013
All rights reserved. No part of this publication may be reproduced, stored in a
retrieval system or transmitted in any form, or by any means, electronic, mechanical,
photocopying, recording, or otherwise, without the prior permission in writing, from
the author.
ISBN 978‐94‐003‐0052‐1
Cover design: Raadhuis voor creatieve communicatie, Alkmaar
This book is no. II of the TIER Research Series, a PhD thesis series published by TIER.
![Page 5: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/5.jpg)
THEY WILL GET THERE!
Studies on Educational Performance of Immigrant
Youth in the Netherlands
ACADEMISCH PROEFSCHRIFT
ter verkrijging van de graad van doctor
aan de Universiteit van Amsterdam
op gezag van de Rector Magnificus
prof. dr. D.C. van den Boom
ten overstaan van een door het college voor promoties
ingestelde commissie,
in het openbaar te verdedigen in de Aula der Universiteit
op woensdag 3 juli 2013, te 13:00 uur
door
Elisabeth Adriana Aleida Maria van Welie geboren te Appeltern
![Page 6: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/6.jpg)
Promotiecommissie
Promotor:
Prof. dr. J. Hartog
Overige Leden:
Prof. dr. J. Dronkers
Prof. dr. H. Maassen van den Brink
Prof. dr. J.G. Mora y Ruiz
Prof. dr. H.G. van de Werfhorst
Prof. dr. A.M.L. van Wieringen
Faculteit der Economie en Bedrijfswetenschappen
![Page 7: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/7.jpg)
In memory of JanKarel Gevers
Especially in our country, where the sometimes comical, but often also foolish race to
the top seems to prevail, de‐specialisation, imagination, overview and synthesis
might not be such a bad antidote (1996).
![Page 8: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/8.jpg)
![Page 9: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/9.jpg)
vii
Acknowledgements
A sense of the whole
When I started the endeavour to write a thesis, my only certainty was my
ambition to add to a better understanding of the school success of pupils of
immigrant descent. I am forever grateful to Joop Hartog, who guided me in such
mysterious ways that only now, in hindsight, I begin to understand how his intuitive,
thoughtful and wise guidance made me find my own way in this journey. Although
we appreciate system dynamics and nonlinear processes differently– as a highly
respected Economist, Joop feels less inclined to this way of thinking than I do as an
Ecologist– I intend to honour him by saying that he demonstrated system dynamics
at its best: a strong image of the complexity of the entire process of writing a thesis,
without being distracted too much by the intricacies of every step I had to take.
Biological time flows unevenly
Motherhood is the fundamental fulfilment of my life and largely surpasses
any other endeavour I undertook; getting children is an all‐overwhelming instant
All texts in Italics except the last one for my parents are from Gunderson L.H. and Holling C.S. (2002). Panarchy: understanding transformations in human and natural systems. Washington, DC: Island Press.
![Page 10: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/10.jpg)
viii
change of purpose. My talented and beautiful son and daughter, Felix and Bérénice,
taught me what transcendental love is, which only has the well‐being and happiness
of the other in mind. Through them I could more profoundly direct my work to the
aim of enhanced equity in educational opportunities for all young people.
Flip into another state
Both at the scale of creating the research institute TIER, and similarly in her
creativity in finding a way around all sorts of bureaucratic impediments that seemed
to block my way to the project of writing a thesis, I admire and respect Henriette
Maassen van den Brink for her perseverance, faithfulness and friendship. Noblesse
comes to mind in her presence.
Innovation undermines bureaucratic rule‐making and stability, but induces greater
resilience
In my last conversation with JanKarel Gevers, before he so sadly passed away,
he once more demonstrated in his beautiful eloquent style how deeply he believed
in academe, and how any bureaucratic impediments simply had to be moved out of
the way of education and research. Those who have known him will remember that
typical inquisitive look he could have, with his head a bit tilted. He looked at me in
that way when he said: "I do not know exactly how this came about, but great men,
with insight and wisdom, always seem to watch over you and guide you; you need
never to worry." Pepe Gines Mora, Roel in't Veld, Fons van Wieringen and Jaap
Dronkers were a source of inspiration for the most part of my career in education.
Koos van der Steenhoven always believed in my endeavour to add to equal
opportunities for all pupils; he created the solid ground I could stand on while
working on this thesis. He pushed research‐based policy making at the government
level over the tipping point; however complicated such changes may be, there will
only be ever more research basis for policies, there is no way back.
![Page 11: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/11.jpg)
ix
The shift from one set of scales to another
Every once in a while I stumbled, and doubted whether I could ever
successfully complete my research; at all those moments Hessel Oosterbeek not only
prevented me from falling, but also gave me at first sight casual suggestions, to try
another approach, read an article that just seemed to come to his mind, or "use less
words and think deeper". All his advice turned out to be exactly right, at exactly the
right time.
Understanding slow variables
Countless researchers, students and politicians are indebted to Hank Levin at
Columbia University, I among them. Hank dedicated a lifetime of work explaining to
the world that the return on investment of education is always positive– and should
be calculated over the lifetime of individuals, not by the mere costs of one year of
schooling per individual.
Growing numbers of children of immigrant descent have entered the school
system over the past decades and there have been times that I was deeply worried
about the tone of the public debate about migrant pupils. I have always believed in
the University and its academic members as good sui generis. At this point I want to
acknowledge the work of Herman van de Werfhorst and Maurice Crul, who in the
best tradition of academe described the steady progress of pupils of migrant
descent, and herewith opposed the negative debate on migrant pupils with sound
research, tirelessly. They strengthened my hope that indeed they will get there.
Adaptiveness is at the heart of understanding
No text survives sloppy editing; each time Juli Behrendt sent back my writing,
her editing skills surpassed my expectations. Apart from her expertise in matters of
style, she demonstrated time and again her profound understanding of my work.
Under her watch I learned to design my papers more clearly, and express myself in
better English. Most of all I valued that I could count on her at all times, as if no time
difference existed between Auckland, in New Zealand, and Amsterdam.
![Page 12: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/12.jpg)
x
Adaptation to faster levels
I could never have succeeded in translating my research questions on the
educational success of migrant pupils into sound analyses, without Ilja Cornelisz,
Bregje Zwaan, and Adam Booij. With creativity and great mastery of statistical
analysis they supported me in getting out of the data what indeed lay hidden inside
them. My visits to Ilja at Columbia University were very inspiring and great fun,
especially our discussions about our joint research paper in the library of Columbia
University– silent and solemn realms in other places‐ where a sign at the entrance
says that debate is explicitly encouraged. As if we needed that encouragement.
Adding another dimension
I devote an important part of my life to music, literature, and poetry; on
many bleak afternoons over the few past years, however, when I got lost in the
complexities of understanding my own data properly, it seemed as if all the Muses
had left me. In hours like these there was always Sebastiene Postma; she could
inspire me instantly again with her vast knowledge of the beauty of language and
literature.
Transfer and storage of experience
I value the experience that I had to struggle, like many pupils of migrant
descent, with bridging two cultures: the world of school practice and of academe. At
times I felt as if I was unable to express myself clearly in either world anymore. Lex
Borghans and Inge de Wolf taught me how to build that bridge; not by
compromising, but with identifying the best of both worlds.
Avete atque valete ("Ave atque vale", "Hail and Farewell", Catullus 84‐54BC)
My father and mother– who both passed away at the beginning of this
millennium– and I, had no easy life together. Although I was well aware of the
tragedies that struck them when they were young children, this did not help me to
![Page 13: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/13.jpg)
xi
have a more understanding view, in hindsight, on sadness and loneliness during my
childhood.
Maybe writing a thesis forces you to deeper understanding– it seems the
most challenging episode in my working life– but for the first time I could think of my
parents with profound kindness and gratitude: these are my mother's genes, which
planted the idea that there is no realm where I would not be entitled to enter, and
my father's genes, which enabled me to complete this grand intellectual odyssey.
![Page 14: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/14.jpg)
![Page 15: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/15.jpg)
xiii
Table of Contents
1 Introduction .................................................................................................... 1
1.1 A brief history of educating pupils with an immigrant background 1
1.1.1 "Black" and "white" schools 3
1.1.2 Was the integration in Dutch society of youth with a migrant background failing? 4
1.1.3 Segregation 5
1.2 Building upon success 6
1.2.1 A focus on local contexts and distinct groups 8
1.3 Outline of the thesis 8
1.3.1 Segregation at zip‐code level 8
1.3.2 Three socio‐economically challenged zip‐code areas 9
1.3.3 Distance measures offer new insights into patterns of secondary school choice 9
1.3.4 Matching relevant research findings with concrete school questions 10
1.3.5 Scientific and societal contributions 11
2 The educational gap and neighbourhood composition .................................. 12
2.1 Introduction 12
2.2 Previous research 17
2.3 Methods and data 22
![Page 16: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/16.jpg)
xiv
2.3.1 Selecting the data 22
2.3.2 Assembling the data set 24
2.3.3 Data from public files 25
2.4 Results 26
2.4.1 Enrolments in academic tracks among Moroccan Dutch, at zip‐code level 26
2.4.2 Enrolments in academic tracks in zips at risk 29
2.4.3 The correlation between the total percentage of migrants at zip level and
enrolments of Moroccan Dutch 31
2.5 Conclusions and discussion 32
2.6 Appendices 35
3 School switching, a concern for migrant pupils .............................................. 41
3.1 Introduction 41
3.1.1 Young people with a migrant background in Amsterdam‐West 42
3.2 Previous Research 45
3.3 Data and Methods 49
3.3.1 Single schools and clusters 49
3.3.2 Migrant students compared to native Dutch peers 49
3.3.3 Three data sets 50
3.4 Results 53
3.4.1 The frequency of switching in the three zip‐code areas 53
3.4.2 The retention rate of schools 56
3.4.3 Switching and upward and downward mobility between track levels 57
3.5 Conclusions and policy implications 61
3.5.1 The three zip areas 62
3.5.2 A hidden problem? 63
3.5.3 Youth in the three zips might benefit from agreements between schools to reduce
switching 63
3.5.4 Schools offering all tracks may be favourable for migrant students 64
3.5.5 Reducing switching may decrease costs 64
3.6 Appendix 66
![Page 17: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/17.jpg)
xv
4 Patterns in secondary school selection in the context of unlimited choice .... 67
4.1 Introduction 67
4.2 Previous research 72
4.3 Data and Methods 80
4.3.1 BRON data and additional data sources 80
4.3.2 Ethnic diverse populations in the four major cities 80
4.3.3 Distance measurements 82
4.3.4 SES indices 83
4.3.5 Indicators of pupils' prior achievement and secondary school quality 83
4.3.6 Limitation 84
4.4 Results 84
4.4.1 A general overview of socio‐economic measurements and distance to school 85
4.4.2 Distance to school 88
4.4.3 Choosing the nearest school or not 91
4.4.4 Selectivity of choice 94
4.4.5 School choice and segregation 97
4.4.6 Benefits of school choice? 101
4.5 Conclusions and discussion 104
4.5.1 Who travels further to school? 104
4.5.2 Differential sorting, mobility increases segregation 104
4.5.3 Upward mobility, an important opportunity for migrant students 105
4.5.4 Policy implications and further research 105
4.6 Appendices 107
4.6.1 Explanation of variables 108
5 A tailor made transfer of scientific knowledge to school practice ................. 119
5.1 Introduction 119
5.2 Previous research 121
5.3 Methodology 126
5.3.1 Six participating school leaders 127
5.3.2 Five separate phases of investigation 129
![Page 18: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/18.jpg)
xvi
5.4 Portraits of the participating schools and their main questions 132
5.4.1 School A: An enriched idea of the school's added value for students with a migrant
family history 133
5.4.2 School B: the advantage of a professional review of academic literature 135
5.4.3 School C: The important next step, translating research outcomes to teacher
practices 137
5.4.4 School D: Taking time to focus on the complex questions we encounter 138
5.4.5 School E: The pleasure of taking time for reflection 140
5.4.6 School F: Time is the most scarce ingredient 141
5.5 Findings 142
5.6 Conclusions and Discussion 147
5.6.1 Principals have questions that can be linked to research outcomes 147
5.6.2 Comparable questions, different approaches 148
5.6.3 Actual applications of academic knowledge and impediments 149
5.6.4 Contributing to the use of existing research in school practice 149
5.6.5 Contributing to research into the valorisation of scientific knowledge 151
5.6.6 Limitations and further research 152
5.7 Appendices 153
6 Conclusions and Discussion .......................................................................... 161
6.1 Conclusions 161
6.2 Policy Implications 166
6.2.1 A different perspective on segregation 166
6.2.2 Stable school careers 167
6.2.3 Validation of academic knowledge 168
6.2.4 The crucial importance of data 168
6.3 Limitations and further research 168
References ........................................................................................................... 171
Samenvatting ....................................................................................................... 183
Biography ............................................................................................................ 191
![Page 19: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/19.jpg)
xvii
List of Tables
Table 2‐1: Total average of 16 year‐olds per zip in the four major cities, and
average number and percentage of 16 year‐old Moroccan Dutch among this
group 27
Table 2‐2: Regression coefficients of the percentage enrolments in academic
tracks among 16 year‐old Moroccans on the share of Moroccan 16 year olds at
zip‐code level 30
Table 2‐3: Correlation of the percentage of 16 year‐old Moroccans per zip, and
the percentage among them in secondary academic tracks, for all zips and for
zips at risk separately 31
Table 2‐4: Correlations between total percentage of all immigrants per zip‐ and
the percentage of 16 year‐old Moroccans in secondary academic tracks,
separately for the four major cities. 32
Table 3‐1: Composition by ethnicity of all 16‐22 year‐olds living in zip‐code areas
1061, 1062 and 1063 in the district Amsterdam‐West, at the reference date of 31
July 2009 53
![Page 20: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/20.jpg)
xviii
Table 3‐2: The share of 16‐22 year‐olds by ethnicity, enrolled in academic
secondary tracks. Individuals were registered at least once in the data file as
being enrolled at this level 54
Table 3‐3: The number of secondary schools attended by 16‐22 year‐olds, by
ethnicity 55
Table 3‐4: The number of secondary schools attended by 16‐22 year‐olds, by
ethnicity, measuring school switching that includes cluster switching 55
Table 3‐5: Categories of school‐exit among pupils of Moroccan descent, who
were enrolled in academic secondary tracks (reference date 31 July 2009) 56
Table 3‐6: School switching among pupils in Amsterdam enrolled in the Year 3 of
secondary education in 2011 58
Table 3‐7: School switching among pupils living in zip‐codes 1061, 1062 and
1063, enrolled in the Year 3 of secondary education in 2011 58
Table 3‐8: Cluster switching among pupils in Amsterdam enrolled in Year 3 of
secondary education in 2011 59
Table 3‐9: Cluster switching among pupils living in zip‐codes 1061, 1062 and
1063 enrolled in Yea 3r of secondary education in 2011 59
Table 3‐10: Comparison of track level at entry and in Year 3 of secondary school,
separate for school switchers and non‐switchers, among pupils in Amsterdam 60
Table 3‐11: Comparison of the track level at entry‐ and in Year 3 of secondary
school, separate for school switchers and non‐switchers, among pupils in zip‐
code areas 1061, 1062 and 1063 60
Table 3‐12: Comparison of track level at entry‐ and in Year 3 of secondary school,
separately for school switchers and non‐switchers, among pupils of Moroccan
descent in Amsterdam. 61
Table 3‐13: Comparison of track level at entry‐ and in Year 3 of secondary school,
separately for school switchers and non‐switchers, among pupils of Moroccan
descent in zip‐code areas 1061, 1062 and 1063 61
Table 4‐1: Summary statistics for pupils living in the four major Dutch cities, at
the individual, neighbourhood and school level 86
![Page 21: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/21.jpg)
xix
Table 4‐2: Regression results (OLS) for distance to secondary school in km, on
individual‐, school‐ and neighbourhood variables (standard errors in
parentheses). 89
Table 4‐3: Summary statistics of the comparison between the group of students
who choose the nearest relevant school (i.e. a school that indeed offers a
desired track) to their home, and those who do not (standard deviations in
parentheses) 92
Table 4‐4: Regression results (OLS) for choosing the nearest school (i.e.
comparing pupils who choose the nearest school with those who choose another
school), with choosing the nearest school = 1 93
Table 4‐5: Regression results (OLS) for the distance difference (i.e. between the
actual and the nearest school), with distance difference = 0 for pupils who
choose the nearest school 95
Table 4‐6: Comparisons of nearest (non‐chosen) school and the actual school,
considering non‐nearest school choosers only (14247 individuals, 88.7 % of the
data set) 96
Table 4‐7: Regression results (OLS) for SES difference (i.e. average SES of chosen
school minus average SES of nearest school), considering non‐nearest school
choosers only (14197 individuals) 99
Table 4‐8: Regression results (OLS) for the difference in the percentage of pupils
with a non‐Western migrant background (i.e. percentage at the actual school
minus the percentage at the nearest school), considering non‐nearest school
choosers only (14197 individuals) 100
Table 4‐9: OLS regressions of individual upward mobility at the entrance of Year
3, considering only pupils who started at the lower vocational secondary level in
Year 1 (4343 individuals); non‐nearest school choosers only 103
Table 5‐1: All school questions clustered in five domains, plus matches with
selected scientific publications and the nature of every publication. 145
![Page 22: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/22.jpg)
xx
Table 5‐2: Categories of selected research publications, per cluster of questions.
(Note that the same publication may be suggested to several schools; e.g. 5
meta‐review studies were matched 15 times with a school question). 145
Table 5‐3: Overview of the outcomes of the evaluative interview. Schools A, B,
and E had an extra intermediate exploratory discussion. 146
Table 5‐4: Application (after Johnson, 1998) of research outcomes by school
leaders after three months. 147
![Page 23: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/23.jpg)
xxi
List of Figures
Figure 2‐1: Data scatter, linear regression (OLS) and LOWESS regression of the
percentage of 16 year‐old Moroccans enrolled in academic secondary tracks per
zip on the percentage of Moroccan Dutch 16 year‐olds as a share of the total
number of 16 year‐olds at zip level in the four major cities in the Netherlands 27
Figure 2‐2: Data scatter and linear regression (OLS) of the percentage of 16 year‐
old Moroccans enrolled in secondary academic tracks on the share of Moroccan
Dutch 16 year‐olds at zip level, separately for the four major cities in the
Netherlands 28
Figure 2‐3: Data scatter, linear regression (OLS) and LOWESS regression of the
percentage of 16 year‐old Moroccans enrolled in academic secondary tracks per
zip at risk on the share of Moroccan Dutch 16 year‐olds at zip level in the four
major cities in the Netherlands. 30
Figure 3‐1: The retention rate in ranking order of all secondary schools
(considering separate schools, not clusters) in Amsterdam, defined as the
percentage of pupils who entered in Year 1, and are still enrolled in Year 3 57
![Page 24: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/24.jpg)
xxii
Figure 4‐1: Linearly fitted lines (OLS), one for each of the 5 ethnic groups on the
correlation between absolute distance to secondary school and relative
neighbourhood SES 91
![Page 25: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/25.jpg)
xxiii
List of Appendices
Appendix 2‐1: Highest completed level of education, by age and ethnicity (2007) 36
Appendix 2‐2: Percentage of pupils with initial qualification for the labour market
(ISCED 3 and higher), by ethnicity and age cohort (2007) 37
Appendix 2‐3: Percentage of pupils with primary education as highest completed
level, by ethnicity and age cohort (2007) 37
Appendix 2‐4: Enrolment of 16 year‐olds by ethnicity in senior general education
(HAVO, ISCED 3) and pre‐university education (VWO, ISCED 3) 38
Appendix 2‐5: Percentage of pupils enrolled in academic track in Year 3 of
secondary school (school year 2011/2012), mean neighbourhood SES (scale ‐4 to
+4), and percentage of pupils eligible for weighted student funding when they
were enrolled in elementary education. 39
Appendix 2‐6: Data scatter, linear (OLS) and LOWESS regression of the
percentage of 16 year‐old native Dutch pupils on the percentage of 16 year‐olds
of Moroccan descent at zip level in the four major cities in the Netherlands 40
Appendix 3‐1: Comparison of the two main data sets. 66
Appendix 4‐1: Maps of Amsterdam, Rotterdam, Utrecht and The Hague. 110
![Page 26: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/26.jpg)
xxiv
Appendix 4‐2: LOWESS regression (Locally Weighted Scatterplot Smoothing;
bandwidth = .6) and linear regression (OLS) of the extra distance travelled to the
preferred school (as compared with the nearest school) and school's mean final
exam score (national standardized exam score) 115
Appendix 4‐3: Regression results (OLS) for difference in the mean relative exam
score at school level (i.e. the actual school exam score minus exam score at the
nearest school), considering non‐nearest school choosers only 116
Appendix 4‐4: Regression results (OLS) for difference in school percentage of
upward mobility to a higher track (i.e. upward mobility at the actual school
minus upward mobility at the nearest school), considering non‐nearest school
choosers only 117
Appendix 5‐1: General characteristics of the six participating schools. 153
Appendix 5‐2: Format of the first interview 153
Appendix 5‐3: Questions for the evaluating interview 155
Appendix 5‐4: Questions and matches per school 156
![Page 27: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/27.jpg)
1
1 Introduction
A person's true joy and felicity lie solely in his wisdom and
knowledge of truth, not in being wiser than others or in others'
being without knowledge of truth, since this does not increase his
own wisdom which is his true felicity.
Baruch de Spinoza, 1670
1.1 A brief history of educating pupils with an immigrant background
In the last decades of the past century, secondary schools, especially in the larger
Dutch cities, experienced substantial changes in the composition of their student
population. These changes were induced by the arrival, from the 1960s onwards, of
large numbers of labour migrants, who were invited to come and work in the
Netherlands. The majority of those labour migrants originated from Turkey and
Morocco, and were predominantly employed in low‐skilled work. In the years to come,
they were, in most cases, followed by their wives and children.
In their countries of origin, many Turkish and Moroccan people did not have the
chance to go to good quality schools, or pursue continued education; moreover,
substantial numbers of them had lived too far from an elementary school to be able to
go to school at all. Their children, among them increasing numbers who were born in the
![Page 28: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/28.jpg)
2
Netherlands, both literally and figuratively had to come a long way to find their rightful
place in the different strata of the Dutch educational system.
In 1987, I was appointed as Principal of a secondary school in Amsterdam‐West
that was in the midst of such significant changes in its student population. Many people
from Morocco and Turkey came to live in this part of the city– and still do– and their
children were arriving in secondary school in large numbers. Our students were not only
very diverse with respect to the country of origin of their parents, but also in their level
of schooling: some were born in the Netherlands, and had acquired a sufficient
proficiency in the Dutch language in elementary school, but others still struggled with
language– albeit they were born in the Netherlands– for example because another
language was spoken at home. However, a substantial number of pupils at that time had
been born in their country of origin, and had therefore been enrolled in Dutch
elementary schools at a later age than most of their classmates. Obviously, their
language proficiency required specialized attention. In order to serve our students well,
however, we had to re‐consider many other aspects of daily school practice as well.
In retrospect, our attitude was very goal‐oriented and practical: we wanted to
guide all our students to the highest possible level of secondary education, given their
individual intelligence and skills; and when the student body changes, teaching needs to
be reconsidered accordingly. In only a few years' time, our school population had
changed from being a student body almost exclusively of Dutch descent, to a vast
majority of pupils with a migrant background. Therefore, in a relatively short period of
time, we had to develop new practices to adapt our school to the needs of our changed
student body. Of course, we developed programmes for second language learners, but
we also acknowledged the importance of meaningful close cooperation with parents,
not only for explaining to them the complex system of tracked secondary education, and
its consequences for tertiary schooling and chances on the labour market in the future.
Importantly, some of our colleagues were skilled interpreters, who could explain the
school system to parents, and discuss the results and prospects of their sons and
daughters in their mother tongue. We also had to cope with, as another example, the
difficult issue of inventing new practices and procedures that truly assessed a pupil's
abilities and that "looked through" the façade of possibly insufficient language
proficiency, which may have masked real potential. We acknowledged, furthermore,
![Page 29: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/29.jpg)
1 Introduction
3
that we had to identify other kinds of stakeholders, and reach out to people with an
influential position in migrant communities. Together with them we mapped out the
route for our pupils towards far higher levels of schooling than had been available to
their parents in their country of origin.
Although we certainly had our share of problems with pupils who did not behave
as we expected them to do, as in any secondary school, we thought about our students
as ambitious young people, willing to learn, and eventually expecting to find a good
position on the labour market. Teachers were in general very dedicated; many among
them would devote their entire career to teaching migrant pupils and remained loyal to
their school until the present day, more than two decades later, even though such
experienced teachers may have a lot of job opportunities in other schools to choose
from.
Interestingly, American‐based studies on educational achievements among
migrant students in general report that, almost as something that goes without saying,
that migrant students express higher educational aspirations for themselves than is the
case among American‐born students. In general, these higher ambitions are explained
by the assumption that their parents, who took the far‐reaching decision to leave their
family and home country, may be a select group who are more often ambitious,
entrepreneurial, and willing to take substantial risks. Parents may even choose to
emigrate specifically for the purpose of finding better schools for their children.
At the time, in our school, we would have recognized these characteristics of
migrant pupils: our students performed remarkably well, and many of them were very
ambitious, and envisioned a future that would surpass all their families' expectations.
1.1.1 "Black" and "white" schools
Increasingly, however, our school and comparable other schools, were classified
as "black schools". And although in public discourse many discussants and columnists
started with some sort of statement that, of course, "black" was just an indication for
schools with many migrant students, and did not mean anything disparaging, the reality
was very different. The designation "black school" became almost equivalent with lower
results, lower chances for pupils, more disciplinary problems and more dropouts. This
dichotomy split the landscape of schools; some schools were even more determined
![Page 30: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/30.jpg)
4
than before to educate migrant pupils to the best of their potential. Other schools,
however, developed strategies to attract mainly native Dutch students, with more
affluent parents. As a result, especially in the four major cities Amsterdam, Rotterdam,
Utrecht, and The Hague, nowadays schools are to a large extent segregated (Ladd, Fiske
and Ruijs, 2009).
Considering the ample scientific evidence for the strong correlation between
parental levels of schooling and the educational prospects of their children, not
surprisingly, average exam results tend to be higher in schools with children from well‐
educated parents. The question remains, though, whether a higher mean exam score
indicates the added value of a school (Opdenakker and Van Damme, 2001), or rather
reflects the pupils' resources at home, including the presence of role models that may
lead to a sense of entitlement to high levels of schooling.
1.1.2 Was the integration in Dutch society of youth with a migrant background failing?
In 2000 Paul Scheffer published an essay in a national newspaper (NRC
Handelsblad) entitled "The Multicultural Drama" that rapidly became a cause célèbre.
Scheffer stated that unemployment, dropout and criminality accumulated among ethnic
minorities, and that he was not hopeful for the future: his expectation was that large
numbers of minority youths would permanently lag behind, and remain without
prospects for a meaningful participation in Dutch society. Importantly, the intent of
Scheffer's essay was to increase opportunities for youth with a migrant background to
participate in Dutch society, his critique being that policy makers, and society as a
whole, had for too long turned a blind eye towards the failing integration of youth with a
migrant background. In his view, this lenient attitude was an ineffective kind of
tolerance, and not in the interest of migrant students. However, the following public
discourse initiated a second dichotomy: those who agreed with Scheffer were
considered to be the new realists, while those who pointed at the continuous
educational improvement of migrant pupils were, oddly enough in a pejorative sense,
referred to as multiculturalists– people who were considered to be in a state of serious
denial. For more than a decade to follow, fundamental shifts in the Dutch political
landscape resulted in an almost categorically negative view of young people with a
migrant background, notably those of Moroccan descent. As a result, the ever growing
![Page 31: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/31.jpg)
1 Introduction
5
shares of migrant pupils who were successful in higher levels of schooling, and did so at
an increasing tempo, were almost ignored because of the strong emphasis on the
assumed educational and societal failure of migrant pupils.
Nevertheless, teachers and school leaders in secondary schools with large
percentages of migrant pupils invested with increasing success in adapting their teaching
to a very diverse student body, and were not at all of the opinion that the education of
migrant pupils was failing; on the contrary, they saw increasing numbers of enrolments
in the highest secondary tracks, and improving exam results. As the Director of a large
secondary school recently shared: "We stopped counting ethnicities and nationalities,
we consider this to be irrelevant; we want to serve, as best as we can, all pupils who are
living in the district around our school. To that end, we constantly work on improving
our programmes, the quality of our assessments, and the final exam results of our
pupils."
While diverse schools largely invested in adequate education for students with a
multitude of backgrounds, the societal discourse seemed to put desegregation of
schools and neighbourhoods centre stage, by implication suggesting that close‐knit
ethnic communities and neighbourhood schools were the cause of the assumed societal
failure of young people with a migrant family history.
1.1.3 Segregation
Among politicians and policy makers, pessimistic views on black schools and on
communities of people with a migrant background seemed to be mainly based on
research in the context of large metropolitan areas in the U.S. (Dobbie and Fryer, 2009;
Payne, 2010; Kerbow, Azcoitia and Buell, 2003). However, some major characteristics of
the Dutch educational system, and characteristics of the larger cities (where the majority
of pupils with a migrant background live) differ substantially from metropolitan areas in
the U.S. First, in the Netherlands, free school choice is universal and no financial
considerations restrict free choice, since all schools are funded equally by the
government, with additional funding for all low SES (socio‐economic situation) pupils
and no tuition fees. Second, unlike large metropolitan areas in the U.S., the major Dutch
cities do not have large impoverished areas that are at substantial distance from more
affluent areas. In most cases, lower SES neighbourhoods may be adjacent to more
![Page 32: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/32.jpg)
6
affluent areas. Third, Dutch cities have a high number of schools that are, furthermore,
all supervised by the Inspectorate of Education. Any seriously underperforming schools
(currently two schools in the four major cities) come under intensified supervision by the
Inspectorate, and typically manage to improve their results within the space of a year.
Fourth, teacher salaries (based on the level of training and teaching) are equal; diverse
schools do in general, not report other problems with the hiring of teachers than those
related to the shortage of teachers on the labour market in some subjects, and do not
report a higher turnover of teachers. In discussions, school leaders of diverse schools
express rather the contrary: the majority of their teachers made a deliberate choice to
teach pupils who, in most cases, are the first in their family to reach secondary track
levels that qualify them for higher education.
Obviously, attending a segregated school and living in a segregated community,
may raise concerns considering, for example, the preparation of migrant youth for a
successful future entry in the labour market. The question remains, however, whether
schools specialized in migrant education– with an emphasis on upward mobility to
higher tracks (Crul, Schneider and Lelie, 2012)– are the better option, or schools with
mixed populations, possibly supported by housing policies.
Remarkably, considering the effect of segregation and desegregation policies,
academic publications contain surprisingly contrasting findings: 1) large scale housing
policies aimed at desegregating impoverished residential areas did not result in any
improvement of educational attainment levels (Oreopoulos, 2007); 2) Desegregating
schools can further burden black pupils (Stuart Wells et al., 2009); 3) School investments
in cooperation with local segregated communities have beneficial effects on the school
results of black children (Dobbie and Fryer, 2009); and 4) neighbourhoods with high
percentages of inhabitants from different countries of origin have negative effects on
school results (Dronkers, 2010).
1.2 Building upon success
The increasing school success of pupils with a migrant background has inspired
this thesis, which is based on the assumption that preventing or repairing failure is not
simply the opposite of building upon success– if only because failing students and
successful students are different individuals. Interestingly, Crul et al. (2012) find, in a
![Page 33: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/33.jpg)
1 Introduction
7
large scale international comparative survey, that in the Netherlands the school results
of migrant pupils are remarkably polarized: being of migrant descent, both considerably
raises the odds of pupils successfully completing far higher levels of schooling than their
parents did, but, at the same time, raises the odds of dropping out– with about equal
percentages. Noguera (2004) describes similarly, that, in the U.S., immigrants are both
more likely to succeed and more likely to fail academically. Importantly, the fact that
these two sides of the distribution represent different individuals, national average
enrolments may not offer an adequate insight into patterns of school success in the case
of migrant pupils. Interestingly, van de Werfhorst and van Tubergen (2007) demonstrate
that, after controlling for parental education and occupational class, "ethnic differences
in achievement vanish, and differences in secondary school type almost disappear. What
remains (…) is not an ethnic penalty, but an ethnic advantage: Turks, Moroccans,
Surinamese and Antilleans choose higher types of secondary schooling than natives with
comparable class backgrounds".
This thesis aims to add to a better understanding of the successful group among
immigrant pupils for two main reasons: first, when success is better understood, policies
to reinforce success factors may add to cost‐effective government policies, since
repairing failure in fact comes too late, and may imply considerable costs (e.g. guiding
dropout students back to school, often repeatedly). Second, as mentioned earlier, the
success of growing shares among migrant students seems almost obscured by the
predominant attention to failure. As a result, notably Moroccan students are almost
categorically portrayed as underperforming.
This negative image may have consequences for the constantly growing group of
high achievers: they currently experience hindrances to find an internship or a job, and
unemployment is substantially higher among young people of migrant descent.
Admittedly, the effort to add to equity and justice for all pupils has been a major
driving force behind my long career in education; however, the point of departure for
the current research is the rational consideration that migrant youth in the larger cities
form half of the future labour force. There are, therefore, obvious economic interests in
educating them to the best of their potential.
![Page 34: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/34.jpg)
8
1.2.1 A focus on local contexts and distinct groups
In order to study the school success of migrant pupils, several preliminary
choices have been made. First, for the purpose of this thesis, access to the highest levels
of secondary education, the academic tracks that grant access to higher education, has
been defined as "success". Second, a focus was put on a better understanding of local
contexts, individual schools, and separate migrant groups (depending on the country of
origin of their parents). To this end, 1) the zip‐code level has been used as a local
geographical unit; 2) five distinct ethnicities have been taken into consideration
separately: pupils of Dutch, Moroccan, Turkish and Antillean/Surinamese descent, as
well as the combined group of "other migrants", that may, however, largely vary in their
socio‐economic characteristics; 3) a special focus was put on the school success of one
specific group, pupils with a Moroccan family history; 4) three specific zip‐code areas
were studied separately; and 5) six secondary schools with a large population of migrant
students were investigated in greater detail.
1.3 Outline of the thesis
1.3.1 Segregation at zip‐code level
In Chapter 2, a basic question underlying this thesis is explored: Is school success,
defined as enrolments in academic secondary tracks, among students of Moroccan
descent, associated with the degree of segregation at the zip‐code level of their area of
residence? Data on the ethnic composition among 16 year‐olds at zip‐code level in the
four major cities have been merged with individual data on enrolments in secondary
academic tracks at that same age, in order to analyse the association between the
residential area and school success.
As described above, contrasting outcomes from academic research have been
published concerning the effect of segregation at the neighbourhood level: at the same
time, existing research points at negative effects of segregated communities and
schools, but also demonstrates both positive and negative effects of desegregation
policies, and, furthermore, reports positive effects of strong (segregated) migrant
communities and neighbourhood schools.
![Page 35: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/35.jpg)
1 Introduction
9
1.3.2 Three socio‐economically challenged zip‐code areas
The aim of Chapter 3 is to study in more detail the school histories of young
people with a migrant background in three residential areas with a largely segregated
population, mainly of Moroccan descent, in the district of Amsterdam‐West. All three
areas have had to cope with serious socio‐economic disadvantages, and have been
depicted as notoriously problematic in the media over the past decade. This analysis is
based on a local data set, covering the school history (all exact exit and entrance dates)
of all 16‐22 year‐olds at the reference date of 31 July 2009. The study starts with the
open question: Does the analysis of municipal data on school histories of the youths in
this specific area, reveal local aspects of education that probably may foster the success
of some pupils, but as yet do not mitigate the problematic school careers of others.
Even at first sight, the database showed a surprisingly high incidence of recurrent
departure (without a diploma) and re‐enrolment in secondary school among pupils with
a migrant background.
The reviewed literature reports disruptive effects of school switching, both for
positive and for negative reasons, and finds that switching raises the odds of later
dropout. School switching appears to be a relatively under‐studied subject in the
Netherlands. In order to be able to compare the frequency of switching among
Moroccan youth living in the three investigated areas, in this chapter a second analysis is
carried out with the use of up‐to‐date national educational data. With this data set,
switching rates for Amsterdam as a whole, and differences among ethnic groups could
be demonstrated, measured in Year 3 of secondary schools. The national data set
confirms the findings based on the local data.
1.3.3 Distance measures offer new insights into patterns of secondary school choice
Chapter 4 is built around geographical distance measurements, notably the
distance to the school concerned and the difference in distance between the nearest
and the more distant preferred school, in the search for underlying patterns of school
choice between different migrant groups‐ and native Dutch pupils. Additionally, distance
measurements allow for controlling for population density, the number of schools
available within 5 km, and "urbanicity" (a measure for economic activity in a given area).
Patterns of school choice may reveal the preferences of parents and their children,
![Page 36: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/36.jpg)
10
notably when a school other than the nearest relevant school is chosen. The underlying
idea was that, in terms of travel time and possible travel costs, choosing a more distant
school requires an extra effort. Can patterns of school choice reveal the motivation for
this extra effort? We considered the extra effort of selecting a more distant school to be
a proxy for the selectivity of choice. Importantly, the choice for the nearest school may
also be a deliberate choice: parents in affluent neighbourhoods, for example, might
prefer the nearest school because the average socio‐economic level of the school
population may reflect the affluence of the residential area. In this case, however, we
cannot know the selectivity of their choice; others may opt for the nearest school to the
home‐address, without making a deliberate choice. For the purpose of this study, a rich
government data set has been used, containing all educational data and many socio‐
economic data of the cohort that was enrolled in the last grade of elementary education
in 2008. The data set has been merged with additional socio‐economic data on the
neighbourhood level, and qualitative data on school performance provided by the
Inspectorate of Education.
1.3.4 Matching relevant research findings with concrete school questions
Chapter 5 presents an in‐depth study of the transfer or the lack thereof– of
applicable outcomes from academic research to actual school questions and practice. Six
Principals of diverse schools in Amsterdam actively participated in this project. This
study has been inspired by, on the one hand, the abundance of high‐quality, recently
published studies on migrant education, and, on the other hand, the observation that
school leaders experience difficulties in selecting a knowledge base for current school
developments and innovative ambitions. Some issues at stake appear to be the timelines
and accessibility of research publications by school leaders and teachers, and the often
contradictory outcomes concerning the same educational theme.
On the basis of semi‐structured in‐depth individual interviews, the main school
questions were identified. Then, an extensive literature review was carried out to match
these questions to up‐to‐date scientific publications that offered opportunities for an
effective translation to school practices.
![Page 37: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/37.jpg)
1 Introduction
11
1.3.5 Scientific and societal contributions
In Chapter 6, general conclusions, based on the four studies in this paper, are
presented. Furthermore, possible implications for school practice and municipal and
national government policies are considered. Finally, the ways in which this thesis may
contribute to the field of scientific knowledge is reflected upon.
![Page 38: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/38.jpg)
12
2 The educational gap and neighbourhood composition
…no disadvantages stem from such freedom (i.e. liberty of
judgement) […] I do not need to go far to find instances of this.
Amsterdam is a fine example of a city which enjoys the fruits of
this liberty,…
Baruch de Spinoza, 1670
2.1 Introduction
The educational achievement gap between native Dutch students and pupils with
a migrant background, as measured by enrolments in the academic tracks of secondary
education, is substantial: in 2008 43.9% of native Dutch students1 were enrolled at this
level– that qualifies for access to higher education– against 19% of youngsters of
Moroccan descent.
1 Crul et al. (2012), rightly states that the vast majority of pupils with a migrant background are born in the Netherlands and have Dutch citizenship. As a consequence, Crul et al. (2012) criticize the habit to call students with a Dutch family history "native Dutch" and prefers to refer to this group as the "comparison group". Nonetheless, in this paper they are still referred to as native Dutch, for considerations of readability.
![Page 39: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/39.jpg)
2 The educational gap and neighbourhood composition
13
In the current chapter, the definitions of immigrant background used by the
National Statistics Office CBS have been used2. A main distinction has been made
between native Dutch and migrant background, with a further focus on a Moroccan
migrant background. The vast majority of pupils in the data set that has been used for
measurements in the current research are themselves not immigrants, they were born
in the Netherlands; migrant background refers to the history of their parents and
grandparents.
At present, about half of all 12‐year‐olds in the four largest cities in the
Netherlands have a migrant background; we are, therefore, looking at a majority of
minorities (Crul et al. 2012), rather than minorities. The vast majority of students with a
migrant background currently enrolled in secondary education were born in the
Netherlands, but, by definition, have at least one parent who was born abroad; they
form what is called the second generation. While the achievement gap indicates that
migrant pupils still as yet lag behind in enrolments in the highest tracks in secondary
education, their constantly increasing enrolment levels over the past decades
demonstrate, however, considerable educational success and intergenerational
improvement. The parents and grandparents of pupils, who are currently enrolled in
secondary schools, have on average lower levels of education. Especially among the
older generation, illiteracy may occur relatively frequently, because in their region of
origin, schools were not always available, and elementary education may not have been
compulsory in their early youth. Considering the strong correlation between parental
education and the level of schooling of their children, this makes the intergenerational
educational improvement of the current generation of young people of migrant descent
even more remarkable.
Nevertheless, public and political discourse over the past decade has been
almost categorically negative about people of all ages with a migrant background and on
the educational achievements of their children, notably those of Moroccan descent. This
2 First generation immigrant: someone born abroad with at least one parent born abroad. Second generation immigrant: someone born in the Netherlands with at least one parent born abroad. Foreign background: someone with at least one parent born abroad. Western immigrant: someone originating from a country in Europe (exclusive of Turkey), North America, Oceania, Indonesia or Japan. Non‐western immigrant: someone originating from Africa, South America, Asia (exclusive of Indonesia and Japan) or Turkey.
![Page 40: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/40.jpg)
14
negative discourse seems to have been inspired by notions of a multicultural drama
(Scheffer, 2000). In political debate, moreover, the supposed failure of the multicultural
society has been a recurrent topic. The discourse seemed to be dominated by this
supposed educational and societal failure of former immigrants and their children.
However, publicly accessible national statistics on education have recorded ever
increasing attainment levels among pupils with a migrant background; these easily
available statistics did not, however, apparently influence the negative image of migrant
groups in society.
This negative debate focused especially on youth with a Moroccan family history:
in 2008, 620 articles on Moroccan youth appeared in Dutch national newspapers, more
than half had a negative tone. In many articles a strongly negative picture was presented
of those neighbourhoods where many people of Moroccan descent live. However, while
the population of people with a Turkish background is roughly equal in size, by
comparison, only 41 articles on Turkish youth were published in the same period, among
which these were about 20 with a negative tendency (source: Ministry of Education).
The ever‐increasing average levels of schooling of the majority of Dutch Moroccan
pupils, referred to as active agents of change in cities by Crul et al. (2012), were virtually
unreported in public discourse. The desegregation of schools and neighbourhoods (in
public discourse called black schools and black neighbourhoods‐ "black" referring not to
race, but to a migrant background) became an important policy goal, since, the
residential concentration of ethnic groups was considered to be a major hindrance for
full participation in Dutch society.
Some characteristics of the Dutch system for secondary education may be in
favour of pupils with a low SES migrant background, who may be the first in their family
to be educated at academic levels: 1) in the Netherlands, all schools are funded equally
by the government; schools are, therefore, independent of local taxes for their funding;
2) there are no tuition fees; 3) children from all low SES parents (including migrant
parents) receive extra funding through a voucher system in elementary education, and
extra funding based on overall school SES composition in secondary education; 4) school
choice has been free for almost a century; no financial or other barriers limit parents
and children in their choice.
![Page 41: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/41.jpg)
2 The educational gap and neighbourhood composition
15
Furthermore, the distances between poorer and more affluent neighbourhoods
are relatively small in Dutch cities; poorer areas are not geographically distant or
isolated from high SES areas. Residential areas where many people with a migrant
background live are, therefore, never very geographically distant from more affluent
areas.
The current chapter sets out to present and describe actual educational
enrolment levels of 16 year‐old youth with a migrant background (reference date 1
October 2008). In order to better understand the effect of stratification at the
neighbourhood level, the association between the ethnic composition at the zip‐code
level and enrolment in academic tracks has been considered, with a focus on enrolments
of youths with a Moroccan background. Several choices have been made to delineate
the research target:
First, the analyses were restricted to pupils in secondary schools, who were living
in the four major cities Amsterdam, Rotterdam, Utrecht and The Hague, since the
majority of migrant pupils live in these four cities.
Second, because this research considers educational success in particular (and
not the full range of educational outcomes), only pupils enrolled in the two highest
tracks of secondary education have been taken into account. The Dutch system of
secondary education distinguishes three main track levels: pre‐vocational education
(further subdivided in four distinct levels), senior general education (HAVO, Dutch
acronym) and pre‐university education (VWO, idem). In the context of this research,
"success" has been defined as access to the two highest levels of secondary education:
senior general education and pre‐university education, in combination here referred to
as academic tracks. Both academic tracks grant access to higher education.3 The choice
3 The following translations of the Dutch educational system in the ISCED (International Standard Classification of Education by UNESCO, update 1997) classification are used:
VMBO: pre‐vocational secondary education, 4 years, ISCED 2, qualifying for senior secondary vocational education.
HAVO: senior general education, 5 years, grade 1‐3 ISCED 2, grade 4‐5 ISCED 3, qualifying for higher education.
VWO: pre‐university education, 6 years, grade 1‐3 ISCED 2, grade 4‐6 ISCED 3, qualifying for higher education.
MBO: senior secondary vocational education, level 1 ISCED 2, level 2‐4 ISCED 3, level 4 qualifying for higher education.
HBO: universities for applied sciences, ISCED 5B.
WO: research universities, ISCED 5A.
![Page 42: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/42.jpg)
16
to focus on the association between neighbourhood composition and access to
academic tracks has been motivated, furthermore, by the 2010 report of The
Netherlands Institute for Social Research that states, contrary to assumptions in public
and political discourse: "Dutch research does not indicate a clear association between
the ethnic composition of a neighbourhood and the socio‐economic position of
individuals. When considering chances of unemployment, dependency on social welfare
and poverty, individual characteristics, notably educational attainment, seem far more
important" (translated from Dutch by the author). Granted the importance of individual
educational attainment, neighbourhood composition may, however, have an indirect
effect on education.
Third, zip‐code level has been used to define the residential area of individual
pupils; and fourth, analyses are based on all 16 year‐olds living in the four major cities at
the reference date of 1 October 2008. Age has been considered to be a more accurate
selection than school grade, since pupils in the same grade may vary considerably in age,
especially in the case of migrant pupils, who may more often follow a longer route
through successive secondary tracks.
This resulted in the following research questions
1. How is ethnic composition at zip‐code level, associated with the access among 16
year‐old Moroccan Dutch pupils to academic tracks of secondary education
(HAVO and VWO), in the four major cities in the Netherlands (Amsterdam,
Rotterdam, Utrecht, and The Hague)?
2. Considering only zips with severe socio‐economic challenges (zips at risk), how is
living in these areas associated with access to academic secondary tracks?
3. Considering the total share of pupils with a migrant background in zip areas– as
opposed to measuring the share of Moroccan pupils specifically in the case of the
first two research question– how is the total ethnic composition of zip areas
related to access to secondary academic levels?
The completion of ISCED level 3 is the internationally agreed initial (or basic) qualification for the labour market. School leavers without an ISCED 3 qualification are regarded as early school leavers or drop outs. ISCED 4 is not used in this text.
![Page 43: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/43.jpg)
2 The educational gap and neighbourhood composition
17
In Section 2.2 previous research is reviewed; methods and data are described in
Section 2.3, the results in Section 2.4, and the conclusions and discussion are presented
in Section 2.5. The results point to a negative association between neighbourhood
ethnic composition and enrolment in academic secondary tracks. The negative slope of
the regression seems, however, to be mainly the result of the presence of a limited
number of zip‐code areas which are burdened by persistent and severe socio‐economic
challenges. In most zip‐code areas enrolments are at the level of national average
enrolments in academic secondary levels, and in many zip areas enrolments are
considerably higher.
2.2 Previous research
A stubborn problem with severe consequences
The achievement gap between students with a migrant background and native
students presents a persistent major challenge to education (Organization for Economic
Cooperation and Development, OECD, 2010). "Migrant students" is a general term used
in academic literature, and may include different countries of origin, different motives
for emigration and different countries of destination. In the Netherlands, the
educational achievement gap between native Dutch pupils and youths whose parents
immigrated from Turkey and Morocco is an important concern. The vast majority of
these students from Turkish or Moroccan background belong to the second generation:
they were themselves born in the Netherlands, but at least one parent was born in the
country of origin. Obviously, the educational achievement gap may have serious further
consequences, both for the individual and for society at large, as is extensively described
in the literature (e.g. Banks, 2007; Groot and Maassen van den Brink, 2003; Heckman,
2008; Belfield and Levin, 2007). The Country Report for the Netherlands within the
framework of the international OECD Reviews of Migrant Education (OECD, 2010),
furthermore, states explicitly (among other recommendations) that efforts to increase
enrolments in high levels of education among migrant students should be intensified,
and career support, notably for students in vocational programmes, should be
prioritized.
Assumedly owing to the complex nature of the gap, finding causal evidence for
effective strategies on the system level, appears to be a tough endeavour: the causal
![Page 44: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/44.jpg)
18
effect of, for example, increasing per capita funding of pupils, reducing class size,
extending school hours, compulsory summer schools or merit pay for teachers and
Principals, cannot be proven (McKinsey & Company, 2007; Dobbie and Fryer, 2009).
Although the current paper does not provide causal explanations for the achievement
gap either, this study aims to contribute to clarifying the position of high‐achieving
migrant pupils, since the majority of studies in the Netherlands, mainly seem to consider
those migrant pupils who lag behind. In addition, surprisingly, Heineck and Riphahn
(2007), in their study on intergenerational transmission of educational attainment of
birth cohorts in Germany from 1929 through 1978, find that: "In spite of massive public
policy interventions and education reforms our results yield no significant reduction in
the role of parental background for child outcomes over the past decades".
However, low educational attainment levels of migrant parents may possibly
have a different cause when compared to native‐born low SES parents: many migrant
parents, notably the older generation of immigrants, often did not have a chance to go
to school at all, because there was no school at a reasonable distances in their country
of origin. Moreover, elementary education may not have been compulsory at the time.
Taking this thought further, low attainment levels of migrant parents may to a lesser
degree point to lower capacities for learning to a lesser degree than could be the case
for low SES parents who actually did have access to schooling over the generations.
Interestingly, considering the intergenerational transmission of IQ, Anger and Heineck
(2009) find that the cognitive abilities of parents are positively correlated with their
children's cognitive skills, also after controlling for educational attainment levels and
family background.
Segregation
A key element in the current political discourse on the participation in Dutch
society of former labour migrants and their children is the assumed negative effect of
segregated residential areas and schools. The Dutch Study Centre for Mixed Schools
(Landelijk Kenniscentrum Gemengde Scholen)– an organization funded by the Ministry of
Education, Culture and Science– is dedicated to the further ethnical integration of
schools and the reduction of the number of schools with a vast majority of migrant
pupils or native Dutch pupils. Although this centre strongly believes in the positive
![Page 45: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/45.jpg)
2 The educational gap and neighbourhood composition
19
effects of mixing school populations, it recognizes, nonetheless, that international
research publications hardly present any findings that demonstrate a positive effect of
mixed schools on learning outcomes (Bakker, 2012). Interestingly, Margolis (2010)
suggests that we may need a broader understanding of local communities– not just the
level of segregation– in order to better understand the stubbornness of the
achievement gap.
According to Ladd et al. (2010), segregation levels of schools are high in the
Netherlands compared to the U.S., and schools tend to be even more segregated than
their surrounding neighbourhoods. They explore the question whether unlimited
freedom of school choice in the Netherlands, actually increases school segregation.
A further complicating factor in establishing the effects of school segregation on
learning outcomes may be that research publications on the negative effects of
segregated schools and residential areas are often based on research findings based on
impoverished neighbourhoods in metropolitan areas in the United States. Circumstances
of extreme poverty, a high incidence of teenage motherhood, widespread
unemployment, and criminality in these urban areas appear to be, however, hardly
comparable to conditions in cities in Europe or Canada according to Oreopoulos (2007).
Another important difference between Dutch and other European cities, on the one
hand, and metropolitan areas in the U.S., on the other hand, may be that the
geographical distances between low SES and affluent neighbourhoods are in general
considerably smaller in European cities. Interestingly, recent research publications show
that estimated positive effects of proximity to affluent neighbourhoods and good
schools may be larger than estimated negative effects of low average SES at
neighbourhood level per se (Durlauf, 2004; Glaeser, Resseger and Tobio, 2008; Massey,
2008).
In addition, Hartog and Zorlu (2009) find that segregation in the Netherlands may
be lower than the public discourse suggests: 95 % of the Moroccan Dutch population, for
example, live in a neighbourhood with less than 30 % Moroccan Dutch residents.
Finally, Dronkers (2010) finds that the diversity at school level, as expressed in
the number of different ethnic groups, has a stronger negative effect on attainment
levels and school outcomes than the total percentage of migrants per se. His findings
would imply that a high percentage of one specific ethnic group would have no or a
![Page 46: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/46.jpg)
20
limited negative effect on educational outcomes. A possible explanation for this effect
could be the "connectedness"4 at neighbourhood level (e.g. contacts among inhabitants,
between parents and school, between inhabitants and neighbourhood services)
resulting in close‐knit communities of people sharing the same background (e.g. Dobbie
and Fryer, 2009).
Desegregation
Extensive research into the effects of the desegregation of schools and
neighbourhoods sometimes leads to ambiguous or unexpected results. A large‐scale,
intensively investigated programme for desegregating severely disadvantaged
neighbourhoods, called the Moving to Opportunities programme5 (MTO), did not result
in any measurable effects on schooling (Oreopoulos, 2007). Furthermore, the enactment
of the 1964 Civil Rights Act in the United States, that barred the use of federal money for
segregated schools, initiated a synchronous desegregation of schools, thus creating
another large scale natural experiment. A study into the effects on black students who
were sent to former white schools (far more numerous than vice versa) after the 1964
Act, particularly shed a light on the negative effects of losing familiar neighbourhood
schools, where teachers were part of the black community (Stuart Wells, Jellison Holme,
Tijerina Revilla and Korantemaa Atanda, 2009). The causal effect on learning outcomes
of a comprehensive school and community approach, based on the idea that good
schools should be brought to disadvantaged children (as opposed to bussing children to
4 "Connectedness" is a term used in Ecology, referring to the density and specialization of niches (indicating a highly developed ecosystem) and the subsequent equilibrium of the system. The term is often borrowed to describe sociological systems, as well as the term "resilience" that describes the equilibrium between the capacity to change and the capacity to resist too much disturbance from the environment. 5 The Moving to Opportunity programme started in 1994 in Baltimore, Boston, Chicago, Los Angeles and New York. Originally the 1970 Housing and Urban development Act of 1970 allowed for the introduction of the Federal Experimental Housing Allowance programme.
![Page 47: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/47.jpg)
2 The educational gap and neighbourhood composition
21
high quality schools elsewhere) in the Harlem Children's Zone (HCZ)6 programme, could,
however, indeed be proven by Dobbie and Fryer (2009), notwithstanding the largely
stratified nature of the area.
Interestingly, as has been mentioned above, parents and children are not in any
way restricted in their choice of school in the Netherlands, which is completely free
under constitutional law; nor are they in any way limited in their choice by financial
constraints, since there is no tuition fee and all schools are equally funded by the
government. Migrant parents and children, however, overall tend to prefer schools with
larger groups of migrant pupils, both in elementary and in secondary education. By
implication, this would mean that compulsory desegregation policies would restrain
parental choice, while we as yet do not sufficiently understand possible beneficial
effects of specialized, albeit stratified neighbourhood schools.
Actually lived lives of individuals
Crul et al. (2012) started a bold endeavour to compare the actually lived lives of
people with an immigrant family history, but born in the country of destination (the
second generation), in six European countries (the Netherlands, Belgium, Sweden,
Austria, Switzerland, and Germany). The aim of their study was the assessment of the
relations between "the integration context", notably the educational system, and the
position of second generation migrants in the six countries. To this end, they
interviewed, by means of a questionnaire, 500 individuals belonging to the second
generation and 500 individuals from the comparison group (of native‐born parentage),
who were between 18 and 35 years old in 2007/2008. Interestingly, considering the
second generation of respondents with a Turkish family history in the Netherlands, Crul
et al. (ibid.) find that the odds of dropping out without a diploma are roughly equal to
6 In the early 1990s, HCZ ran a pilot project that brought a range of support services to a single block. The idea was to address all the problems that poor families were facing: from crumbling apartments to failing schools, from violent crime to chronic health problems. HCZ created a 10‐year business plan, to ensure its best‐practice programmes were operating as planned. HCZ was in the vanguard of non‐profit organizations that began carefully evaluating and tracking the results of their work. Those evaluation results enabled staff to see whether programmes were achieving their objectives, and to take corrective actions if they were not. In 1997, the agency began a network of programmes for a 24‐block area: the Harlem Children's Zone Project. In 2007, the Zone Project grew to almost 100 blocks. Today the Children's Zone® serves more than 8,000 children and 6,000 adults. Overall, the organization serves more than 10,000 children and more than 7,400 adults. The FY 2010 budget for the agency overall is over $75 million (text retrieved from HCZ website).
![Page 48: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/48.jpg)
22
the odds of being exceptionally successful and achieving far higher attainment levels
than their parents. They identified– among other factors– two specific characteristics of
the Dutch educational system that work out positively for pupils of migrant descent:
opportunities for upstreaming and downstreaming in secondary schools, and longer
alternative routes up the vocational ladder, eventually granting access to higher
education. The authors report that "between half to two thirds of Turkish second
generation occupy a stable lower‐ to upper‐middle‐class position…Some members have
taken a spectacular step in just one generation".
2.3 Methods and data
2.3.1 Selecting the data
The main goal of this study is a better understanding of the success of Moroccan
pupils; for this reason, access to the highest levels of secondary education at age 16, has
been defined as the measure of success for the purpose of this research. Typically, pupils
pass their final exam in these two secondary academic tracks at age 17 and 18,
respectively. At age 16, therefore, the number of pupils who have already passed the
final exam is negligible. The two highest secondary tracks have been combined in this
research since both tracks qualify the pupils for access to higher education; they will be
further referred to as academic tracks. At age 16 pupils are typically enrolled in Grade 4
of secondary schools. For this research, selecting pupils at age 16 was preferred,
however, and not enrolment in the fourth year of secondary school. At grade level pupils
may be of a different age because of grade repetition, and as the result of an age‐range
of almost 12 to almost 13 when entering secondary education; this would complicate
measurements of the proportion of the population of 16 year‐olds enrolled in academic
tracks at the zip‐code level.
This chapter focuses, in particular, on pupils with a Moroccan immigrant
background. The interpretation of research outcomes on educational achievement in
different ethnic groups is often complicated by the frequent use of combined migrant
groups, notably the indication Non‐western and Western immigrants. These widely used
combination of ethnic groups in research publications and national statistics, neglect
considerable differences per country of origin, notably the educational system, and
different reasons for emigration by distinct groups (de Heus and Dronkers; 2008,
![Page 49: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/49.jpg)
2 The educational gap and neighbourhood composition
23
Noguera, 2000). The vast majority of Moroccan Dutch 16 year‐olds belong to the second
generation; the measurements presented in this chapter do not distinguish between
first and second generation. In the Netherlands different groups in the population are
indicated by ethnicity rather than race, as is for instance more common in the United
States.
The four major cities (Amsterdam, Rotterdam, Utrecht and The Hague) by 4‐digit
zip‐code have been chosen as the spatial entity, for methodological reasons: the zip‐
code, although not without its own limitations, is internationally comparable and more
strictly defined than neighbourhood, district, community, or borough.
In middle‐sized and small cities, the number of Moroccan Dutch (and other
ethnicities) is often too small at zip level for reliable analyses; the majority of Moroccan
Dutch pupils live in the four major cities. Therefore, analyses have been carried out only
for the four major cities. Also to avoid too small numbers in some zips in the four major
cities, only zips with more than five Moroccan Dutch pupils were selected for analyses,
unless clearly indicated otherwise.
Separate measurements have been carried out for zip areas that qualify for extra
financial support by the Dutch Ministry of Housing, Spatial Planning and the
Environment7; these socio‐economically severely disadvantaged zips are referred to as
zips at risk in this text.
Apart from the main data set that was especially assembled for the purpose of
this research, also publicly accessible data published by Statistics Netherlands (the
National Statistical Office) and the Central Financial Institution (CFI, part of the Dutch
Ministry of Education, Culture and Science) have been reworked. These tables, based on
these public data bases, are presented in the Appendices (2.6).
7 In 2007 the Ministry of Housing, Communities and Integration (Tweede Kamer der Staten Generaal, (2007) started a programme to support, with extra funding, neighbourhoods with an accumulation of problems. Qualification for extra funding is based on a list of possible disadvantages. Among the neighbourhoods that participate in this programme are many zips with large numbers of inhabitants with a migrant background. A list of 340 zip‐codes (out of around 4800 in total in the Netherlands) that qualified for extra support was selected by the Ministry based on four groups of indicators:
Socio‐economic indicators like average income, employment and educational attainment;
Infrastructural indicators like size, age, condition and price of houses;
Social problems like vandalism, harassment and perceived unsafe situations;
Physical annoyances like noise, traffic, street litter and pollution; The Ministry gave priority to a selection of 40 communities out of these 340, covering in total 83 zip‐codes, among these 59 in the four major cities. At least five Moroccan pupils of 16 years old live in 47 out of these 59 zip‐codes in the major cities; these are used for measurements in this paper.
![Page 50: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/50.jpg)
24
2.3.2 Assembling the data set
The relevant data necessary for the measurement of the correlation between the
percentage of 16 year‐old Dutch Moroccans per zip‐code, and their access to academic
tracks in secondary education, are managed by two different institutions, CFI and
Statistics Netherlands: all address (zip) information that can be traced back to the
individual is collected by Statistics Netherlands whereas all educational data are
registered in what is called the BRON8 database by CFI9.
The data set CFI made available (reference date 1 October 2008) contains the
following information: The data set covers all 16 year‐olds by ethnicity, who are enrolled
in academic tracks in secondary education, by 4‐digit zip‐code level in the four major
cities. Nationwide the group of 16 year‐olds consists of around 315.000 individuals;
around 28.000 live in the four major cities.
The above data cannot be used to measure the percentage of Moroccan Dutch
pupils enrolled in academic levels, as a share of their own ethnic group at zip level, since
the data do not cover the total number of 16 year‐old Moroccan Dutch per zip‐code.
To measure the share that enrols in secondary academic tracks, the CFI data
were linked at the 4‐digit zip level to a tailor‐made data set made available by Statistics
Netherlands, with the following description:
The data set covers all 16 year‐olds (not only those in academic tracks) by 4‐digit
zip‐code and ethnicity.
Individuals are regarded as migrant pupils when at least one parent was born
abroad (definition Statistics Netherlands).
The file contains the year and the month of birth, but not the day. Therefore all
individuals born on 1 October 1990, up to and including 30 September 1990, are
included in this file.
The reference date is 26 September 2008
8 BRON: Dutch acronym: Basis Register Onderwijs Nummer. In the new millennium the Dutch government started the phased introduction of one single registration number for all citizens, thus combining for instance registration numbers for education, family related data, income and social security, further training, medical and police records. 9 During the current research, CFI has been in the process of merging, which has resulted in a new agency, called DUO (Dutch acronym, Dienst Uitvoering Onderwijs). DUO is the office for all financial affairs in education, the organization of final exams, and the management of student funding) and is part of the Ministry of Education, Culture and Science.
![Page 51: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/51.jpg)
2 The educational gap and neighbourhood composition
25
By merging these two data sets, the share of Moroccan youth at zip‐code level
could be calculated, as well as the percentage of them enrolled on academic tracks. The
data set contains 2947 16 year‐old Moroccan Dutch living in the four major cities: 1290
in Amsterdam, 719 in Rotterdam, 501 in Utrecht, 437 in The Hague.
2.3.3 Data from public files
CFI and Statistics Netherlands also allow public access to parts of their databases,
and regularly publish updated statistics considering, for example, education,
demography, economic development and the labour market. The publicly accessible site
Statline by Statistics Netherlands offers tools to adapt public data to specific research
questions. Some of these tables and figures were reworked as a reference for findings in
this research. Figures based on publicly available data are included in the Appendices
and show: 1) rising levels of educational attainment among people with a Moroccan
immigrant background age 25 to 65– 2) decreasing percentages of people with
elementary education only– 3) rising numbers of people with at least an initial
qualification for the labour market10; and 4) raising enrolment levels in academic tracks
among 16 year‐old pupils with a migrant background between 2003‐2009. The tables in
the Appendices show a sharp rise in educational attainment levels for all ethnicities
when the cohorts 55‐65 year‐olds are compared to 25‐35 year‐olds (Appendices 2‐1, 2‐2,
2‐3, and 2‐4). This increase is, however, different for distinct ethnic groups. 40% among
native‐born in the cohort 55‐65 year‐olds, did not accomplish ISCED 3 level, the
minimum requirement for the labour market; among the 25‐35 year‐old native‐born,
16% did not reach this level, a decrease of 24 percentage points (Appendix 2‐1). People
with a migrant background improved average levels of schooling far stronger: 80%
among Moroccan Dutch 55‐65 year‐olds against 34% among 25‐35 year‐olds did not
accomplish ISCED 3, a decrease of 46 percentage points (Appendix 2‐1).
Correspondingly, the percentage with only elementary education as the highest
10 Initial qualification for the labour market is a standard introduced by the OECD (Organization for Economic Co‐operation and Development) and is defined as the completion of at least the ISCED 3 level. The ISCED qualification (UNESCO 1997) is precisely introduced to make international agreements possible on required educational attainment levels, thus facilitating the international labour market and stimulating student mobility. The OECD member states agreed that ISCED 3 is the basic qualification for entering the labour market. Of course ISCED 3 refers to different nomenclature for educational tracks in the member states; in the Netherlands it refers to a diploma at pre‐university (VWO), senior general (HAVO) level or advanced levels of senior secondary vocational education (MBO).
![Page 52: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/52.jpg)
26
completed level decreases strongly for all groups‐ but mostly for Turkish and Moroccan
Dutch (Appendix 2‐3). Furthermore, among Moroccan Dutch 25‐35 year‐olds, 16%
obtained a bachelor or master's degree or a PhD (38% native‐born in the same cohort
reached this level); this was 0% among 55‐65 year‐olds Moroccan immigrants (Appendix
2‐1).
2.4 Results
The tables and figures presented in this Section follow the three research
questions:
Question 1, which considers the association between ethnic composition at zip
level and access to academic secondary tracks: Table 2‐1, as an introduction,
presents an overview of the number of 16 year‐olds at zip‐code level, and the
share of them who are of Moroccan descent. Figure 2‐1 depicts the regression of
enrolments among youth with a Moroccan background in academic tracks, and
ethnic composition at the zip‐code level, in the four major cities combined;
Figure 2‐2 shows regressions for enrolments separately for the four major cities.
Question 2, which looks specifically at zip areas at risk, with severe socio‐
economic challenges: Figure 2‐3 illustrates enrolments in zips at risk; Table 2‐3
compares all zips, and zips at risk. Regression coefficients are listed in Table 2‐2.
Question 3, which considers the total percentage of immigrants at zip level (all
ethnicities) and access to academic secondary tracks: Table 2‐4 presents the
correlation between enrolments in academic tracks among Moroccan Dutch
youth, and the total ethnic composition (including youths with a Turkish,
Antillean/Surinamese or other immigrant background) of the zip‐code area
where they live.
2.4.1 Enrolments in academic tracks among Moroccan Dutch, at zip‐code level
Table 2‐1, as an overview, shows the average number of 16 year‐olds per zip, and
the average number and percentage of them who are of Moroccan descent. Amsterdam
has the highest number of youngsters of Moroccan descent, Utrecht the highest average
percentage at zip‐code level.
![Page 53: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/53.jpg)
2 The educational gap and neighbourhood composition
27
Table 2‐1: Total average of 16 year‐olds per zip in the four major cities, and average number and percentage of 16 year‐old Moroccan Dutch among this group
Amsterdam Rotterdam Utrecht The Hague Total
# zips 42 37 22 20 121
Total average of 16 year‐olds at zip level 125 135 70 133 111
Average # Moroccan 16 year‐olds at zip level 26 18 21 19 19
% Moroccan among 16 year‐olds at zip level 21% 13% 30% 14% 17%
Notes: Only zips with > 5 Moroccan 16 year‐olds included.
Figure 2‐1: Data scatter, linear regression (OLS) and LOWESS regression of the percentage of 16 year‐old Moroccans enrolled in academic secondary tracks per zip on the percentage of Moroccan Dutch 16 year‐olds as a share of the total number of 16 year‐olds at zip level in the four major cities in the Netherlands
Note: Only zips > 5 Moroccans of 16 years old have been included.
Figure 2‐1 shows the regression (LOWESS and linear) of the percentage of 16
year‐old Moroccan Dutch pupils enrolled in academic secondary tracks on the share of
Moroccan Dutch among 16 year‐olds at zip‐code level. LOWESS weights the data point
by point within a moving band containing 40% (in this case) of the data set, and has
been added to check that the linear fitted line was not biased because of the relatively
small number of data points (zips). Figure 2‐1 shows that the association between zip
020
4060
80%
in a
cade
mic
trac
ks
0 20 40 60% Moroccan 16 year olds per ZIP
![Page 54: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/54.jpg)
28
composition and enrolments in academic tracks is negative (coefficient ‐0.21**, Table
2‐2). Measurements show a relatively high dispersion of data points in zips where up to
20% of 16 year‐old Moroccans live. As a reference for these findings, an overview of the
national dynamics in enrolment levels among 16 year‐old students with a migrant
background (2003‐2009) has been included in Appendix 2‐4.
Figure 2‐2: Data scatter and linear regression (OLS) of the percentage of 16 year‐old Moroccans enrolled in secondary academic tracks on the share of Moroccan Dutch 16 year‐olds at zip level, separately for the four major cities in the Netherlands
Enrolments in academic tracks among students of Moroccan descent, separately
for the four major cities in Figure 2‐2, indicate higher access in Amsterdam and
Rotterdam, as compared to Utrecht and The Hague. An explanation for these differences
in access between the four major cities is beyond the scope of this research, since the
data do not allow for further socio‐economic analyses that might explain these
differences. Appendix 2‐5, based on new educational data on current enrolments in
secondary Year 3, has been added, however, because this new data set does indeed
cover socio‐economic indicators: Appendix 2‐5 confirms that enrolments in secondary
020
4060
80%
in a
cade
mic
trac
ks
0 20 40 60% Moroccan 16 year olds per ZIP
Amsterdam RotterdamUtrecht The Hague
![Page 55: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/55.jpg)
2 The educational gap and neighbourhood composition
29
tracks are higher in Amsterdam, but does not indicate a higher average SES at zip level
as a possible explanation; neither does the relatively low average SES at zip level in
Rotterdam in Appendix 2‐5, offer more insight into the higher enrolment levels for
Rotterdam in Figure 2‐2. Finally, as a reference, Appendix 2‐6 shows the association
between ethnic composition at zip level and enrolments among native Dutch students.
Although their average enrolments are overall higher than in the case of migrant
students, enrolments among native Dutch decline with increasing percentages of
migrant youths at zip level. A further analysis of socio‐economic indicators that might
explain these findings is presented in Chapter 4 of this thesis.
2.4.2 Enrolments in academic tracks in zips at risk
In 2007 the Ministry of Housing, Communities and Integration earmarked a
budget for the improvement of living conditions in neighbourhoods with accumulated
socio‐economic problems. Priority was given to a selection of 40 neighbourhoods,
covering in total 83 zip‐codes, among which 59 were in the four major cities. When only
zip‐codes with more than five 16 year‐old Moroccan Dutch are considered, 52 zips at risk
remain for measurements in this Section. Young people in these zips are supposedly
extra challenged to find their way to academic tracks in secondary education, because
average levels of schooling among parents in zips at risk are low, household income is
below average, and unemployment is high. General theory identifies these
circumstances as negative predictors for educational attainment levels.
Figure 2‐3 shows a negative slope of the regression line (coefficient ‐0.25**,
Table 2‐2), much like the measurements on all zips in Figure 2‐1, with a flattening out of
the LOWESS regression line for zips with around 20‐40% of 16 year‐olds of Moroccan
descent. Zips with lower percentages of Moroccan Dutch 16 year‐olds, do better than
the national average for this group (see Appendix 2‐4), while those who live in zips with
more than 40% of Moroccan Dutch 16 year‐olds fall behind. The socio‐economic
characteristics for zips at risk mentioned above are negatively correlated to average
levels of schooling in many studies (e.g. De Witte, 2010). Although causal explanations
for the enrolments in academic secondary tracks among Moroccan pupils are beyond
the scope of this research, the results do not indicate that students are less likely to
![Page 56: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/56.jpg)
30
enrol in academic tracks in zips at risk when compared to the average enrolments in all
zips combined (Figure 2‐1).
Figure 2‐3: Data scatter, linear regression (OLS) and LOWESS regression of the percentage of 16 year‐old Moroccans enrolled in academic secondary tracks per zip at risk on the share of Moroccan Dutch 16 year‐olds at zip level in the four major cities in the Netherlands.
Table 2‐2: Regression coefficients of the percentage enrolments in academic tracks among 16 year‐old Moroccans on the share of Moroccan 16 year olds at zip‐code level
All zips At risk Amsterdam Rotterdam Utrecht The Hague
% Moroccans ‐0.21** ‐0.25** ‐0.23 ‐0.26 ‐0.22 ‐0.14
(0.09) (0.10) (0.18) (0.26) (0.15) (0.27)
Constant 25.16*** 27.52*** 28.28*** 26.64*** 22.43*** 20.23***
(2.28) (3.09) (4.91) (4.70) (4.91) (4.47)
N 121 52 42 37 22 20
Mean 20.7 20.9 22.8 22.5 15.9 18.2
Notes: Standard errors in parentheses. */**/*** denote significance at the 10/5/1% confidence level.
020
4060
80%
in a
cade
mic
trac
ks
0 20 40 60% Moroccan 16 year olds per ZIP at risk
![Page 57: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/57.jpg)
2 The educational gap and neighbourhood composition
31
Table 2‐3: Correlation of the percentage of 16 year‐old Moroccans per zip, and the percentage among them in secondary academic tracks, for all zips and for zips at risk separately
City Correlation P‐value #zips
All zips Zips at risk All zips Zips at risk All zips Zips at risk
Amsterdam ‐0.194 0.032 0.219 0.910 42 15
Rotterdam ‐0.168 ‐0.284 0.321 0.200 37 22
Utrecht ‐0.313 ‐0.429 0.156 0.289 22 8
The Hague ‐0.122 ‐0.392 0.607 0.385 20 7
Total ‐0.204 ‐0.322 0.025 0.020 121 52
Notes: Significant correlations p≤ 0.05 in bold.
In addition to the regression analyses, the correlation between the percentage of
16 year‐old Moroccan Dutch and their enrolments in academic tracks was measured as
well (Table 2‐3).
Caution is needed because measurements in Table 2‐3 are based on small
numbers of zips, notably the numbers of zips at risk. Note that in the case of Rotterdam,
22 out of 37 zip areas qualified for extra government funding because of persistent
socio‐economic disadvantages. No significant correlations were found for 16 year‐old
Moroccans and enrolments among them in academic tracks, both in all zips– and in zips
at risk, when considering the four cities separately; a significant negative correlation was
found, however, in total counts of the four cities combined, probably as a result of the
larger total number of zips (and a smaller standard error).
2.4.3 The correlation between the total percentage of migrants at zip level and enrolments of Moroccan Dutch
Table 2‐4 presents correlations between enrolments of Moroccan Dutch and the
total percentage of 16 year‐old migrants at zip level (all immigrant groups combined,
Moroccan, Turkish, Surinamese/Antillean and other). Note that in Table 2‐4, unlike all
the other main measurements, all zips were included where at least one 16 year‐old
individual of Moroccan descent lives (in other measurements only zips with > 5
Moroccan 16 year‐olds were included). Only in Rotterdam is the correlation negative
and significant (‐0.416, P‐value of 0.002). For the other three cities, the correlation is
smaller and not significant.
![Page 58: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/58.jpg)
32
Table 2‐4: Correlations between total percentage of all immigrants per zip‐ and the percentage of 16 year‐old Moroccans in secondary academic tracks, separately for the four major cities.
City Correlation P‐value # zips
Amsterdam ‐0.074 0.579 59
Rotterdam ‐0.416 0.002 52
Utrecht ‐0.237 0.178 34
The Hague 0.150 0.338 43
Total ‐0.086 0.241 188
Notes: Significant correlation p≤0.05 in bold
2.5 Conclusions and discussion
Three questions have been explored in the current chapter: 1) Is enrolment in
academic secondary tracks of 16 year‐olds with a Moroccan background, related to the
percentage at zip level of Moroccan 16 year‐old youngsters; 2) When considering only
zips with persistent socio‐economic challenges, do the results differ from measurements
in all zips; and 3) Is de total percentage of youths with an immigrant background
associated with enrolments at academic secondary levels of youths with a Moroccan
background.
A major assumption underlying policymaking at the government and the
municipal level is the assumed negative effect of segregated neighbourhoods on
educational outcomes. The results presented in this chapter nuance this assumption:
enrolment percentages at zip‐code level show a large dispersion, leading to a question
for further research: Why do zip areas with the same percentages of 16 year‐old youths
of Moroccan descent, have quite different enrolment rates. In zip areas with up to 20%
16 year‐old youngsters of Moroccan descent, average enrolments are on the level of
average national enrolments for their group (20.3% in 2009). Overall, however, the
association is negative, with a regression coefficient of ‐0.21** (all zips considered): for
every 10% increase of Moroccan 16 year‐olds per zip, enrolments in secondary tracks
decrease by 2.1 percentage points. The decrease is 2.5 percentage points considering
zips at risk.
Earlier analyses by Hartog and Zorlu (2009) show that 95% of people of
Moroccan descent live in a neighbourhood with ≤ 30% of people who have this same
background, indicating that segregation is more moderate than public and political
debate suggests. Considering the zips with ≤ 30% Moroccan youths at zip level, no
![Page 59: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/59.jpg)
2 The educational gap and neighbourhood composition
33
convincing association was found between zip composition and enrolments in high level
tracks.
The findings in the current study do not ignore there may be valid reasons for
school desegregation: learning and living together in mixed schools may add in
important ways to skills and valuable experiences that enhance the quality of the future
labour force in a multicultural society. However, in order to be able to make balanced
policy decisions, further research should also elucidate the effect of specific school
policies in schools with a majority of migrant pupils (e.g. specialized further teacher
training, cooperatives with migrant parents, and joint community‐based programmes) as
the effect of desegregation. Furthermore, unavoidably, desegregation policies, like
centralized school assignment, would limit free school choice in the process. Finally, the
added value of specialized, albeit segregated, schools may change over time: whereas
societal groups that are in the emancipatory process of substantially increasing their
educational attainment levels may benefit from specialized schools, over time a more
balanced school composition could become more important.
Extrapolating the index of the increase in enrolments in academic tracks among
Moroccan pupils between 2003 and 2009 (31%: see Appendix 2‐4), and assuming this
speed of increase could be sustained, Moroccan youths could be expected to catch up
with current enrolments among native‐born students around the year 2025. However,
attainment levels of such a large extent affect both the prospects of the individual and
the potential of the workforce as a whole, so further research into strategies that could
speed up the closing of the gap remain crucially important.
While the overall ethnic composition at zip level (all ethnicities, including
Moroccan background) does not show a significant correlation with secondary academic
enrolments, there is one exception: in Rotterdam the correlation is significant and
negative (‐0.416, Table 2‐4). Interestingly, this finding corresponds to other research
which demonstrates that Rotterdam seems to face extra challenges: for example, early
school leaving (De Witte, 2010). At first sight, the findings for Rotterdam in Table 2‐4
seem to be in contrast to the findings in Figure 2‐2: in Figure 2‐2 enrolments in academic
tracks were regressed on the percentage of Moroccan youths at zip‐code level, and
show higher enrolments both in Amsterdam and Rotterdam in comparison with the
other two cities. However, in Table 2‐4, unlike in Figure 2‐2, the correlation between
![Page 60: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/60.jpg)
34
enrolments and the total percentage of youths of migrant descent is considered. A
possible explanation for these divergent outcomes may be the different ethnic
composition at zip‐code level in Rotterdam: Moroccan youngsters in Rotterdam tend to
live in neighbourhoods with on average higher percentages of people with a migrant
background (all ethnicities considered) than is the case in the other three cities, but with
a lower average percentage of Moroccan 16 year‐olds than is the case in the other three
cities. Furthermore, interestingly, updated BRON data (listing enrolments in academic
tracks in Year 3 of secondary school in 2011/2012) show a considerably lower average
SES index for neighbourhoods in Rotterdam (Appendix 2‐5). Their enrolments in
academic tracks, as measured at the entrance of Year 3 in secondary schools, seem
substantially lower (16%) than is, for example, the case for Amsterdam (27%), although
Utrecht shows the same low percentage of enrolments as Rotterdam.
In summary, most Moroccan pupils live in neighbourhoods with many
inhabitants of migrant descent. Whether Moroccan pupils would have been more
successful in qualifying for high secondary tracks, as a result of desegregation policies
through restrictions of free school choice or housing policies, is a very intricate question.
Such policies would be expected to have far‐reaching consequences for families and
individuals. Moreover, neither can we estimate adequately whether these interventions
would indeed speed up the closing of the educational achievement gap. Although this
study demonstrates that an increase in the ethnic segregation of Moroccan youths at zip
level significantly reduces enrolments in academic tracks by around 2 percentage points
per 10% increase of Moroccan youths, the size of the effect may not counterbalance the
impact of desegregation policies of neighbourhoods and schools, especially not vis‐à‐vis
the deeply valued right to the free choice of where to live and where to go to school in
the Netherlands.
The findings presented in the current chapter have inspired further research
projects, which are presented in the following chapters of this thesis. One project is
based on a detailed analysis of individual school histories of youth living in three zip
areas in the district Amsterdam‐West, which face substantial socio‐economic challenges
(Chapter 3). Furthermore, the analysis of the association between the ethnic
composition of the area of residence and school choice also inspired an investigation of
the distance travelled to school, controlled for a rich set of socio‐economic indicators
![Page 61: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/61.jpg)
2 The educational gap and neighbourhood composition
35
(among which ethnic background). Since school choice is free and not restricted by any
financial barriers in the Netherlands, distance to school may reveal interesting
differences in patterns of school choice (Chapter 4).
Finally, the current research raises some important qualitative questions, for
example, on the motivations of migrant students to opt for academic tracks, and, on
availability of information for parents on the consequences of school choice with
reference to future success on the labour market of their children.
2.6 Appendices
The figures presented below may offer relevant additional information to the
main findings presented in Section 2.4 (Results). These statistics are not based on the
tailor‐made data set used for main measurements in this chapter, but either on the total
BRON database, or the publicly accessible data source Statline11, managed by Statistics
Netherlands.
Increasing educational attainment
The research questions in this paper put a focus on 16 year‐old Moroccan Dutch
in academic tracks in secondary education. Moroccan labour immigration started in the
1960s. Most Moroccan immigrants came to the Netherlands to be employed, at that
time, in unskilled labour. Their average educational attainment levels were (very) low,
and so were the levels of schooling of their wives and children who followed in the years
to come. The table in Appendix 2‐1 shows how attainment levels have increased over
time. The completion of ISCED 3 is the minimum requirement or initial qualification for
the labour market, a standard agreed among OECD countries.
As mentioned above, the completion of ISCED 3 is an important boundary level
for access to the labour market. Between Moroccan Dutch aged 55‐65 and aged 25‐35,
the initial qualification for the labour market has increased by 47 percentage points
(Appendix 2‐2). The percentage of Moroccan Dutch with only elementary education as
the highest completed level has decreased accordingly, by 48 % points, as is illustrated in
Appendix 2‐3.
11 www.cbs.nl
![Page 62: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/62.jpg)
36
Appendix 2‐4 shows that in 2008 the national average enrolment in academic
tracks was 19% among Moroccan Dutch, against 43.9% among native‐born pupils. The
index for the relative increase in enrolments in academic tracks between 2003 and 2009
is, however, different for native‐born and Moroccan migrants: the relative increase for
native‐born over this period is 17% (100 x (45.1 : 38.5) = 117), and 31% for Moroccan
youths (100 x (20.5 : 15.6)= 131).
Appendix 2‐1: Highest completed level of education, by age and ethnicity (2007)
Ethnicity Level of education Age
25‐35 35‐45 45‐55 55‐65
Native Dutch Primary education 3% 4% 8% 12%
ISCED 2 13% 18% 22% 28%
ISCED 3 45% 47% 41% 35%
Bachelor 25% 19% 19% 16%
Masters, PhD 13% 11% 9% 8%
Unknown 1% 1% 0% 0%
Moroccan Primary education 12% 24% 50% 60%
ISCED 2 22% 25% 20% 20%
ISCED 3 46% 36% 30% 20%
Bachelor 11% 3% 0% 0%
Masters, PhD 5% 7% 0% 0%
Unknown 5% 5% 0% 0%
Turkish Primary education 15% 27% 39% 64%
ISCED 2 29% 22% 21% 18%
ISCED 3 42% 40% 25% 18%
Bachelor 8% 3% 0% 0%
Masters, PhD 5% 5% 7% 0%
Unknown 0% 3% 7% 0%
Notes: ISCED 2: VMBO pre‐vocational secondary education; ISCED 3: MBO, level 4 (senior general vocational education); HAVO (senior general education); VWO (pre‐university education).
Source: Statistics Netherlands, Statline, based on the EBB questionnaire to the labour force.
![Page 63: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/63.jpg)
2 The educational gap and neighbourhood composition
37
Appendix 2‐2: Percentage of pupils with initial qualification for the labour market (ISCED 3 and higher), by ethnicity and age cohort (2007)
Source: Statistics Netherlands, treated.
Appendix 2‐3: Percentage of pupils with primary education as highest completed level, by ethnicity and age cohort (2007)
Source: Statistics Netherlands, treated.
0
20
40
60
80
100
Native Dutch Moroccan Turkish
% with basic qualification
Ethnicity
25-35
35-45
45-55
55-65
0
20
40
60
80
100
Native Dutch Moroccan Turkish
% with primary education only
Ethnicity
25-35
35-45
45-55
55-65
![Page 64: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/64.jpg)
38
Appendix 2‐4: Enrolment of 16 year‐olds by ethnicity in senior general education (HAVO, ISCED 3) and pre‐university education (VWO, ISCED 3)
2003 2004 2005 2006 2007 2008 2009
Total Senior general 20.1% 20.4% 21.1% 21.3% 22.0% 22.3% 22.7%
Pre‐university 16.1% 16.9% 17.5% 18.0% 18.8% 19.1% 19.5%
combined 36.2% 37.3% 38.7% 39.3% 40.8% 41.3% 42.3%
Native‐born Senior general 21.4% 21.7% 22.4% 22.6% 23.4% 23.5% 24.2%
Pre‐university 17.1% 18.0% 18.7% 19.2% 20.0% 20.4% 20.9%
combined 38.5% 39.7% 41.1% 41.8% 43.4% 43.9% 45.1%
Moroccan Senior general 10.9% 10.8% 12.2% 12.2% 12.0% 13.4% 13.5%
Pre‐university 4.7% 5.1% 5.5% 5.5% 6.5% 5.6% 7.0%
combined 15.6% 15.9% 17.7% 17.7% 18.5% 19.0% 20.5%
Turkish Senior general 11.8% 11.9% 12.7% 12.3% 12.8% 13.4% 13.9%
Pre‐university 4.9% 6.0% 5.6% 6.5% 6.8% 6.7% 6.7%
combined 16.7% 17.8% 18.3% 18.7% 19.6% 20.0% 20.6%
Notes: The two secondary tracks combined are referred to in the text as academic tracks, both qualifying pupils for access to higher education.
Source: CFI, BRON data. I am deeply grateful to Gert Korteweg of the Ministry of Education, Culture and Science for his technical help in assembling this table and his collegial advice.
![Page 65: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/65.jpg)
2 The educational gap and neighbourhood composition
39
Appendix 2‐5: Percentage of pupils enrolled in academic track in Year 3 of secondary school (school year 2011/2012), mean neighbourhood SES (scale ‐4 to +4), and percentage of pupils eligible for weighted student funding when they were enrolled in elementary education.
Amsterdam
Ethnicity % in academic track in Neighbourhood SES % weighted N
Native‐born 62 0.45 2 1737
Moroccan 27 ‐0.89 60 968
Turkish 28 ‐0.83 56 538
Surinamese/Antillean 23 ‐0.52 31 762
Other immigrant 44 ‐0.30 33 1054
Total 42 ‐0.24 30 5059
Rotterdam
Ethnicity % in academic track Neighbourhood SES % weighted N
Native‐born 47 0.38 4 1777
Moroccan 16 ‐1.36 56 592
Turkish 19 ‐1.27 52 659
Surinamese/Antillean 19 ‐0.84 39 846
Other immigrant 37 ‐0.62 39 900
Total 32 ‐0.47 30 4774
The Hague
Ethnicity % in academic track Neighbourhood SES % weighted N
Native‐born 56 1.47 2 1668
Moroccan 22 ‐1.01 56 454
Turkish 16 ‐0.87 48 569
Surinamese/Antillean 33 ‐0.00 30 528
Other immigrant 43 0.51 24 699
Total 41 0.47 23 3918
Utrecht
Ethnicity % in academic track Neighbourhood SES % weighted N
Native‐born 53 1.29 2 1107
Moroccan 16 ‐0.16 57 419
Turkish 26 ‐0.11 52 205
Surinamese/Antillean 34 0.82 20 91
Other immigrant 56 0.91 21 219
Total 42 0.79 21 2041
Source: BRON data.
![Page 66: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/66.jpg)
40
Appendix 2‐6: Data scatter, linear (OLS) and LOWESS regression of the percentage of 16 year‐old native Dutch pupils on the percentage of 16 year‐olds of Moroccan descent at zip level in the four major cities in the Netherlands
020
4060
80 %
in a
cade
mic
trac
ks a
mon
g N
ativ
e D
utch
20 40 60 80 100% migrants per ZIP
![Page 67: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/67.jpg)
41
3 School switching, a concern for migrant pupils
Justice and injustice are also called equity and inequity…
Baruch de Spinoza, 1670
3.1 Introduction
Enhancing continuity and stability in secondary school careers by the reduction
of switching between schools might, according to the literature reviewed in this chapter,
be a promising approach to speed up the reduction of the current educational
achievement gap between native Dutch pupils and pupils with an immigrant
background. Switching, however, seems to be a relatively understudied subject in the
Netherlands. This explorative study sets out to take stock of patterns of switching
among different ethnic groups, notably among pupils with a Moroccan immigrant
background, in three zip‐code areas in the district of Amsterdam‐West, compared to the
rest of Amsterdam.
During 1969s and 1979s, large numbers of Turkish and Moroccan immigrants
entered the Dutch labour market, mostly to be employed in unskilled work. In the
following decades, most of these workers were followed by their wives and children.
Therefore, the vast majority of their children and grandchildren, who are currently
![Page 68: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/68.jpg)
42
enrolled in secondary education, belong to the second‐generation immigrants12.
Enrolments of migrant pupils in the two highest levels of secondary education (the
academic tracks, specified further below) have continuously increased over the past
decades, but so did enrolment percentages of native‐born pupils. The width of the gap,
therefore, remains substantial: in 2009, nationally, 45.1% of 16 year‐old native Dutch
pupils were enrolled in secondary academic tracks, against 20.5% among pupils form
Moroccan descent and 20.6% of pupils from Turkish descent (Ministry of Education,
Culture and Science).
In general, in this research references to a migrant background indicate second‐
generation immigrants; according to the definition used by Statistics Netherlands, this
means that a person was born in the Netherlands, but has at least one parent who was
born abroad. The focus group in this study are pupils belonging to the second generation
Moroccan immigrants. Two levels of school organization are distinguished in this text:
the individual school, on the one hand, and the cluster level of schools that share a
common governing board, on the other hand. This distinction has been made in order to
be able to measure all transfers between schools, as well as school switching that
includes transfer across cluster borders. We will also consider upward and downward
mobility (referring to moving to another track within the multi‐tracked system for
secondary education in the Netherlands) and its association with switching. Students can
change to another track within the same school, or do this after switching to another
school. When a pupil moves up to a higher track in the same school, this is not counted
as switching; switching implies the transfer to another school
3.1.1 Young people with a migrant background in Amsterdam‐West
In the three zip‐code areas studied in this chapter, inhabitants with a Moroccan
background form the largest group. In the words of Crul and Holdaway (2009):
12 In this text the following definitions used by Statistics Netherlands (the National Statistics Office, www.cbs.nl) are used:
First generation immigrant: a person born abroad with at least one parent born abroad.
Second generation immigrant: a person born in the Netherlands with at least one parent born abroad.
Foreign background: a person with at least one parent born abroad.
Western immigrant: a person originating from a country in Europe (exclusive of Turkey), North America, Oceania, Indonesia or Japan.
Non‐western immigrant: someone originating from Africa, South America, Asia (exclusive of Indonesia and Japan) or Turkey.
![Page 69: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/69.jpg)
3 School switching, a concern for migrant pupils
43
"Amsterdam‐West is the 'capital' of Dutch Moroccans in the Netherlands (…) the second
largest national origin group in the city. It is also the fastest growing group (…) more
than 40% are younger than 19 years old (…)"
The achievement gap between native‐born and migrant pupils in the
Netherlands, and more in general between children from low SES parents and children
from more affluent families in many countries all over the world, has become a complex
and persistent problem. Moreover, the consequences of the gap are serious and costly,
both for the individual and for society at large, hence the abundance of research into the
nature and possible solutions of this problem. As will be further described below,
because of this complexity, the causal effect of widely implemented interventions to
narrow the gap– for example, by reducing class size, increasing per capita funding of
pupils, merit pay for teachers and Principals, intensifying testing and making schools
more accountable– could not unequivocally be proved. Moreover, research publications
that actually do describe causal effects on children's chances of school success, often
identify circumstances that are outside the span of control of schools, such as low‐
educated parents, or impoverished residential areas with an accumulation of socio‐
economic problems; such circumstances, in most cases, have to be regarded as a given
by school leaders. As a further illustration of the stubbornness of the problem, the width
of the achievement gap between native‐born and migrant pupils in secondary academic
tracks in the Netherlands, apparently so far could not have been reduced sufficiently by
large scale interventions by the Dutch government, e.g. extra funding for all low SES
pupils, equal funding of all schools by the government, and completely free school
choice (according to the constitutional law of 1917). Poor neighbourhoods are not
further held back by underfunded schools, since these are not dependent on local tax
revenues or tuition fees.
The underrepresentation in secondary academic tracks of the two largest
migrant groups in the Netherlands– people of Turkish and Moroccan descent– affects a
growing number of young people: nationally, in 2008 16% of all individuals aged 0‐ 20
years had a migrant background; in the four major cities (Amsterdam, Rotterdam,
Utrecht and The Hague), around 50% of the young people of that age are of immigrant
descent. Most Turkish and Moroccan families have a weaker socio‐economic position, a
considerable lower average family income compared to the native Dutch, and live more
![Page 70: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/70.jpg)
44
often below the poverty line (Herweijer, 2009). Their opportunities to fully participate in
society in meaningful ways, seem further challenged, in recent years by the polarizing
nature of the public debate on a culturally diverse society. According to Schinkel (2008),
in his critique on this polarizing debate, migrants tend to be regarded as being "outside"
Dutch society.
Schools in the major cities in the Netherlands are fully aware of their crucial role
in providing the best opportunities for migrant pupils through high‐quality education.
However, as mentioned above, schools are confronted with evidence‐based results from
academic research into factors that seem to explain the gap, which are mostly outside
the influence of school management and policies. Interestingly, however, research into
the effect of instability in school careers, caused by repeated switching of pupils to other
schools, may offer promising new opportunities: reducing student mobility lies to a large
extent within the span of control of individual schools– with the exception, of course, of
changing schools because parents move to another area. While the currently available
data for this research cannot identify statistical causal effects, the findings may offer
further practicable insights into the nature and pattern of switching.
This chapter proceeds to answer the following questions:
What is the frequency of school switching among 16‐22 year‐olds, with a focus
on pupils of Moroccan descent, living in three socio‐economically disadvantaged
zip areas in Amsterdam, considering separately the transfer between single
schools, and across clustered school boundaries?
What is the average frequency of school transfer, considering all secondary
schools in Amsterdam, measured as the school retention rate in Year 3?
Is switching associated with upward and downward mobility between secondary
tracks, comparing the three zips with the rest of Amsterdam?
Previous research is reviewed in Section 3.2. Data and methods are described in
Section 3.3; results are presented in Section 3.4. Conclusions and suggestions for further
research are presented in Section 3.5. The results indicate that living in the three
investigated zip‐code areas and being of Moroccan descent are factors associated with
an increase in the frequency of switching. Switching is associated with a reduction of
![Page 71: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/71.jpg)
3 School switching, a concern for migrant pupils
45
upward mobility, and an increase in downward mobility between secondary tracks,
especially for Moroccan pupils, and more so in the three zip areas.
3.2 Previous Research
The achievement gap and switching schools
The complexity of reducing the achievement gap is illustrated by the difficulties
researchers encounter in their many attempts to prove the causal effect of widely
implemented interventions aimed at improving the educational outcomes of
underrepresented groups. For example, Dobbie and Fryer (2009), in their investigation
into the effects of the Harlem Children's Zone initiative13 on results in Maths and English
Learning Arts, could not prove the causal and sustainable effect of investments in school
quality alone. Interestingly, they could prove, however, that improving the quality of
schools combined with investments in the community– thus creating more stable and
family‐supported school careers– had a positive effect on outcomes.
Enhancing the stability of schools careers by investigating "school careers as they
point to the ways in which young adults navigate the system" (Baysu and De Valk, 2012),
seems to be a promising approach, as the research which is reviewed here below may
illustrate: all types of intermediate school departure (obviously with the exception of
obtaining a diploma), seem to have a negative influence on the width of the gap, and to
reduce the odds of achieving a successful secondary school completion.
Hanushek, Kain and Rivkin (2001) found an overall negative effect of school
transfer for virtually all reasons for school changing: positive motives for switching, such
as moving to a more affluent neighbourhood, or upward mobility to a higher track or a
better school, and negative reasons, such as downward educational mobility or being
expelled from school, all turned out to be harmful to student's prospects. Their research
underlines, however, that the different mobility effects need to be disentangled in order
to identify the separate effects of specific types of school changing.
Swanson and Schneider (1999) identified three categories of pupil mobility: 1)
movers: moving to another residential area, but staying in the same school; 2) changers:
going to another school, but continuing to live in the same area; and 3) leavers: changing
both the area of residence and school. Their study shows that all mobility increases the
13 www.hcz.org
![Page 72: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/72.jpg)
46
odds of dropout, yet has different effects in the long or short run. The prospects of some
students may improve in the long run, possibly because a proportion of this group
moves on to a better quality school or to a higher level of education; in the short run,
however, the interruption of the school career burdens all students, and the risks are
higher for low SES students. Furthermore, the negative impact of student mobility
increases with grade and age; after the 10th grade (15‐16 years), a substantial negative
effect was found on student's cognitive growth. Better results in the long run for a
certain proportion of students in the U.S. may, furthermore, partially be the result of
"Tiebout effects" (1956): the tendency to move residence until the best neighbourhood
is found, given the family income; however, schools in the U.S. depend partly on locally
different tax revenues, implying that more affluent neighbourhoods may have higher
quality schools because of better funding. In contrast, in the Netherlands all schools are
funded equally by the government.
Rumberger and Larson (1998) were able to prove that transfer and dropout form
a continuum in educational disengagement; similar factors predict both, and dropout
increases with the number of school changes. The authors identify a personalized school
climate (where teachers know students well), an academic inspiring curriculum, and an
engaging social climate, as factors that enhance a beneficial connection between pupil
and school. Research by Murnane (2009) adds yet another interesting argument to this
debate on the negative effects of school mobility: substantial levels of pupil mobility at
school level also burdens those pupils who do not switch themselves, because teachers
are required to "devote scarce instruction time to assessing the skills and knowledge of
the new entrants, and socializing them to classroom norms of behaviour." Especially
children from poorer families are burdened by more often attending high‐turnover, low‐
quality schools (Hanushek et al., 2001), and tend to move recurrently within the same
urban school cluster (Kerbow et al. 2003); such children are referred to as "city
migrants" by Nakagawa, Stafford, Fisher and Matthews (2002).
Considering the harmful effects of school transfer, the question seems relevant
whether the "holding capacity" of a school should be considered as a critical measure for
school effectiveness, instead of only assessing average test scores (Rumberger and
Thomas, 2000). This view is further supported by Payne (2010), who identified six factors
that might reduce the persistence of underperformance in urban schools: instruction
![Page 73: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/73.jpg)
3 School switching, a concern for migrant pupils
47
time, intellectually ambitious instruction, a strong and focused community of teachers, a
challenging academic climate in combination with strong social support for pupils, the
diagnostic ability of teachers and a whole school approach. Notably, continuity of social
support for pupils, and guidance by means of formative assessments may benefit from
stability of school careers and long‐term relations with teachers.
At the individual level, not surprisingly, high absenteeism, misbehaviour, and low
social engagement with school and classmates predict school switching and dropout.
The widespread practice, described by Rumberger and Larson (1998), of administratively
transferring troublesome students to other schools, may worsen the prospects that
these underperforming students. In the first place, school changing increases the risk of
these students will dropout, while at the same time a cluster of schools may nonetheless
end up with the same number of troublesome pupils; they were just moved around
within the cluster from one school to the next.
Kerbow et al. (2003) summarize the factors that actually cause the negative
consequences of non‐promotional changing of schools as follows: 1) frequent movers
miss continued exposure to key concepts of learning and higher‐order skills; 2) moving
reduces overall the pace of learning; 3) ability grouping may be biased because teachers
have less time for assessments; 4) the chance to be placed in high ability groups may be
reduced and 5) cooperation with parents may be jeopardized. Migrant parents who
were interviewed in the context of the OECD review on migrant education in the
Netherlands (Shewbridge, Kim, Wurzburg and Hostens, 2010) expressed the opinion that
the complex tracked system for secondary education in the Netherlands puts an even
greater emphasis on stable contacts between parents and school and sustained
cooperation.
Stability of school careers and cooperation with parents
School effectiveness– especially for children from low SES parents– is positively
correlated with tightly knitted bonds between school, parents and community:
"connectedness matters" could be the dictum of Maroulis and Gomez (2008).
Furthermore, Margolis (2010) writes: "Teachers need to be anthropological
ethnographers of students (…). They work to understand students through learning
centred conversations that are sometimes focused directly on the immediate content
![Page 74: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/74.jpg)
48
being explored, but other times may be centred around personal, social, or cultural
issues that impact individual and group learning over time".
The point of view that teachers and school need to stay connected to the places
where pupils live is further supported by Miller Marsh and Turner‐ Vorbeck (2010), who
argue that the dissonance between the culture at home and that at school– a substantial
barrier for low SES children– can best be mitigated by strong neighbourhood schools and
continuous school careers. Finally, switching schools, notably for migrant pupils, may
threaten the sense of security that pupils have experienced by being among classmates
at their original school, and even increase exposure to hostility at another school (Leigh
McGarrigle, 2010).
Focus on school switching
In the Netherlands, the reduction of early dropout is one of the main themes on
the political agenda, on both the government and the municipal level. Obviously,
dropout may have severe consequences for young people. Moreover, over the lifetime
of the individual, the costs for society may be substantial as a result of reduced tax
revenues, and the costs of social security due to unemployment– that is often related to
dropout. On the other hand, the returns on investment of continued education, as Levin
extensively reports throughout his work (e.g. Levin, 2011, Belfield and Levin, 2007),
exceed all the costs of supportive programmes to guide disadvantaged young people
towards a diploma. Although over a lifetime the returns on investment in education are
substantial, in the short run, avoiding dropout and often repeatedly guiding youths back
to school requires expensive interventions due to the time‐consuming nature of the
work. Non‐promotional school changing appears to be driven by comparable factors
such as early school leaving, and similarly predicts higher odds of later dropout.
However, the reduction of switching could be far less expensive than the reduction of
dropout: enhancing the "holding capacity" of schools means, in practice, different ways
of organizing already existing student guidance and counselling practices, and to
agreements between schools in order to avoid creating groups of "city migrants"
(Nakagawa et al., 2002), who frequently migrate between schools and may eventually
end up as dropouts after all.
![Page 75: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/75.jpg)
3 School switching, a concern for migrant pupils
49
In summary, the literature predicts disruptive effects of school switching,
through the loss of familiar classmates, teachers, and curriculum. Discontinuity of school
careers challenges the intellectual growth of pupils, as measured by formative
assessments. Pupils enrolled in high turnover schools, who do not switch themselves,
lose time to learn because teachers have to spend time on taking on board every
newcomer. Finally, switching reduces chances to upward mobility and school
completion. These findings have inspired the current study.
3.3 Data and Methods14
3.3.1 Single schools and clusters
Inspired by Kerbow et al. (2003), we measured school switching at two levels,
and used the following definitions: 1) School switching: switching school, irrespective of
switching within a cluster or across cluster borders; and 2) cluster switching: switching to
another school in another cluster. The level of a single school and a cooperating school
cluster could be distinguished by means of the BRIN15 code, the school identification
number. Every single school has 4‐digit BRIN identification code; when schools
cooperate in a cluster two digits are added to the individual school code. Hence
switching between single schools could be compared to switching across borders of
school clusters. In practice, there may be occasional exceptions to the rule of four plus
two additional digits in the case of clusters, typically caused by the different histories of
mergers between schools. Therefore the BRIN codes of all schools and school clusters in
Amsterdam have been visually checked, in order to be able to trust the frequencies of
switching we found. While the literature describes different types of more or less formal
school clustering, as a rule, in the Netherlands, clustered schools share a common
board, and have a Principal per individual school.
3.3.2 Migrant students compared to native Dutch peers
We measured switching at two different points in time during the school career:
1) switching counted after exiting secondary education (with a diploma or early
dropout); and 2) switching measured at the entrance of Year 3 in secondary school.
14 I am profoundly grateful to Bregje Zwaan, Adam Booij and Ilja Cornelisz, who supported me with their technical
genius in carrying out Stata analyses. 15 BRIN: Dutch acronym, Basis Registratie Instellingen.
![Page 76: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/76.jpg)
50
Furthermore, we compared switching in the three zip areas with the rest of Amsterdam,
separate for different ethnicities and with a focus on pupils of Moroccan descent.
Finally, we measured how switching is associated with upward and downward mobility
between tracks in secondary school. The characteristics of the data sets will be
described below, as well as the choices that have been made in carrying out
measurements; Appendix 3‐1 presents a comparative overview of the data sets.
3.3.3 Three data sets
Data for the three zip‐codes (DMO data)
This data set covers the following information: on the reference date of 31 July 2009,
3168 16‐22 year‐olds live in zip area 1061, 1062 and 1063 in the district of Amsterdam‐
West. We choose this age range, instead of a cohort, for several reasons. We are
interested in the total frequency of switching when pupils exit secondary education.
Depending on the track in secondary education16, pupils exit school typically at age 16,
17 or 18. We also wanted to include, however, pupils who repeat classes or even leave
school and come back again. Therefore we decided to include youths up to 22 years of
age. Furthermore, a single year cohort would imply rather small numbers of individuals
when we consider only these three zip areas. Finally, we were interested in a "picture"
of the youths living in these three zip areas, at a single reference date. We acknowledge,
however, that cohort data offer valuable other information. Therefore, we repeated our
analyses with a second data set, covering cohort data, as described below.
The DMO data set covers both school‐going youths and those who have already left
secondary education, either with or without a diploma. In measurements of the
frequency of school transfers, all individuals who were still enrolled in secondary school
16 Official translations of the Dutch educational system in the ISCED classification (International Standard Classification of Education by UNESCO, update1997):
VMBO: pre‐vocational secondary education, 4 years, ISCED 2, qualifying for senior secondary vocational education;
HAVO: senior general education, 5 years, grade 1‐3 ISCED 2, grade 4‐5 ISCED 3, qualifying for higher education;
VWO: pre‐university education, 6 years, grade 1‐3 ISCED 2, grade 4‐6 ISCED 3, qualifying for higher education;
MBO: senior secondary vocational education, level 1 ISCED 2, level 2‐4 ISCED 3, level 4 qualifying for higher education;
HBO: universities for applied sciences, ISCED 5B;
WO: research universities, ISCED 5A.
The completion of ISCED level 3 is the internationally agreed initial (or basic) qualification for the labour market. School leavers without an ISCED 3 qualification are regarded as early school leavers or dropouts.
![Page 77: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/77.jpg)
3 School switching, a concern for migrant pupils
51
at the reference date were dropped, since this might have biased our measures: these
pupils were still enrolled in the 4‐, 5‐ and 6‐year programmes of the different secondary
tracks, and some of them may still switch before they exit secondary school.
The data set covers the age, gender and ethnicity of every individual; all individuals were
registered by an anonymous code. The name, address and BRIN code (identification
code) of schools were also registered. Every enrolment and school exit (with or without
a diploma) was registered at the exact date, at the individual level. When individuals
drop out of school, the Social Development Service registers them in a special "priority
file", to guide pupils back to school as quickly as possible. Thus we could follow every
pupil into re‐enrolment or final dropout. The total number of 3168 individuals in the
data set was taken into account for measuring the ethnic composition of the zip‐code
area, as well as the proportion enrolled in the highest academic tracks of secondary
education, as a proxy for the current achievement gap between native Dutch pupils and
pupils with a migrant background.
Several additional choices were made to prepare the data for analysis: tracks in
secondary education have been clustered in two main trajectories, referred to as the
"vocational track" and the "academic track": 1) VMBO, the vocational track (4 years, pre‐
vocational education, four sub‐tracks and some 40 different vocational programmes that
correspond with occupations on the job market), on the one hand; and 2) HAVO (5
years, senior general education) plus VWO (6 years, pre‐university education) combined
as academic tracks, on the other. Individuals enrolled in senior vocational institutions
were dropped; these institutions do not belong to secondary education (the scope of
this study) and are an advanced vocational level after secondary school. Pupils typically
leave elementary education at age 12; however, a limited number of individuals
between 16‐22 years of age were still registered as being enrolled in elementary
education. These students probably arrived in the Netherlands only recently, or may
experience severe learning disabilities; these individuals were dropped from the
analysis. Individuals with a missing BRIN identification code were also dropped. When
individuals re‐entered exactly the same school, this was not counted as switching. This
reduced the number of individuals in all counts of school switching from 3168 to 2653. A
further 564 individuals were dropped because they were still enrolled in secondary
schools. Thus 2089 individuals remained for analysis.
![Page 78: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/78.jpg)
52
This tailor‐made data set was kindly made available by the Social Development Service
of the municipality of Amsterdam (Dutch acronym: DMO).
Amsterdam school retention rates (DMO data, school year 2009/2010)
Additionally, as a reference for the frequency of school switching among pupils living in
the three zips, the frequency of switching measured at school level (i.e. the school's
retention rate) for all schools in Amsterdam was also measured. To this end the Social
Development Service granted the use of data on the retention rate of secondary schools
in Amsterdam (51 schools). The school's retention rate, for the purpose of this paper,
has been defined as the percentage of pupils who are still enrolled in Year 3, in the same
school where they enrolled initially in Year 1.
Pupils' position in secondary school Year 3 (BRON data), comparing the three zips with the rest of Amsterdam.
Having started in 2001, DUO, the government agency that collects all educational data
linked to the individual social security number, is in the process of constructing a new
educational database called BRON. This database will eventually cover the complete
educational history of all individuals. From 2008 on, all pupils in elementary education
are recorded in the BRON files, starting with pupils in the final grade of elementary
school in that year; at the time of the current research, this cohort had reached the third
year of secondary school. The BRON data make it possible to compare switching in the
district Amsterdam‐West with the frequency of switching in the rest of Amsterdam.
Importantly, with the BRON data, a first attempt could be made to measure how
switching is associated with upward and downward mobility between different
secondary tracks, a significant performance indicator. The BRON data set used for
measurements in this study contains 4815 pupils who were enrolled in the last grade of
elementary education in 2008 (a summary of data characteristics is given in Appendix
3‐1). Among these, 310 pupils lived in Amsterdam zip areas 1061, 1062 and 1063. For a
part of the total cohort (2593 individuals, 414 of Moroccan descent) the data set lists
every school transfer, plus every transfer to another secondary track (upward or
downward) at the same school or another school. Considering the three zips, data on
track mobility were available for 190 individuals (98 of Moroccan descent). The data
allowed for measuring the association between school switching and track switching.
![Page 79: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/79.jpg)
3 School switching, a concern for migrant pupils
53
Importantly, whereas the number of schools attended by one individual were counted
to answer Research question 1, in the case of upward and downward mobility through
secondary tracks and its association with switching, we used a dummy‐coded variable
(switching =1, non‐switching = 0).
Additionally, we used the BRON data to estimate the average retention rate of schools
attended by pupils living in the three zips.
3.4 Results
3.4.1 The frequency of switching in the three zip‐code areas
To set the scene, Table 3‐1 provides information on the ethnic composition of
zip‐code 1061, 1062 and 1063 in the district Amsterdam West, based on all 16‐22 year‐
olds in the data file, 3168 individuals. As can be seen from Table 3‐1, youths of
Moroccan descent form the largest group with a share of 40%. In total 83% of all 16‐22
year‐olds in the area have an immigrant background.
Table 3‐1: Composition by ethnicity of all 16‐22 year‐olds living in zip‐code areas 1061, 1062 and 1063 in the district Amsterdam‐West, at the reference date of 31 July 2009
Ethnicity Amount %
Native Dutch 543 17.1
Moroccan 1267 40.0
Turkish 647 20.4
Surinamese 229 7.2
Antillean/Aruban 28 0.9
Other 454 14.3
Total 3168 100
Source: DMO
As a further illustration, the total number of 3168 individuals was also taken into
account to measure enrolments in the academic tracks of secondary education as a
proportion of each ethnic group; these indicative measurements have been included as
an approximation of the width of the achievement gap in the three zip areas (Table 3‐2);
the available data allow for the measurement of the percentage of individuals who were
registered at least once as being enrolled in the academic secondary track. The share of
them who obtained a diploma at this level is not considered (it was not in the data file
for all individuals).
![Page 80: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/80.jpg)
54
Importantly, the percentage of native Dutch individuals in academic tracks may
be overestimated here: native Dutch youngsters, who grew up in other cities, tend to
come to live in the area when they start their university studies, attracted by cheaper
housing opportunities. From the data we cannot derive whether these individuals were
born in the area or elsewhere; but since the name of all schools is registered, we can
identify that their secondary school was outside the city border. Unfortunately, this did
not allow us to drop these individuals, since we could not be sure whether they lived
outside the city border or travelled a further distance to school. This may explain why
the share of them that has been registered at the academic secondary level, is actually
well above the national average (45.1% in 2009). Enrolments of 15.9% among Moroccan
pupils in the investigated zip area, and 15.2% of Turkish pupils are, however, well below
the national average (nationally 20.5% and 20.6%, respectively, in 2009). Note that we
do not encounter this problem in measurement based on the BRON data (from Table 3‐6
on); individuals in this data set have not yet reached the age of access to university.
Table 3‐2: The share of 16‐22 year‐olds by ethnicity, enrolled in academic secondary tracks. Individuals were registered at least once in the data file as being enrolled at this level
Ethnicity Total In academic tracks %
Native Dutch 543 317 58.4
Moroccan 1267 202 15.9
Turkish 647 98 15.2
Surinamese 229 35 15.3
Antillean/Aruban 28 9 32.1
Other 454 121 26.7
Total 3168 782 24.7
Source: DMO
Table 3‐3 presents, separately for the six distinct ethnic groups, the number of
secondary schools that were attended by pupils counted after they had exited secondary
education. The default situation would be that a pupil attends one single secondary
school, except when parents move to another area. As can be seen from Table 3‐3, there
are differences, notably in the percentages enrolled in only one school, between
different groups: 67.0% of pupils with a Moroccan background attended only one
secondary school; this is also the case for 72.5% of Turkish students. The total average
![Page 81: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/81.jpg)
3 School switching, a concern for migrant pupils
55
over all groups of enrolments in only one school was 69.4%. Considering native Dutch
students, 77.4% did not switch.
Table 3‐3: The number of secondary schools attended by 16‐22 year‐olds, by ethnicity
No. of schools
Ethnicity
1 % 2 % 3 % 4 % Total %
Native Dutch 151 77.4 36 18.5 7 3.6 1 0.5 195 100
Moroccan 647 67.0 262 27.1 53 5.5 4 0.4 966 100
Turkish 346 72.5 102 21.4 27 5.7 2 0.4 477 100
Surinamese 111 64.5 51 29.7 8 4.7 2 1.2 172 100
Antillean/Aruban 13 76.5 3 17.7 0 0.0 1 5.9 17 100
Other 181 69.1 62 23.7 15 5.7 4 1.5 262 100
Total 1449 69.4 516 24.7 110 5.3 14 0.7 2089 100
Source: DMO.
Table 3‐4: The number of secondary schools attended by 16‐22 year‐olds, by ethnicity, measuring school switching that includes cluster switching
No. of schools Ethnicity
1 % 2 % 3 % 4 % Total %
Native Dutch 165 84.6 26 13.3 4 2.1 0 0.0 195 100
Moroccan 765 79.2 175 18.1 25 2.6 1 0.1 966 100
Turkish 408 85.5 62 13.0 7 1.5 0 0.0 477 100
Surinamese 133 77.3 35 20.4 3 1.7 1 0.6 172 100
Antillean/Aruban 35 88.2 2 11.8 0 0.0 0 0.0 17 100
Other 207 79.0 43 16.4 12 4.6 0 0.0 262 100
Total 1693 81.0 343 16.4 51 2.4 2 0.1 2089 100
Source: DMO.
The measurement of the number of schools attended was repeated at the higher
aggregate level of a school cluster, in order to compare the proportion of switching
between all separate schools, to switching across borders of clustered groups of
cooperating schools (Table 3‐4). Measuring at the higher aggregate level of clustered
schools reduces the number of switches, indicating that pupils also migrate within a
cluster of schools. Interestingly, considering school switching that includes cluster
switching, the percentage of pupils of Turkish descent who do not switch (85.5%) is close
to the percentage among native Dutch pupils (84.6%). Youngsters of Moroccan descent,
however, switch more often, both in counts at the separate school level and at the
cluster level. The substantial difference of 12.2 percentage points considering Moroccan
![Page 82: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/82.jpg)
56
pupils who attend only one school, between counts at the level of single schools and
clusters, indicates they move rather frequently within a cooperating school cluster.
Table 3‐5 presents 16‐22 year‐old Moroccan students on academic track levels,
who have exited secondary school (with or without a diploma), or have moved on to
another school. Three types of outflow status are specified: 1) departure from school
with a diploma; 2) dropout without a diploma; and 3) switch to another school.
Considering pupils who exit their first school of enrolment, a majority of 87.4% leave
with a diploma; 10.3% switch and continue in another school. However, when pupils
exit, for example, their third school, only 39.1% do so with a diploma. When exiting the
third school, no less than 43.5% drop out, and do not re‐enrol again in a next school.
Table 3‐5: Categories of school‐exit among pupils of Moroccan descent, who were enrolled in academic secondary tracks (reference date 31 July 2009)
No. of schools Cause
1 % 2 % 3 % 4 % 5 % 9 % Total
Drop out 2 2.3 18 24.7 10 43.5 6 50.0 4 66.7 1 100 41
Other school 9 10.3 4 5.5 4 17.4 3 25.0 1 16.7 0 0.0 21
Diploma 76 87.4 51 69.9 9 39.1 3 25.0 1 16.7 0 0.0 140
Total 87 100 73 100 23 100 12 100 6 100 1 100 202
Source: DMO.
3.4.2 The retention rate of schools
To illustrate the occurrence of switching at school level, Figure 3‐1 depicts the
retention rate for all schools in Amsterdam (51 schools). The retention rate is defined
here as the percentage of pupils who enrolled in a secondary school in Year 1, who are
still enrolled in the same school in Year 3. Separate schools are observed here, clustering
is not considered. While the average retention rate for schools in Amsterdam is 85%,
schools differ substantially within a range of 55% to 98% retention in Year 3. Since the
schools in the municipal data file have anonymous codes, unfortunately, we cannot
derive directly from these data whether pupils living in the three zips attend a school
with a higher or lower retention rate. We could, however, estimate the average
retention rate of schools attended by pupils living in the three zip areas, by means of the
BRON data set. Therefore we considered all Amsterdam schools where pupils living in
the three zip area have enrolled (in the Netherlands 88.7% of pupils in secondary
![Page 83: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/83.jpg)
3 School switching, a concern for migrant pupils
57
education are enrolled in another school than the one closest to their home, as will be
described in Chapter 5 of this thesis), and derived the retention rate of these schools by
using the total number of individuals enrolled in a school, according to our BRON data
set. We found that the average retention rate of the schools where pupils living in the
three zip areas were enrolled is 71.2%, a very low retention rate considering Figure 3‐1.
When we distinguish between the retention rates of schools attended by switchers (≥ 1
switch) and non‐switchers living in the zip areas who were still enrolled in Year 3:
switchers attend a school with a mean retention rate of 50.1%, and non‐switchers
attend a school with a retention rate of 79.7%.
Figure 3‐1: The retention rate in ranking order of all secondary schools (considering separate schools, not clusters) in Amsterdam, defined as the percentage of pupils who entered in Year 1, and are still enrolled in Year 3
Source: DMO; school year 2009/2010.
3.4.3 Switching and upward and downward mobility between track levels
Observations in the BRON data have not been collected for long enough to allow
counting all switches during secondary school. We will use these data to distinguish
switchers (at least one school switch) and non‐switchers; for the cohort in the highest
grade of elementary school in 2008, we can only measure switching behaviour up to the
Year 3 in secondary education
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
0 10 20 30 40 50 60
Series1
![Page 84: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/84.jpg)
58
Table 3‐6 presents school switching in Amsterdam during the first three years of
secondary school, as a dichotomous variable. As a comparison, Table 3‐7 depicts the
same measurement for zip areas 1061, 1062 and 1063. In the three zips areas, switching
is less frequent among native Dutch pupils and more frequent among immigrant pupils
(but, except for Moroccan pupils, the number of observations in the three zip areas is
very small). All over, the frequency of switching is higher in the three zip areas in the
district Amsterdam‐West, than overall in Amsterdam.
Table 3‐6: School switching among pupils in Amsterdam enrolled in the Year 3 of secondary education in 2011
Switch Ethnicity
None % ≥1 % Total %
Dutch 1422 81.7 319 18.3 1741 100
Surinamese/Antillean 581 76.7 177 23.4 758 100
Turkish 337 73.3 123 26.7 460 100
Moroccan 577 70.4 243 29.6 820 100
Other 774 74.7 262 25.3 1036 100
Total 3691 76.7 1124 23.3 4815 100
Source: BRON data.
Table 3‐7: School switching among pupils living in zip‐codes 1061, 1062 and 1063, enrolled in the Year 3 of secondary education in 2011
Switch Ethnicity
None % ≥1 % Total %
Dutch 20 90.9 2 9.1 22 100
Surinamese/Antillean 13 72.2 5 27.8 18 100
Turkish 58 72.5 22 27.5 80 100
Moroccan 106 68.4 49 31.6 155 100
Other 24 68.6 11 31.4 35 100
Total 221 71.3 89 28.7 310 100
Source: BRON data.
In Tables 3‐8 and 3‐9, the above counts have been repeated for switching across
school cluster borders.
Differences between ethnicities are much smaller now, except between native
Dutch and immigrant students. The frequency of cluster switching is roughly 2/3 of
school switching, indicating that 1/3 of school switching occurs within clusters, and 2/3
crosses cluster boundaries: the majority of school switching occurs across cluster
borders. But, as is the case for school switching, cluster switching among immigrant
![Page 85: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/85.jpg)
3 School switching, a concern for migrant pupils
59
students is more frequent in the three zips than in the rest of Amsterdam, whereas
among native Dutch students it is lower in the three zips (but numbers are very small
here).
Table 3‐8: Cluster switching among pupils in Amsterdam enrolled in Year 3 of secondary education in 2011
Switch Ethnicity
None % ≥1 % Total %
Dutch 1520 87.3 221 12.7 1741 100
Surinamese/Antillean 654 86.3 104 13.7 758 100
Turkish 386 83.9 74 16.1 460 100
Moroccan 674 82.2 146 17.8 820 100
Other 880 84.9 156 15.1 1036 100
Total 4114 85.4 701 14.6 4815 100
Source: BRON data.
Table 3‐9: Cluster switching among pupils living in zip‐codes 1061, 1062 and 1063 enrolled in Yea 3r of secondary education in 2011
Switch Ethnicity
None % ≥1 % Total %
Dutch 21 95.5 1 4.6 22 100
Surinamese/Antillean 14 77.8 4 22.2 18 100
Turkish 63 78.8 17 21.3 80 100
Moroccan 123 79.4 32 20.7 155 100
Other 28 80.0 7 20.0 35 100
Total 249 80.3 61 19.7 310 100
Source: BRON data.
In the tables below, we take a closer look at the nature of pupil mobility: pupils
can move up to a higher secondary track, move down to a lower track, or stay at the
same level; we considered all steps of upward and downward mobility, from vocational
(VMBO) to the senior general track (HAVO), and from the latter to the pre‐university
level (VWO). Moving between tracks can occur both for pupils staying in the same
school, and for those who enrol in another school.
We use another subset for measurements now, and consider only individuals
who changed track: 2593 individuals in Amsterdam, where 414 are of Moroccan
descent; 190 individuals in the three zips, where 98 are of Moroccan descent.
Interestingly, the vast majority of pupils in Amsterdam stay in the same track, also after
switching (Table 3‐10). Upward mobility occurs more often when pupils stay in the same
![Page 86: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/86.jpg)
60
school (22.7%) than among pupils who transfer to another school (9%). The reverse is
the case for downward mobility: 11% downward mobility when changing track in the
same school, against 17.3% among those who switch to another school. Table 3‐11
presents similar measurements, considering the three zip areas. Correspondingly, the
vast majority of pupils continue at the same level, including after switching. However,
both after switching and when staying in the same school, upward mobility is lower in
comparison with the rest of Amsterdam. Furthermore, switching is more strongly
associated with downward mobility in the three zips.
Table 3‐10: Comparison of track level at entry and in Year 3 of secondary school, separate for school switchers and non‐switchers, among pupils in Amsterdam
Switch Level
None % ≥1 % Total %
Level(s) down 223 11.0 98 17.3 321 12.4
Same level 1344 66.3 417 73.7 1761 67.9
Level (s) up 460 22.7 51 9.0 511 19.7
Total 2027 100 566 100 2593 100
Source: BRON data.
Table 3‐11: Comparison of the track level at entry‐ and in Year 3 of secondary school, separate for school switchers and non‐switchers, among pupils in zip‐code areas 1061, 1062 and 1063
Switch Level
None % ≥1 % Total %
Level(s) down 14 9.8 9 19.2 23 12.1
Same level 108 75.5 36 76.6 144 75.8
Level(s) up 21 14.7 2 4.3 23 12.1
Total 143 100 47 100 190 100
Source: BRON data.
Finally, Tables 3‐12 and 3‐13 compare the mobility between tracks among pupils
specifically with a Moroccan background, in Amsterdam (Table 3‐12) and in the three
zips (Table 3‐13). 19.5% of pupils of Moroccan descent move up to a higher track in
Amsterdam when they stay in the same school; this is only the case for 7.6% of the
switchers, who also have a higher rate of downward mobility (13.2%). In the three zip
areas in the district Amsterdam‐West, Moroccan youths staying in the same school
realize an upward mobility of 17.1% when they stay in the same school (caution because
of small numbers).
![Page 87: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/87.jpg)
3 School switching, a concern for migrant pupils
61
Table 3‐12: Comparison of track level at entry‐ and in Year 3 of secondary school, separately for school switchers and non‐switchers, among pupils of Moroccan descent in Amsterdam.
Switch Level
None % ≥1 % Total %
Level(s) down 22 7.1 14 13.2 36 8.7
Same level 226 73.4 84 79.3 310 74.9
Level (s) up 60 19.5 8 7.6 68 16.4
Total 308 100 106 100 414 100
Source: BRON data.
Table 3‐13: Comparison of track level at entry‐ and in Year 3 of secondary school, separately for school switchers and non‐switchers, among pupils of Moroccan descent in zip‐code areas 1061, 1062 and 1063
Switch Level
None % ≥1 % Total %
Level(s) down 5 7.2 3 10.7 8 8.2
Same level 53 75.7 23 82.1 76 77.6
Level (s) up 12 17.1 2 7.1 14 14.3
Total 70 100 28 100 98 100
Source: BRON data.
In summary, more pupils switch in the three zip‐code areas 1061, 1062 and 1063,
than in the rest of Amsterdam. Furthermore, comparing all ethnic groups, the frequency
of switching is substantially higher among migrant students. Notably for Moroccan
students living in the three zip areas, the incidence of switching is comparatively high at
31.6% (Table 3‐7). Differentiating switching further into upward and downward mobility
through secondary tracks demonstrates that upward mobility is higher for all groups,
when they stay in the same school. 17.1% of pupils of Moroccan descent in the three zip
areas, move up to a higher track when staying in the same school (Table 3‐13).
3.5 Conclusions and policy implications
The main data set analysed here is the DMO data file, kindly made available by
the Social Development Service of the municipality of Amsterdam. This data set covers
the detailed school history of youth living in the three studied zip areas. This local data
set has been an important source of information about key policy issues for the
municipality of Amsterdam, notably the reduction of early dropout, and is, furthermore,
far more detailed than national data sets. However, its architecture makes the BRON
data much more accessible for research. Ideally, the architecture of the national data set
![Page 88: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/88.jpg)
62
should be compatible with the municipal data set, in order to support local policy
makers in translating more general national educational analyses to local practice which
fit every school's specific context. While currently the two data sets we used have a
largely different architecture, we hope that our comparison of the data sets may support
further initiatives to make the data compatible, a promising endeavour considering the
fascinating opportunities the BRON data will offer in the near future, when we will be
able to follow the 2008 cohort from elementary school until tertiary education, and,
finally, on the labour market. By then, for example, we will be able to estimate the
causal effects of school switching on school success.
3.5.1 The three zip areas
Overall, a larger share of Dutch pupils in comparison with migrant pupils, do not
switch. The share of pupils with a Moroccan background who do not switch is the
lowest, at 67% (Table 3‐3). Whereas 22% of native Dutch pupils switch school, 33% of
Moroccan pupils switch. Among both groups, the share of school switchers who also
switch cluster is about 2/3: 21% of Moroccan pupils switch cluster, whereas only 15% of
their native Dutch peers switch cluster (Table 3‐3 and Table 3‐4). These findings suggest
that Moroccan migrant students may be extra‐burdened by the assumedly serious
disruptive effects of switching, as the literature predicts. Considering that the
achievement gap is most wide in the case of pupils of Moroccan descent, especially in
their case, agreements between schools to reduce switching, or, when switching is
unavoidable, to securely guide the transfer, may offer promising opportunities, within
the span of control of schools, at relatively low costs. Clustered schools have the
advantage that they can make agreements between their own schools to reduce
switching. Although the currently available data do not yet allow for demonstrating
causal evidence, the association between obtaining a diploma at academic secondary
levels and staying in the same school during the whole school career, convincingly points
in the same direction as the causal evidence presented in the reviewed literature in
Section 3.2 (e.g. Hanushek et al., 2009). Considering these Moroccan youths who
remained in one school, 87.4% obtained a diploma on the academic secondary level
(Table 3‐5). Those who exited a second school already showed a dropout percentage of
24.7%.
![Page 89: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/89.jpg)
3 School switching, a concern for migrant pupils
63
3.5.2 A hidden problem?
Remarkably, the vast majority of switchers continue in the same track in another
school. This raises the question why they did switch. This finding may refer to a specific
problem; pupil movements between schools may partly be the result of disciplinary
school policies, specifically expulsion policies. Both the municipality of Amsterdam and
the Inspectorate of Education are well aware of increased risks of later school dropout,
when a pupil is expelled from school. Therefore the school has a duty to report every
pupil that has been expelled. However, sending a pupil to another school under the
same board within a cluster may de jure not imply that the pupil has been expelled, and
the school may not report these transfers to the Inspectorate. In this way involuntary
departure may be disguised as voluntary departure‐ and behavioural problems that
induced change to another school may persist. To estimate the occurrence of not
reporting school departure within clusters would be a question for further research.
The substantial differences in retention rates of all secondary schools in
Amsterdam (Figure 3‐1), which range from roughly 55% to 95% retention (measured
between Years 1 and 3), illustrate that indeed the frequencies of switching do differ
substantially, which leaves ample room for improvement.
3.5.3 Youth in the three zips might benefit from agreements between schools to reduce switching
The analyses based on the national BRON data confirm that pupils with a migrant
background switch more often than native Dutch students, and that switching is higher
still among migrant pupils living in the three studied zip areas.
The BRON data made it possible, furthermore, to carry out a first exploration of
possible relations between switching and educational outcomes. Considering the
tracked system of secondary education in the Netherlands, switching especially appears
to reduce upward mobility to a higher secondary track: In Amsterdam 22.7% of all pupils
move up to a higher track; this is, however, only the case for 9% among switchers, while
in their case downward mobility is 17.3% (Table 3‐10). These differences between
switchers and non‐switchers are far more articulate in the three zip areas (Table 3‐11):
among switchers only 4.3% move up to a higher track. Furthermore, looking specifically
at students of Moroccan descent (total of Amsterdam) who did not switch until Year 3 of
secondary school, a remarkable 19.5% has improved their track level; even in the three
![Page 90: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/90.jpg)
64
zip areas, this is still 17.1%. Interestingly, Van de Werfhorst and Hofstede (2007) in their
study on relative risk aversion (in this case the avoidance of downward mobility) find
that it is rather concerns with mobility (i.e. the risk involved in moving to a higher track)
than with social class that are associated with ambitions to progress to higher levels of
schooling. However, switching substantially reduces the upward mobility of Moroccan
youths to 7.6% in Amsterdam and 7.1% in the three zip‐code areas (Tables 3‐12 and
3‐13). Upward mobility potentially may have a substantial positive effect on the closing
of the achievement gap, since migrant pupils more often may need some extra time to
move up to their level of abilities. Staying in the same school, the current research
suggests, improves the chances of upward mobility. The study by Van de Werfhorst and
Hofstede (ibid.) seems to support this finding, and, moreover, staying in the same school
when moving to a higher track, may reduce the experienced risk of mobility, notably for
migrant pupils.
3.5.4 Schools offering all tracks may be favourable for migrant students
As Crul and Schneider describe in their comparative study on school success of
pupils with a Turkish background in Germany and in the Netherlands (2009), the tracked
Dutch system offers opportunities to follow a longer route through successive higher
levels of secondary education, eventually allowing access to higher education.
Stimulating upward mobility within the same familiar school, which offers a broad
choice of tracks, may be in favour of low SES students, since this may minimize the
disruptive effect of school transfer for students. This assumption would counter,
however, the current trend in the Netherlands to organize vocational and academic
secondary tracks in separate schools, and also even diversify further between academic
tracks.
3.5.5 Reducing switching may decrease costs
Finally, active policies to reduce switching might also be worthwhile considering
in terms of cost‐effectiveness. As has been mentioned earlier in this chapter, policies to
reduce dropout tend to be very costly, because of the labour intensive, complex and
tailor made nature of such strategies. Dropout may be the final result of problems that
worsened over a long period of time. The literature referred to in Section 3.2 offers
convincing arguments for regarding school changing as a predictor for later dropout, and
![Page 91: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/91.jpg)
3 School switching, a concern for migrant pupils
65
the findings in the current paper suggest the same tendency: dropout increases with the
number of schools attended. Presumably, the costs of the reduction of switching may be
considerably lower than the repair of dropout, since such policies can be agreed upon
between schools and mainly require transparency in the reasons for student mobility;
clustered schools in particular have the opportunity to make agreements to reduce
switching between their own schools.
In Summary, when pupils of Moroccan descent stay in the same school while
enrolled in the academic track in the three zips, 87.4% obtains a diploma; already when
they have been enrolled in two schools (one switch), the percentage of diplomas drops
to 69.9%, and dropout is 24.7% (5.5% transfer to yet another school). These last
percentages are especially worrying, because here we consider Moroccan pupils who
qualified for the academic secondary level: this dropout directly widens the
achievement gap, although we cannot know from our data which share eventually takes
up their studies again after the age of 22. Furthermore, a first exploration of the effect
of switching, in terms of upward or downward mobility between secondary tracks
(measured in Year 3) resulted in substantial differences: upward mobility among
Moroccan pupils who do not switch, is 19.5% in Amsterdam and 17.1% in the three zips,
almost 1 in 5 pupils: these track improvements may add in important ways to the closing
of the achievement gap. However, among switchers, upward mobility decreases to 7.6%
and 7.1% (three zips) respectively. As a consequence, agreements between schools in
Amsterdam to reduce switching may be expected to have positive results on the closing
of the gap. We hope that insights into the occurrence and effects of switching described
in our study may motivate schools to reduce switching, and to be transparent about
expelling pupils as a disciplinary measure, although this may be a more complex issue.
Interesting opportunities for further research will arise, once the BRON data
cover the exit‐exam results in secondary education of the cohort that was enrolled in
the final year of elementary school in 2008. Hopefully this paper may add to convincing
the Ministry of Education, Culture and Science of the importance of also making these
high‐quality data available for scientific research in the future.
![Page 92: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/92.jpg)
66
3.6 Appendix
Appendix 3‐1: Comparison of the two main data sets.
DMO data set Amsterdam municipal data
BRON data set National educational data
# Individuals 3168 4815
Characteristics All 16‐22 year‐olds living in zip‐code areas 1061, 1062 and 1063 in district Amsterdam‐West. Exact date of all school transfers, exact BRIN code of all schools, plus the BRIN code of clusters of schools. Ethnicity of all pupils, plus track level in secondary school and exit status when leaving secondary education.
The cohort, living in Amsterdam, which was enrolled in the last grade in elementary school in 2008. Postal code of the residence of all pupils; BRIN code of all schools on the school and cluster level. Ethnicity, home address and track level of all pupils.
Reference date July 2009 The cohort enrolled in the last grade of elementary school in 2008, living in Amsterdam. The data were updated to 2011, when pupils typically were enrolled in Year 3 of secondary school.
Selection of sample
Kept: all pupils who had left secondary school at the reference date, with or without a diploma, whose complete history of school attendance plus school BRIN code was registered: 2089 individuals. The total number of individuals (3168) was used to indicate neighbourhood composition and the educational achievement gap between native Dutch and second generation immigrant pupils.
Kept: all pupils whose complete school transfer history was listed in the data set; 2593 individuals in Amsterdam, 190 of them living in the three zip areas 1061, 1062, 1063.
Time of measurement
After exit from secondary school In Year 3 of secondary school
Switching variable
Count of the number of schools in which a pupil has been enrolled.
Dummy‐coded variable (switching =1, non‐switching=0). No distinction between once or repeated switching.
The third data set (DMO) covers the percentage, at school level, of pupils who
are still enrolled in the same school in Year 3, as the school they entered in Year 1;
reference date: school year 2009/2010.
![Page 93: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/93.jpg)
67
4 Patterns in secondary school selection in
the context of unlimited choice
I should also add further that this liberty (i.e. the liberty of
judgment) is absolutely essential to the advancement of the arts
and sciences; for they can be cultivated only by those with a free
and unfettered judgment.
Baruch de Spinoza, 1670
4.1 Introduction
Considering the stubborn problem of unequal access to higher levels of
education for different socio‐economic groups, offering broader options for school
choice is a policy measure that generally raises high expectations. Interestingly, in the
Netherlands, free school choice has been practically unlimited for almost a century, by
the constitutional law of 1917. This law also gives parents the right to found a new
school according to, for example, a specific religious or pedagogical concept, on the
condition that a sufficient number of pupils will attend the new school, and no
comparable school is available within reasonable distance. In practice, because of the
density of schools in the Netherlands, the constitutional right to found a new school is
nowadays rarely exercised. Free school choice is, moreover, guaranteed because all
This chapter is based on: van Welie, E.A.A.M., Hartog, J. and Cornelisz, I. (2013). Forthcoming.
![Page 94: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/94.jpg)
68
schools are funded equally by the state, both schools with a board based on private law
or those with a board based on public law. Additionally, in elementary and secondary
schools no tuition fee is required, except for a contribution for certain activities outside
the curriculum, like school festivities. This contribution is, however, not obligatory; when
parents cannot afford to pay, the child is still entitled to participate. Finally, schools are
independent of local taxes (since all funding comes from the Ministry of Education), and
all children in elementary education growing up in low SES families are additionally
funded through a voucher system, the money following the pupil. Secondary schools
receive extra funding when the student population crosses a threshold percentage of
pupils who live in low SES neighbourhoods (the threshold percentage is lower for
vocational tracks and higher for academic tracks). This results in schools with relatively
more low SES pupils actually receiving more funding than schools with more children
from affluent parents. In summary, the long history of free school choice, equal funding
by the state, no tuition fees, and extra funding for low SES pupils, offers interesting
opportunities to study possible differences in school choice between distinct groups in
society.
In the current chapter we compare patterns of school choice among native Dutch
youth and youth with a migrant family history. In most cases, these migrant pupils'
grandfathers came to the Netherlands as unskilled workers in the 1969s and 1970s. The
average level of education of the children and grandchildren of these labour immigrants
has shown a steady increase over the past decades. However, considering the
importance of attainment levels that fit pupils' capacities– both for the individual and
for society at large– we still need a deeper understanding of the actual choices migrant
pupils make, in order to develop policies that may accelerate the process of finding
equal access to high levels of schooling. According to the OECD17 (2011), a knowledge‐
based economy may lose part of the population's potential to expand scientific and
cultural knowledge, and may receive lower tax revenues over the lifetime of individuals,
when substantial numbers of citizens remain under‐schooled.
In this chapter we focus on secondary schools, and diversify between secondary
vocational tracks (VMBO) on the one hand, and two secondary academic levels (HAVO
17 OECD: Organization for Economic Cooperation and Development.
![Page 95: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/95.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
69
and VWO), on the other18. In the Dutch educational system VMBO is the lowest level,
and VWO the highest level. The achievement gap in Dutch secondary education is
typically described as the lower enrolment levels in academic tracks among youth with a
migrant background, as compared to their native Dutch peers. Both academic tracks
offer access to higher education; in the text we refer to these two levels combined as
the academic level; in some tables we diversify between HAVO and VWO.
In Europe and the U.S. alike, the debate on raising educational opportunities for
underrepresented groups in the higher strata of education seems to concentrate on
extending school choice, assuming that this will enhance the desegregation of schools
along the lines of race, ethnicity, or social status; the underlying assumption for
desegregation policies in general, is that low SES children may benefit from peers
belonging to other social strata.
Interestingly, geographical distance measurements offer opportunities to
disentangle the effects of school choice, school composition, neighbourhood
composition, school density, and urbanicity19. Using the right to select a school freely,
which may result in choosing another school than the one closest to the pupils'
residence, unavoidably involves the extra effort of a longer distance to travel to school.
In the present chapter, we use measurements of the distance from home to the chosen
secondary school, relative to the nearest available school, controlling for indicators at
the individual, neighbourhood and school level, in order to be able to analyse whether
and how patterns of school choice differ between ethnic groups.
Analysing the effect of secondary school choice is complicated by the fact that
measuring school success– as the combined effect of individual, neighbourhood, school
quality, and educational system variables– requires data on prior achievement in
elementary education to start with. Interestingly, in 2001, the Dutch Ministry of
18 Tracks in the Dutch system for secondary education, plus the ISCED translation (International Standard Classification of Education by UNESCO, update1997):
VMBO: pre‐vocational secondary education, 4 years, ISCED 2, qualifying for senior secondary vocational education.
HAVO: senior general education, 5 years, grade 1‐3 ISCED 2, grade 4‐5 ISCED 3, qualifying for higher education (professional universities).
VWO: pre‐university education, 6 years, grade 1‐3 ISCED 2, grade 4‐6 ISCED 3, qualifying for higher education (research universities).
19 Urbanicity: a measure for the intensity of human activity in a given area, based on the number of addresses per km² (definition Statistics Netherlands).
![Page 96: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/96.jpg)
70
Education, Culture and Science started to create a longitudinal educational data set
based on individual social security numbers, covering all students and their complete
educational history. In 2008, all pupils in the final year of elementary education were
recorded for the first time in this new database (students in other strata in the
educational system had preceded them). In the coming years, when the total school
history of all individuals will be recorded, performance in elementary education can
actually be linked to the exit‐exam outcomes of secondary education at the individual
level20. At the time of our research, the cohort that was enrolled in the final year of
elementary education in 2008 has reached the Year 3 in secondary school. Therefore,
we could relate prior achievement in elementary education to achievements at the
entrance of the third secondary year.
Among policy makers and researchers alike, positive effects of equal
opportunities in education are expected from: 1) free school choice; 2) equal
(government) funding of schools; 3) independence of local taxes; and 4) extra funding
for low SES pupils. These four conditions have all been met system‐wide in Dutch
schools. Although the outcomes of the Dutch system of education rank among the top‐
ten in international comparisons (PISA, 2009), the pattern of unequal access to higher
levels of schooling among low SES youth per se, resembles patterns in countries without
the four above mentioned conditions.
While we use distance to the preferred non‐nearest school as a measure for the
selectivity of choice, we acknowledge that choosing the proximity allocation may also be
based on selective choice; however, in this case we do not know whether choice was
based on qualitative considerations, or whether parents and pupils simply chose the
nearest school because it was close to their home. Both high SES and low SES pupils,
notably youth with a migrant background, may have reasons to choose the school
closest to the residence‐ or prefer another school, albeit these reasons may be different
between both groups. Obviously, choosing the nearest school takes the least effort in
terms of travelling time and costs. Since there is no school bus system in the
20 We are indebted to Cees Vermeulen at DUO (the Dutch government organisation that collects all educational data), who provided us with an extended data set of great quality. We acknowledge his unique professionalism in designing and loading this new database. Moreover, we valued his advice in finding our way in the database. We are also deeply grateful for the support of Erik Smits and Rob Kerstens (Director‐ General of DUO).
![Page 97: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/97.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
71
Netherlands, avoiding the dangers of traffic may also be a consideration for parents.
There may be other reasons though: native Dutch pupils living in affluent areas may
prefer a school close to home, because this school presumably mirrors the high SES
composition of the residential area. Migrant students, on the other hand, may choose
the nearest school because this school may be more specialized in migrant education,
for example with an extra emphasis on language proficiency. Migrant students may also
have reasons to choose a school where they form the majority. In summary, preferring
the nearest school may, or may not be, a deliberate choice; however, on the basis of our
available data we cannot diversify between different motives behind the choice for the
nearest school. We assume, however, that making the extra effort to travel further to
school is the result of active choice.
We explore the following research questions:
1. Does distance to school, as a measure of selectivity of school choice, differ
between ethnic groups, and does the group of pupils who choose the nearest
school differ from those who opt for a school further away?
2. How is school choice associated with ethnic segregation?
3. Do pupils benefit from selective school choice?
The outline of the chapter is as follows: previous research is reviewed in Section
4.2; the data sources we used are described in Section 4.3; the results are presented in
Section 4.4; conclusions, discussion and possible policy implications for policies are
presented in Section 4.5. The results show that, of all pupils in secondary schools living
in the four largest cities, 88.7 % are not enrolled in the school nearest to their residence.
Pupils of Dutch origin, living in poorer neighbourhoods, have a stronger tendency to
choose a school at a further distance than migrant students in such neighbourhoods.
Differences between the group of pupils who choose the nearest school and those who
choose the non‐nearest school are marginal. Prior achievement in elementary
education, secondary school average exam scores, and school‐level upward mobility are
only marginally associated with distance travelled to school. The school's percentage of
students with a non‐Western immigrant background and the school population's
average SES, however, do significantly drive school choice. Choice patterns seem to
reveal ethnic segregation by choice among migrant pupils.
![Page 98: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/98.jpg)
72
4.2 Previous research
Our data allow for measurements of socio‐economic pupil, neighbourhood, and
school characteristics; for qualitative characteristics on the individual and school level;
and for distance measurements. In the literature reviewed in this section, we have made
an attempt to address this wide array of factors related to the patterns and geographic
distribution of school choice. Segregation, neighbourhood schools, differential effects
related to distinct pupil characteristics, postponed tracking, distance from home to
school and extended school choice are the key issues we explored in academic
literature: 1) We studied: 1) research that points out the benefits of mixed desegregated
schools, identifying peer effects and parental support by better‐educated parents as
main mechanisms behind higher expected outcomes; but 2) in contrast, research that
finds positive effects of segregated neighbourhood schools in the case of migrant pupils,
that indicates the importance of close cooperation between parents and school; 3) yet
other research findings that demonstrate differential effects for different groups. For
example, students in vocational trajectories who might benefit from neighbourhood
schools, and, on the other hand, low SES, talented students, who might have better
opportunities in a school with a higher average SES; 4) scientific research that considers
the effect of postponement of the age of tracking in secondary schools, as well as the
availability of comprehensive schools that offer all tracks (thus enhancing chances for
upward mobility); 5) research that considers the distance to school specifically and
different patterns of school choice between ethnic groups. And, finally, 6) the expected
effects of extending school choice.
School composition
The expected mechanism behind desegregation policies (typically involving
greater distances to school) is often based on: 1) the assumed positive peer effects
between pupils from various backgrounds; 2) assumed positive effects of more support
in school activities by higher SES parents; and 3) a probable teacher selection effect,
because high quality teachers might prefer high SES schools. Desegregation in this
context involves the reduction of ethnic segregation between schools. If residential
areas are ethnically segregated, this implies aiming for a school composition that
deviates from the ethnic composition of the neighbourhood where the school is located.
![Page 99: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/99.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
73
Research outcomes on the effects of school composition appear, however, to be
ambiguous. While, according to some authors, the overall effect may be limited, other
authors warn that assumedly underestimated compositional effects or differential
effects should be treated with caution (e.g. Palardy, 2008, Vigdor and Ludwig, 2010).
Thrupp, Lauder and Robinson (2003), in their comparison of school compositional and
peer effects in the United States, the United Kingdom, Belgium and New Zealand, argue
that the little consensus over these effects might be due to inadequate theorizing and
research design. In order to be able to attribute differences in school quality and
outcomes to school or pupil characteristics, it is a prerequisite, so they argue, to be able
to distinguish between the effects of school leadership and instructional quality on the
one hand, and pupil composition and peer effects on the other hand. The authors
emphasize that peer and school effects can, in fact, only be measured adequately, when
prior achievement at the individual pupil level is known. This means in the Dutch context
that the pupils' CITO21 score at the end of elementary education, plus the elementary
teacher's advice for the appropriate track level in secondary education, should be linked
to later actual achievements in secondary education, in order to assess the secondary
school's added value.
Ho Sui‐Chu and Willms (1996) find that children perform better in schools with a
high mean SES of parents, and illustrate that this positive effect is indeed mediated by
parental involvement. The educational systems in the United States and in the
Netherlands differ, however, in important characteristics related to these research
outcomes: although in the United States many schools with low SES pupils receive extra
funding schools also depend on local tax revenues; poor neighbourhoods may more
often have poor schools with a high turnover of teachers, which further burdens poor
children. In the Netherlands all children from low SES parents receive extra funding, and
all schools receive equal standard funding from the government. Schools in the
Netherlands, therefore, hardly depend on financial support by parents. However,
parental assistance in school processes and projects may vary also in the Netherlands,
depending on average levels of schooling of parents.
21 The final test of elementary education; scores range between 500‐ and 550; the national average is 535.
![Page 100: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/100.jpg)
74
Opdenakker and van Damme (2001) suggest that composition effects may be
larger than many researchers assume, but the mechanism behind these effects might
not so much be a peer effect, but also the result from the influence of school
composition on school processes, notably an orderly learning environment and
structural and effective cooperation between teachers. Therefore, the authors suggest
that school effectiveness measurements might be biased in favour of high SES schools,
where high standard outcomes might wrongly be attributed to high standard school
processes, while in fact these are the emergent effects of school composition.
Interestingly, Opdenakker and van Damme (ibid.) find that low SES, high ability students
might be twice as sensitive to school composition as low SES, low‐ability students.
Neighbourhood schools
Ryan Wells (2010) raises the issue that policies aimed at the enhancement of the
levels of schooling of low SES youth, may in fact work out differently for migrant pupils,
than for low SES native‐born pupils. Ryan Wells (ibid.), in accordance with often‐
reported research findings in the Unites States, reports that migrant youths on average
express higher expectations of their future attainment levels than native‐born pupils.
Whether this is also the case in the Dutch context, we do not know. Interestingly, Wells
(ibid.) also reports, seemingly in contrast with these higher expectations among migrant
youth, that immigrant status is associated with (self‐reported) lower expectations in
high SES schools‐ all else being equal. He assumes that the rationale behind these
findings is that migrant pupils may rather benefit from the presence of successful
migrant students as a reference and role model, in schools with high percentages of
migrant students, and may feel isolated in high SES schools.
Especially in the tracked and fairly complex system for secondary education in
the Netherlands, access to information is an important prerequisite for parents in order
to be able to choose the best school for their children. Parents need to be
knowledgeable specifically about the consequences of the choice for a particular
secondary track, considering their children's future options for tertiary education.
According to Cabrera and La Nasa (2001): "SES gaps are reduced, if not eliminated, once
a number of school‐based and family‐oriented factors are taken into account… [these]
practices are as important, if not more, than is family's SES in becoming college
![Page 101: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/101.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
75
qualified…" Following this line of reasoning, primary schools in migrant neighbourhoods
may play an important role in informing parents about the options when choosing a
secondary school. In practice, schools with large numbers of migrant pupils may put
family‐based policies that involve cooperation with parents much higher on the agenda
than schools with more affluent, well‐educated parents.
Finally, Bauder (2002) severely criticizes policies to send migrant pupils to schools
outside their residential area, for the sole purpose of desegregating schools: this may
wrongly stigmatize students attending neighbourhood schools in the view of future
employers, because by implication the message appears to be that neighbourhood
schools in migrant communities are of insufficient quality.
Mixed and differential effects
Cullen, Jacob and Levitt (2000) investigated the effect of expanded parental
choice within the Chicago Public School System. This educational reform resulted in
about half of all the students opting out to attend another school. The authors describe
how this intervention, in the first place, dramatically increased sorting. Disproportionally
more motivated students opted out, and this indeed raised the odds of graduating in
their school of choice. They found no evidence of a negative effect on students who
remained in their assigned neighbourhood school. The overall effect on the
desegregation of schools, however, turned out to be limited. In sum, the authors
demonstrate that expanded choice affects students and their parents differently,
resulting in more motivated students to opt for a new school. With the exception of
what are called the "Career Academies"22, the increased chance to graduate for those
who opt for another school may be correlated with motivation, rather than with the
quality of the chosen school. Finally, in the words of the authors, the "greatest puzzle" is
why so many students opt for another school, while the academic benefits turn out to
be limited. This is an interesting question in the Dutch context as well, where parental
choice is free, but especially migrant students tend to choose a school on average closer
to their residence address than native‐born students.
22 Career Academies promote a college‐preparatory curriculum and career‐focused education in different fields. Students get the opportunity to visit local business, and shadow business professionals in various career areas.
![Page 102: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/102.jpg)
76
Rumberger and Palardy (2005) raise the question of (reversed) causality when
finding that schools with mostly lower‐income students tend to be organized differently
from schools with affluent students: Should school reform introduce effective school
characteristics in schools with many low‐income students? Or are effective school
characteristics the emergent result of having affluent students? In the first option,
school characteristics and processes may make the difference; in the second option, the
composition of students affects outcomes. Furthermore, Rumberger and Palardy (ibid.)
explore whether school policies and organizational characteristics may have a
differential effect on black and white students. Their study demonstrates that teacher
expectations (albeit raising teacher expectations structurally may be a complicated
endeavour) and an academic school climate foster the success of low SES students; if
schools with many low SES students were to focus on these quality aspects, the authors
do not expect that desegregation would offer any extra educational advantages.
Finally, Konstantopoulos and Borman (2011) gained new interesting insights from
their replicated analyses of Coleman's famous data (1966), with current, more advanced,
statistical instruments. Coleman (ibid.) found that the pupils' background was a stronger
predictor for student achievement than school quality. While these findings were
reaffirmed by Konstantopoulos and Borman (ibid.), because of the current availability of
multi‐level models for statistical analyses, they could also demonstrate, however,
significant between‐school variance: "Our results also indicated that schools play
meaningful roles in distributing equality or inequality of educational outcomes to
females, minorities, and the disadvantaged." Similar to Coleman, they found that within‐
school variance is in fact larger than in‐between school variance; but notwithstanding
this, they could also demonstrate that "40 % of the total variability in achievement was
attributable to differences among schools, and ...that schools have nontrivial effects [ ...]
on the achievement gap."
Opportunities for upward mobility
Migrant students may need more time to discover their ambitions and capacities,
for example, because they may lack role models in their own family. Especially when
Dutch is not the language spoken at home, they may also need some extra time to
acquire the level of Dutch language proficiency that is needed for academic tracks in
![Page 103: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/103.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
77
secondary education. This would imply that broad secondary schools that offer all tracks
(vocational and academic), and which are specialized in assessing the capacities of
migrant students repeatedly during their school career, would be an advantage for
migrant students. Recent studies seem to support this hypothesis.
Pekkarinen, Uusitalo and Kerr (2009) made use of the unique opportunity of a
system‐wide school reform in Finland to investigate the effect of postponed tracking on
the correlation between the labour market position of parents and the level of schooling
of their children. Between 1972 and 1977, the former Finnish two‐tracked system was
replaced by a comprehensive secondary school system that shifted the age of choosing
between a professional or an academic track, from age 11 to 16. This later selection
resulted in a substantial decrease of 23 % in intergenerational income elasticity.
Comparable research outcomes (Hanushek and Woessman, 2006, Brunello and
Checchi, 2007, Bauer and Riphahn, 2006) demonstrate that early selection and tracking
negatively affects educational outcomes of low SES students. Considering the
intergenerational elasticity of high‐ and middle‐income students in the Swiss educational
system, Bauer and Riphahn (2006) find that "early tracking increases the absolute
benefit of having highly vs. mid‐way educated parents and magnifies the relative
advantage of highly educated parents." Finally, also Bjorklund and Salvanes (2010) also
find that postponement of tracking may potentially reduce intergenerational
correlations with parental schooling.
Interestingly, Van Elk et al. (2011) make use of the opportunity that in the
Netherlands early tracking at age 12, and postponed tracking at age 13/14, exist in
parallel. As described above, schools, furthermore, may offer all secondary tracks or may
be specialized in either vocational or academic tracks. Considering schools that offer all
tracks (lower vocational and academic), and that start with one or two comprehensive
years, they find that pupils starting in the lower secondary vocational track, have a 26 %
chance to complete higher education later on. Note that without upward mobility to an
academic track, the secondary lower professional track does not directly qualify pupils
for access to higher education. In schools that only offer the secondary professional
tracks (and no academic tracks), however, this chance is 21 %. The findings of van Elk et
al. (ibid.) suggest that upward mobility is more feasible, when all tracks are offered
within the same school. A 5 percentage point increase in the chance to complete higher
![Page 104: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/104.jpg)
78
education, after attending broad secondary schools in the case of students starting on
the pre‐vocational level, seems of extra importance for low SES migrant students, who
may need some more time to discover their ambitions and capacities.
The distance between home and school
Interestingly, Nihad Bunar (2010) explored the question why the children of
immigrants in Sweden would prefer the (segregated) neighbourhood school even when
they know the school is low‐performing and located in a high‐poverty area, while– as is
the case in the Netherlands– school choice is free. Bunar (ibid.) interviewed migrant
pupils in two urban schools in Malmö and Stockholm, and found that neither school
quality, nor a lack of information on school quality, nor the costs of travelling to a school
at greater distance, can fully explain the choice to stay in an urban school in the
proximity of the residential area. Bunar (ibid.) writes: "The answer is to be found in the
process of negotiation taking place within the realms of families and peer groups
oscillating around the importance of relationships that provide safety, the feeling of
belonging and cultural recognition, on the one hand. On the other hand, there are
detrimental effects of categorization and stigmatization attached to immigrants,
neighbourhoods and schools."
Harris, Johnston and Burgess (2007) report in their geodemographic analysis of
ethnicity and school choice in Birmingham (England) that the likelihood of attending the
nearest state‐funded secondary school indeed varies with the ethnic composition of the
neighbourhood. They find that white pupils are always more likely not to attend the
nearest school, and that this likelihood is further increased by greater exposure to other
ethnic groups in the residential area. Harris et al. (ibid.) find evidence that pupils may
prefer to attend a school that is more representative for a pupils' own ethnic group; this
may result in a stronger segregation at the school level, as compared to the
neighbourhood level. Allen (2007) reports similar findings that demonstrate that
choosing another school than the nearest school tends to increase social and ability
sorting. Allen (ibid.) developed a new model, based on the availability of the pupils' zip‐
codes, to compare actual school choice data to the proximity counterfactual. Allen (ibid.)
provides us with a telling illustration of how choosing another school than the nearest
one implies a larger average distance to school as a consequence: "The proximity
![Page 105: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/105.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
79
allocation indicates that the typical journey currently made by a pupil is 60% longer than
the minimum necessary. In fact, over 5 million kilometres of additional travel are made
by 11 ‐ 16 year‐olds every school day…"
Finally, Andersson, Malmberg and Östh (2012) describe how the liberalization of
school choice in Sweden (20 years ago) indeed increased the average distance travelled
to school, but that foreign‐born students travel shorter distances. They find,
furthermore, that Swedish‐born students choose more distant schools, and tend to do
so more often when residential areas have larger proportions of foreign‐born students,
or larger proportions of socio‐economically disadvantaged families.
Extended school choice
Overall, the better match of pupils and schools is the rationale behind, the
expected educational gains of extended school choice (Gibbons, Machin, Silva, 2008).
Secondly, extended school choice is often thought of as an important strategy to reduce
segregation along lines of SES, ethnicity or race (e.g. Cullen et al., 2000). Ladd et al.
(2009) find, however, that schools are to a large extent segregated in the larger Dutch
cities, notwithstanding the fact that school choice has been free and universal for almost
a century. They suggest that free choice may lead to Dutch parents to avoid multi‐
cultural schools, thus increasing segregation.
In summary, the literature reviewed above predicts that school choice may be
related to ethnicity and segregation, to school‐ and neighbourhood socio‐economic
composition and to school quality; furthermore, those effects may be different
depending on pupil characteristics. This has inspired us to choose distance
measurements as the core of our research: first, distance measurements allow for a
comparison of school choice patterns between native Dutch pupils and youth with a
migrant background, relative to the socio‐economic characteristics of their residential
area. Second, measuring distance to the more distant schools as a proxy for the
selectivity of choice, allows for analyses of school characteristics that may drive school
choice, like average exam scores and the percentage of upward mobility to a higher
track, but also school ethnic composition and school average SES. Third, distance
measurements may shed a light on the possible benefits of deliberately choosing a
school at further distance.
![Page 106: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/106.jpg)
80
4.3 Data and Methods
4.3.1 BRON data and additional data sources
Starting in 2001 (based on a new law), DUO23, the government agency that
collects all educational data linked to the individual social security number, is in the
process of constructing a new educational database called BRON24. This database covers
all relevant socio‐economic background characteristics of pupils, including their
complete educational history. From 2008 on, all pupils in elementary education will be
recorded in the BRON data, starting with pupils in the final grade of elementary school in
that year; at the time of our research this first cohort had reached the third grade of
secondary education. In the Dutch tracked secondary system, many schools are (partly)
comprehensive for the first two years. In the third year, however, the majority of pupils
have been placed in a specific track. This offered us the opportunity to relate the track
level in the third year at secondary school, to prior achievement in elementary
education. We have added a list with the definitions of all variables we used in this
chapter in Section 4.6.1.
4.3.2 Ethnic diverse populations in the four major cities
In the current chapter we focus on a comparison of school choice patterns of
pupils with a Dutch background, and those with a migrant family history. Since the
majority of migrant students live in the four major cities (Amsterdam, Rotterdam,
Utrecht, and The Hague), we based our analyses on youths living in these cities. Another
consideration for our choice to concentrate on these four cities was our assumption that
in smaller cities, and especially in the countryside, school choices may structurally differ
from those taken in urban environments. For example, considering the much lower
density of schools in the countryside, school choice could be driven to a large extent by
the sheer presence of a school.
Our total data set contains in total 170.465 individuals living in the Netherlands,
who were enrolled in the last grade of elementary school in 2008. The data cover
information from 2008 up to and including 2011. From this total data set we selected
pupils living in the four major Dutch cities throughout the years 2008‐2011, 17.192
23 Dutch acronym: Dienst Uitvoering Onderwijs. 24 Dutch acronym: Basis Register Onderwijs; official educational database.
![Page 107: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/107.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
81
individuals. Next we dropped 997 individuals with a distance from home to school ≥ 10
km (elementary school) or ≥ 20 km (secondary school). These eliminations resulted in
16.195 individuals. We dropped, furthermore, 124 individuals because of incomplete
distance data.
This resulted in N = 16.071 individuals we kept for analyses. This includes 2624
pupils living in one of the four cities, who were enrolled in a school outside the city (but
within 20 km). The four major cities differ in size, urbanicity and unobserved
characteristics. Therefore, we added city dummies in our estimates to account for such
aggregate effects. When we introduced the characteristics of the nearest school as an
explaining variable, this reduced the number of individuals to 11.023, since we do not
have relative exam scores for the nearest school in the case of all individuals.
In our analysis of the difference between pupils who choose the nearest school
to their residence, and those who do not, our counts demonstrate that 1824 pupils
chose the nearest school, while 14.247 chose a school at further distance (total 16.071).
For measuring individual upward mobility to a higher track in secondary school,
we selected all pupils who were enrolled in Year 1 of the lower vocational track (VMBO),
and enrolled in Year 3 at the time of our research: 4343 pupils.
Unless clearly indicated otherwise, we distinguish five ethnic groups: 1) native
Dutch pupils; pupils with a 2) Surinamese/Antillean‐, 3) Turkish‐, or 4) Moroccan
background; and 5) the combined group of all other non‐Dutch backgrounds, referred to
in the tables and figures as "other background". The last group is very diverse and
includes many nationalities and a wide variance in educational attainment and in
motives for immigration (e.g. the children of high skilled labour immigrants, but also the
children of low educated refugees). The data set includes 6368 pupils of Dutch origin,
2450 of Moroccan descent, 1995 of Turkish descent, and 2269 came from Suriname or
the Dutch Antilles. 2928 individuals had another type of immigrant background; for 43
individuals we did not know their ethnicity. The vast majority of pupils with a Turkish or
Moroccan background were born in the Netherlands; they belong to the second
generation immigrants. For the sake of sufficient statistical power, in some tables a
![Page 108: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/108.jpg)
82
distinction was made between the combined group of all immigrants with a non‐
Western migrant background25 and others.
4.3.3 Distance measurements
For our distance measurements we used the BRON information on the pupils'
residence (4‐digit postal code) and the school address (6‐digit postal code). A student's
geographic coordinates refer to the centroid of the neighbourhood of residence;
neighbourhoods are defined by their corresponding 4‐digit postcode. In compliance with
privacy laws, the data do not allow the student to be located at the 6‐digit postal code
and the individual household level. The average number of individuals per 4 digit
postcode area is 100.7 (SD 52.3)26. Correspondingly, in our data set clusters of around
100 students on average, residing in the same 4‐digit zip‐code area, share the same
residential location. For each school, geographic coordinates are available at the exact 6‐
digit zip‐code school location level.
Using these geographic coordinates for students and schools, Euclidian distances
were calculated in order to derive the distance‐to‐school measures. For each student,
the school‐distance measure is thus defined as the Euclidian distance between the
centroid of the student's 4‐digit zip‐code neighbourhood, and the 6‐digit zip‐code school
location he/she attends, measured in kilometres. We constructed a data set that covers
the distance to the nearest relevant school (i.e. the nearest school that indeed offers the
track the pupil is enrolled in) and the distance to the actually attended school (if
different), for every individual.
We used three types of distance measurements: 1) the absolute distance from
home to school; 2) the difference in distance between the nearest and the actual school;
3) a dummy for not attending the nearest school.
25 The following definitions by Statistics Netherlands (CBS, the national statistics office, www.cbs.nl) have been used:
Western immigrant: someone originating from a country in Europe (exclusive of Turkey), North America, Oceania, Indonesia or Japan.
Non‐Western immigrant: someone origination from Africa (in the Netherlands the majority group of immigrants from Morocco), South America, Asia (exclusive of Indonesia and Japan) or Turkey.
26 Percentiles of the numbers of individuals in our data set living in one 4‐digit postal code area: 41 (10%); 61 (25%); 90 (50%); 125 (75%); 174 (90%).
![Page 109: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/109.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
83
For some analyses, we also make use of density measures, capturing the number
of schools within a reasonable distance of the pupils' residence. We used the data by
Statistics Netherlands (CBS) on the number of schools within a 5 km radius.
4.3.4 SES indices
We included information on the student's residential area characteristics. In
particular, we make use of data collected by the Netherlands Institute for Social
Research (SCP). These data provide us with a socio‐economic index for every
neighbourhood, known as the "status scores".27 Information for these status scores is
collected through household (telephone) surveys, one household per 6‐digit zip‐code.
Next, these data are aggregated at the 4‐digit zip‐code level. The data include
neighbourhood mean income, employment and level of schooling (all self‐reported by
inhabitants). In our figures we present poverty indices in the reverse order, compared to
SCP: in all our figures on neighbourhood SES and distance to school, the standardized
poverty score ranges from ‐4 (poor) to +4 (affluent). Lastly, we merged our data with
additional data on neighbourhood characteristics that are updated annually (e.g.
demographics and urbanicity), provided by Statistics Netherlands (CBS).28
4.3.5 Indicators of pupils' prior achievement and secondary school quality
Prior achievement in elementary education is of crucial importance for the
assessment of results in secondary schools. At the end of elementary education, children
have a final test (the CITO test). The score on this test, plus the recommendation of the
pupils' elementary teacher, typically determine the level of enrolment in a specific track
in Dutch secondary education. The BRON database we used contains both the score on
the final elementary CITO test, and the teacher's recommendation, but unfortunately,
we only had the results of the CITO test in the case of 3208 individuals and the
elementary teacher's advice in the case of 6081 pupils. We tested these subsets of the
data on all control variables we use; we found that the regression coefficients of the
controls remained stable. Therefore, we trusted that the subsets did not differ from the
total set of 16.071 individuals in major ways. Furthermore, we compared the subsample
with only the CITO score, and the group with only the teacher's advice. The pupils
27 SCP data set "Statusscores Postcodegebieden 2006". 28 CBS data set "Buurtkaart met Cijfers 2008" (update 2).
![Page 110: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/110.jpg)
84
without CITO scores had a slightly lower mean teacher's advice score, but similar
individual and neighbourhood characteristics. We acknowledge the limitation of
introducing a smaller subsample with information on prior achievement of pupils;
however, we considered this additional information relevant in combination with the
data on school quality we use in our analyses, in our search for quality‐driven elements
(at the individual and school level) of school preference. The Inspectorate of Education
(2011) found no evidence that children with a migrant background receive a lower
advice from their elementary teacher than native Dutch children.
We used data provided by the Inspectorate of Education for our measurements
of school mean exam scores and upward mobility to higher tracks, as indicators of
secondary school quality (see Section 4.6.1 for definitions of the variables).
In the tables we present below, we diversified between three different main
tracks in Dutch secondary education: VMBO (lower secondary vocational education),
HAVO (senior general education), and VWO (pre‐university education), plus the
combined advice HAVO/VWO. We used VMBO (the lowest level) as a reference, and the
other levels as dummy variables. We included pupils who are eligible for extra funding
because of special educational needs; these pupils can, in principle, enrol in every
school. In the Netherlands only pupils with specific special needs (e.g. blind children)
attend special schools; these pupils are not listed in our data set.
4.3.6 Limitation
We had to accept some limitations, because, as we described earlier, the
development of the BRON data set is currently in progress. As a consequence, for the
time being, we could not link results on (future) final secondary exams to prior
achievement in elementary education. We attempted to bypass this barrier somewhat,
by using mean exam results at the school level (collected by the Inspectorate of
Education) to estimate whether distance to the preferred school, controlling for
individual achievement in elementary school, is associated with school quality, as
expressed in the average exam score.
4.4 Results
In this Section we start with a general overview of summary statistics of the total
data set of 16071 pupils living in the four Dutch major cities, who were enrolled in the
![Page 111: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/111.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
85
last grade of elementary education in 2008; most of them were enrolled in Year 3 of
secondary school (2011) at the time of our research (Table 4‐1). Next we present a first
overall analysis of the association between distance to school and control variables at
the individual, neighbourhood, and school level; we added the characteristics of the
nearest school as an explaining variable for school choice (Table 4‐2). We continue with
a comparison of the group of pupils who choose the nearest school, and those who opt
out for a school at greater distance; we present the summary statistics in Table 4‐3, and
regression analyses in Table 4‐4. In Table 4‐5 we consider the distance difference
between the nearest and the actual school, as a proxy for the selectivity of choice,
assuming that pupils who take the extra effort to travel further made a more deliberate
choice. In Table 4‐6 we present a comparison between the nearest school and the actual
school, with a focus on patterns of ethnic segregation. We investigate these patterns
specifically by considering the SES difference and the difference in percentages of
migrant pupils, when comparing the nearest (not chosen) and the actual school (Tables
4‐7 and 4‐8). Finally we look at upward mobility to a higher track, as a specific example
of a possible benefit of selective school choice in Table 4‐9.
4.4.1 A general overview of socio‐economic measurements and distance to school
As a visualization of our data, we have added in Appendix 4‐1 the maps of the
four cities, which present a view of the ethnic composition of neighbourhoods, the
average distance travelled to school, and the distribution of secondary schools. The
maps give a first impression of the association between the ethnic composition of an
area and the distance travelled to school.
In Table 4‐1, we set the scene and present the summary statistics for the four
ethnicities (including native Dutch) that we distinguish in this chapter, plus the
combined group "other immigrants". Counts were carried out on the pupil‐, school‐ and
neighbourhood level. In addition to socio‐economic variables, we present school quality
indicators (mean exam score and percentage of pupils with upward mobility to a higher
track) and indicators of pupils' prior achievement in elementary school (the CITO score
and elementary teacher's advice for the track level in secondary school).
![Page 112: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/112.jpg)
86
Table 4‐1: Summary statistics for pupils living in the four major Dutch cities, at the individual, neighbourhood and school level
Dutch Sur./Ant. Turkish Moroccan Other Imm. Total mean mean mean mean mean mean
VARIABLE (SD) (SD) (SD) (SD) (SD) (SD) individual level distance to primary school (km) 1.16 1.29 0.93 0.91 1.23 1.12 (1.21) (1.44) (1.17) (1.15) (1.4) (1.27) distance to secondary school (km) 3.49 3.30 2.92 2.67 3.17 3.21 (2.4) (2.48) (2.02) (1.89) (2.20) (2.28) school outside municipality border 0.21 0.17 0.13 0.10 0.12 0.16 (0.41) (0.38) (0.34) (0.30) (0.33) (0.37) CITO‐score 537.44 529.32 528.77 528.29 533.51 533.16 (9.93) (10.59) (9.68) (10.64) (11.03) (11.03) VMBO advice 0.42 0.72 0.75 0.76 0.54 0.58 (0.49) (0.45) (0.43) (0.43) (0.50) (0.49) HAVO advice 0.15 0.11 0.12 0.12 0.15 0.14 (0.36) (0.31) (0.32) (0.32) (0.36) (0.34) HAVO/VWO advice 0.18 0.10 0.07 0.08 0.14 0.13 (0.39) (0.30) (0.26) (0.26) (0.35) (0.34) VWO advice 0.24 0.07 0.06 0.05 0.17 0.15 (0.43) (0.26) (0.23) (0.22) (0.37) (0.36) individual upward mobility in Year 3 0.18 0.09 0.07 0.09 0.17 0.14 (0.39) (0.29) (0.26) (0.29) (0.38) (0.35) male 0.50 0.50 0.51 0.47 0.51 0.50 (0.50) (0.50) (0.50) (0.50) (0.50) (0.50) non‐western background 0.00 1.00 1.00 1.00 0.60 0.53 (.) (.) (.) (.) (0.49) (0.50) 2nd generation non‐western background 0.00 0.85 0.91 0.93 0.81 0.52 (.) (0.36) (0.28) (0.26) (0.39) (0.50) weighted student funding 0.03 0.33 0.52 0.57 0.32 0.27 (0.10) (0.43) (0.44) (0.43) (0.42) (0.40) one‐parent household 0.06 0.26 0.02 0.02 0.12 0.09 (0.24) (0.44) (0.15) (0.12) (0.32) (0.28) nearest school chooser 0.12 0.14 0.10 0.10 0.11 0.11 (0.32) (0.34) (0.30) (0.30) (0.31) (0.32) distance difference chosen and nearest school 2.40 2.39 2.09 1.85 2.29 2.26 (2.26) (2.44) (1.96) (1.79) (2.12) (2.17) neighbourhood level neighbourhood SES index 0.84 ‐0.47 ‐0.91 ‐0.90 ‐0.12 0.00 (1.33) (1.33) (1.16) (1.14) (1.44) (1.50) urbanicity 3.42 3.60 3.81 3.81 3.67 3.60 (0.76) (0.57) (0.40) (0.39) (0.53) (0.63) distance to nearest relevant school (km) 1.08 0.91 0.83 0.82 0.88 0.95 (0.84) (0.64) (0.63) (0.64) (0.65) (0.74) # of relevant schools within 5 km 15.44 17.36 21.19 20.85 18.47 17.81 (8.00) (9.52) (9.07) (9.26) (9.35) (9.10) school level # of tracks offered at secondary school 2.76 2.93 2.99 3.07 2.86 2.88 (1.18) (1.30) (1.41) (1.28) (1.30) (1.27) relative exam scores secondary school 0.09 ‐0.04 ‐0.10 ‐0.13 0.02 0.01 (0.26) (0.25) (0.27) (0.26) (0.29) (0.28) % upward mobility at school level 15.50 19.10 21.05 22.23 17.28 17.93 (9.26) (8.73) (9.78) (10.27) (9.51) (9.79) N 6386 2269 1995 2450 2928 16071
The mean average distance to secondary school is lowest for Moroccan youth at
2.67 km. Native Dutch pupils more often attend the nearest school (12%) than migrant
groups, with the exception of youth with a Surinamese/Antillean background (14%). Like
native Dutch students, Surinamese and Antillean students travel a larger mean distance
to school, but at the same time more often choose the nearest school. Presumably
Surinamese and Antillean pupils prefer a school with a large share of pupils with the
same background; consider for example Amsterdam: a large share of Surinamese and
Antillean pupils live in the Bijlmer district in the south‐eastern part of Amsterdam, and
![Page 113: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/113.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
87
tend to go to local schools with large percentages of pupils belonging to this group; this
might explain why they more often attend the nearest school than is the case for other
groups. At the same time, Surinamese and Antillean pupils who do not live in the Bijlmer
district, may still prefer to attend a school with many Surinamese and Antillean students
in the Bijlmer district; this could explain the higher mean distance to school.
Native Dutch pupils prefer more often prefer the nearest school; the odds are
that these students more often live in more affluent areas than is the case for migrant
students, and tend to attend a local school with a population that mirrors the affluent
neighbourhood. For example 57% of Moroccan pupils are eligible for weighted student
funding in elementary education, against 3% among native Dutch pupils.
The mean neighbourhood SES index differs substantially between the different
ethnic groups, by more than 1 Standard Deviation, and between native Dutch pupils and
all other groups. The share of Dutch pupils enrolled in a school outside the city
boundaries, at 21% differs considerably from, for example, students with a Moroccan
background (10%). This alludes to our further findings further below, which illustrate
that native Dutch pupils tend to prefer the nearest school when they live in an affluent
neighbourhood, but leave lower SES residential areas to go to school elsewhere more
often than migrant pupils living in the same area.
Considering pupils' prior achievement, the mean CITO score of native Dutch
pupils is close to 1 Standard Deviation higher than, for example, it is for pupils of
Moroccan descent. Substantially more often, migrant children are advised to proceed in
secondary vocational education (the ranking order of secondary tracks is VWO (highest),
HAVO, VMBO); 24% of native Dutch students are advised to proceed to the highest
secondary track (pre‐university track) from their elementary school teacher, against 5%
and 6% of pupils with, respectively, a Moroccan or Turkish background.
Successful progress in secondary school, as measured by upward mobility to a
higher track, differs substantially between native Dutch and migrant students: among
immigrant groups, 7‐9% move up to a higher secondary track at the start of Year 3 of
![Page 114: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/114.jpg)
88
secondary school, whereas this is the case for 18% of the Dutch students29. Besides the
percentage of upward mobility, we considered the school's relative exit‐exam score to
be an important indicator of school quality; we defined this indicator as the
performance at each track level, measured by the mean exit‐exam score, compared with
the mean exit‐exam score of all other schools in the four major cities which offer this
same track. The school mean exam score is higher in the case of native Dutch pupils.
4.4.2 Distance to school
In Table 4‐2 we present OLS regressions for the individual distance travelled to
school. Except for Surinamese/Antillean students, immigrant students travel less far to
school than native Dutch students. This conclusion still holds with the inclusion of
additional controls. The magnitude of the difference declines if we add school and
neighbourhood characteristics, but is restored if we add school advice and CITO score.
Specification 1 (individual student characteristics) demonstrates that poorer students
(who were eligible for weighted student funding in elementary education), among
whom there are many Turkish and Moroccan migrant students, on average travel less
far to school.
In Specification 2 we introduce neighbourhood characteristics: an increase in
neighbourhood SES decreases distance to school. This may seem in contrast with what
we find at the individual level: poorer students (as measured by the mark‐up on funding
per pupil in elementary education– weighted student funding) travel less far to school.
However, while poorer migrant students tend to attend their neighbourhood school, the
same also appears to be the case for affluent native Dutch youth in higher SES
neighbourhoods.
29 Data files of the Inspectorate of Education show, as a reference, that native Dutch pupils are enrolled in schools with a mean 15.5 % upward mobility (measured after the completion of final exams), while Turkish pupils are enrolled in schools with a mean 21.1 % upward mobility, and Moroccan pupils in schools with a mean 22.2 % upward mobility. This indicates that migrant students more often attend a school with a larger upward mobility, but do so less themselves (as table 1 shows) than native‐born students.
![Page 115: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/115.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
89
Table 4‐2: Regression results (OLS) for distance to secondary school in km, on individual‐, school‐ and neighbourhood variables (standard errors in parentheses).
(1) (2) (3) (4) (5) distance to distance to distance to distance to distance to VARIABLES sec. school sec. school sec. school sec. school sec. school individual level male 0.10*** 0.08** 0.05 0.16** 0.20** (0.038) (0.038) (0.047) (0.063) (0.081) weighted student funding ‐0.30*** ‐0.18*** ‐0.19** ‐0.28*** ‐0.09 (0.076) (0.056) (0.074) (0.106) (0.155) Surinamese/Antillean ‐0.11 ‐0.13 ‐0.12 ‐0.21 ‐0.18 (0.124) (0.089) (0.097) (0.137) (0.216) Turkish ‐0.41*** ‐0.18* ‐0.23 ‐0.35** ‐0.48* (0.120) (0.102) (0.138) (0.158) (0.258) Moroccan ‐0.65*** ‐0.52*** ‐0.54*** ‐0.65*** ‐0.65*** (0.111) (0.082) (0.107) (0.148) (0.243) Other Immigrant Background ‐0.24*** ‐0.18*** ‐0.21*** ‐0.30*** ‐0.34** (0.091) (0.064) (0.076) (0.097) (0.161) one‐parent household 0.08 0.12* 0.12 0.07 ‐0.02 (0.088) (0.069) (0.083) (0.102) (0.133) HAVO advice track dummy ‐0.84*** (0.109) HAVO/VWO advice track dummy ‐0.89*** (0.122) VWO advice track dummy ‐0.61*** (0.168) CITO test score ‐0.02*** (0.007) Non‐western x CITO test score 0.00 (0.000) neighbourhood level neighbourhood SES index ‐0.21*** ‐0.14** ‐0.04 0.05 (0.052) (0.068) (0.078) (0.125) Surinamese/Antillean x SES 0.14** 0.13* 0.12 0.14 (0.059) (0.072) (0.096) (0.124) Turkish x SES 0.19*** 0.19** 0.16* 0.17 (0.064) (0.076) (0.090) (0.113) Moroccan x SES 0.19*** 0.21** 0.15 0.04 (0.069) (0.085) (0.097) (0.169) Other Immigrant Background x SES 0.08** 0.09* 0.06 ‐0.02 (0.041) (0.046) (0.058) (0.094) urbanicity ‐0.33*** ‐0.42*** ‐0.19 ‐0.28 (0.113) (0.157) (0.155) (0.187) # of relevant schools within 5 km ‐0.03*** ‐0.04*** ‐0.05*** ‐0.05*** (0.008) (0.009) (0.009) (0.011) distance to nearest relevant school (km) 0.59*** 0.47*** 0.45*** 0.51*** (0.117) (0.139) (0.143) (0.163) Utrecht municipality dummy ‐0.13 0.12 0.10 0.07 (0.193) (0.250) (0.255) (0.303) The Hague municipality dummy ‐0.55*** ‐0.67*** ‐0.49** ‐0.88*** (0.139) (0.189) (0.193) (0.257) Rotterdam municipality dummy ‐0.51*** ‐0.50*** ‐0.32* ‐0.53*** (0.137) (0.174) (0.177) (0.201) school level relative exam scores nearest school ‐0.32 ‐0.30 ‐0.17 (0.266) (0.262) (0.322) # of tracks offered at nearest school ‐0.02 ‐0.01 ‐0.10 (0.051) (0.054) (0.080) % upward mobility at nearest school ‐0.49 0.09 ‐0.38 (0.606) (0.570) (0.704) average SES index at nearest school ‐0.25*** ‐0.27*** ‐0.42*** (0.093) (0.104) (0.130) Constant 3.44*** 4.89*** 5.48*** 5.08*** 18.22*** (0.117) (0.600) (0.828) (0.805) (3.695) Observations 16,071 16,060 11,023 6,081 3,208 R‐squared 0.02 0.14 0.14 0.16 0.18 Adj. R‐squared 0.0200 0.139 0.139 0.158 0.171
Notes: Robust clustered standard errors (at neighbourhood level) in parentheses *** p<0.01, ** p<0.05, * p<0.1
![Page 116: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/116.jpg)
90
In Specification 3 we introduce the characteristics of the nearest relevant school
as an explaining variable. The two school quality indicators we consider in this chapter–
mean exam score and percentage upward mobility– do not seem to drive school choice.
A higher average SES of the nearest school, however, decreases distance to school, in
accordance with what we described above: native Dutch pupils in richer areas tend to
attend their neighbourhood school.
In Specification 4 we add the elementary teacher's advice for the pupils'
appropriate track in secondary school. Pupils who are advised to follow the academic
secondary tracks (HAVO and VWO), relative to the lowest professional track (VMBO),
travel less far to school. This again seems related to what we described above: native
Dutch pupils more often have affluent parents and live in higher SES residential areas;
the children of more affluent, better‐educated parents more often enrol in academic
tracks (e.g. 58% among native Dutch, and 25% among pupils of Moroccan descent, Table
4‐1). We have seen that affluent native Dutch pupils tend to attend their neighbourhood
school, hence the lower mean distance to school among pupils in academic tracks.
More migrant than native Dutch students start in secondary vocational tracks
(VMBO); on average, enrolment in a vocational track may require somewhat longer
travel distances: in contrast to academic tracks, at the vocational level a wide range of
professional programmes is offered (that correspond to professions on the labour
market), but not all these options are offered in every school. Therefore, enrolling on
the preferred vocational programme may imply further travelling to school.
Finally, in Specification 5, we add the CITO test score; unfortunately, similar to
the elementary teacher's advice, these analyses are based on a substantially smaller
subsample. We decided, however, to include these regressions, because we found
hardly any instability to controls. We trusted, therefore, that the subsample may not
differ in major ways from the total data set. The effect of the pupils' CITO score on
distance to school is marginal but statistically significant.
As we presented in Table 4‐1, only 10‐14% (depending on the ethnicity) of pupils
chooses the nearest school to their residence. For this reason, we looked further into
possible differences between the two groups, those who choose the nearest school, and
those who do not.
![Page 117: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/117.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
91
Figure 4‐1 demonstrates these two effects. The interaction of ethnicity with
neighbourhood SES, furthermore, diminishes the effect of ethnicity on distance to
school: when neighbourhood SES increases, the odds are that Turkish and Moroccan
students travel further to school. We see an opposite effect between native Dutch and
migrant students here: while the first group tends to choose a school closer to home in a
high SES area, our analyses estimate the opposite effect for migrant students.
A greater distance to the nearest relevant school: a greater distance to the
nearest school may point at a lower population density and, accordingly, lower school
density. In that case, a preferred school may tend to be further away than in a densely
populated area.
Figure 4‐1: Linearly fitted lines (OLS), one for each of the 5 ethnic groups on the correlation between absolute distance to secondary school and relative neighbourhood SES
4.4.3 Choosing the nearest school or not
Only 1824 (11.3%) pupils chose the nearest school to their residence, while
14.247 (88.7%) pupils chose another school, illustrating that the right to choose freely is
largely exercised (Table 4‐3). Note that preferring the nearest school may also be a
deliberate choice, however, in this case we cannot know; therefore, the percentage of
pupils who actively choose a school could in reality even be higher. In these descriptive
![Page 118: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/118.jpg)
92
measurements we used the higher aggregate level of the combined group of non‐
Western immigrants, in order to avoid losing too much statistical power if we were to
specify 1824 pupils (nearest school choosers) into five ethnicities.
Table 4‐3: Summary statistics of the comparison between the group of students who choose the nearest relevant school (i.e. a school that indeed offers a desired track) to their home, and those who do not (standard deviations in parentheses)
Non‐Nearest Nearest mean meanVARIABLE (SD) (SD)individual level distance to primary school (km) 1.13 1.08 (1.28) (1.19) distance to secondary school (km) 3.49 1.04 (2.26) (.85) CITO test score 532.96 534.53 (11.06) (10.77) VMBO advice 0.59 0.47 (.49) (.5) HAVO advice 0.13 0.18 (.34) (.39) HAVO/VWO advice 0.13 0.15 (.34) (.36) VWO advice 0.15 0.19 (.35) (.4) individual upward mobility in Year 3 0.14 0.15 (.34) (.36) male 0.50 0.51 (.5) (.5) non‐western background 0.53 0.53 (.5) (.5) 2nd generation non‐western background 0.52 0.52 (.5) (.5) weighted student funding 0.27 0.25 (.4) (.39) one‐parent household 0.09 0.09 (.28) (.29) nearest school chooser 0.00 1.00 (.) (.) distance difference chosen and nearest school 2.55 0.00 (2.14) (.)neighbourhood level neighbourhood SES index ‐0.03 0.17 (1.49) (1.57) urbanicity 3.62 3.42 (.61) (.76) # of relevant schools within 5 km 18.17 15.01 (9.04) (9.17) distance to nearest relevant school (km) 0.94 1.04 (.72) (.85)school level # of tracks offered at secondary school 2.86 3.04 (1.28) (1.2) relative exam scores secondary school 0.01 0.00 (.28) (.27) % upward mobility at school level 17.89 18.32 (9.87) (9.15) N 14247 1824
Measurements overall demonstrate only marginal differences between the
groups of nearest and non‐nearest school choosers. The main differences we found
concern the elementary teacher's advice to follow the vocational track (VMBO) and
neighbourhood SES (richer native Dutch pupils who choose the nearest school). As we
described above, attending the preferred VMBO programme may require further
![Page 119: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/119.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
93
travelling, since not all schools offer all vocational programmes. We found virtually no
difference between the two groups considering immigrant family background or
eligibility for weighted student funding in elementary school.
Table 4‐4: Regression results (OLS) for choosing the nearest school (i.e. comparing pupils who choose the nearest school with those who choose another school), with choosing the nearest school = 1
(1) (2) (3) (4) (5)VARIABLES nearest nearest nearest nearest nearest male 0.00 0.00 0.01 0.01 0.02 (0.006) (0.006) (0.007) (0.009) (0.012) weighted student funding ‐0.01 0.01 0.01 ‐0.00 ‐0.01 (0.009) (0.009) (0.011) (0.015) (0.023) Surinamese/Antillean 0.03 0.03* ‐0.03 ‐0.03 ‐0.02 (0.017) (0.014) (0.022) (0.024) (0.036) Turkish ‐0.01 0.01 ‐0.03 ‐0.03 0.01 (0.014) (0.014) (0.022) (0.025) (0.040) Moroccan ‐0.01 0.01 ‐0.03 ‐0.01 0.01 (0.015) (0.013) (0.024) (0.028) (0.037) Other Immigrant Background ‐0.00 0.01 ‐0.02 ‐0.01 ‐0.01 (0.011) (0.009) (0.014) (0.018) (0.027) one‐parent household ‐0.01 ‐0.01 ‐0.02 0.00 ‐0.00 (0.009) (0.008) (0.011) (0.016) (0.027) neighbourhood SES index 0.00 ‐0.01 ‐0.01 ‐0.01 (0.007) (0.010) (0.012) (0.016) Non‐western x SES ‐0.01 0.01 0.00 0.00 (0.008) (0.010) (0.012) (0.015) urbanicity ‐0.05*** ‐0.04** ‐0.06** ‐0.05* (0.017) (0.022) (0.025) (0.030) # of relevant schools within 5 km ‐0.00*** ‐0.00*** ‐0.00*** ‐0.00*** (0.001) (0.001) (0.001) (0.001) distance to nearest relevant school (km) ‐0.02 ‐0.00 ‐0.00 ‐0.00 (0.017) (0.026) (0.028) (0.033) Utrecht municipality dummy ‐0.04* ‐0.06* ‐0.09** ‐0.09** (0.024) (0.034) (0.037) (0.041) The Hague municipality dummy ‐0.01 ‐0.01 ‐0.02 ‐0.03 (0.018) (0.025) (0.027) (0.032) Rotterdam municipality dummy ‐0.03 ‐0.02 ‐0.03 ‐0.01 (0.019) (0.023) (0.024) (0.034) relative exam scores nearest school 0.06 0.06 0.05 (0.036) (0.042) (0.056) # of tracks offered at nearest school ‐0.01 ‐0.01 ‐0.00 (0.006) (0.007) (0.011) % upward mobility at nearest school 0.00 ‐0.04 ‐0.06 (0.076) (0.091) (0.121) average SES index at nearest school 0.02 0.02 0.04 (0.017) (0.018) (0.025) % non‐western at nearest school 0.00 ‐0.02 ‐0.06 (0.059) (0.053) (0.084) % non‐western at nearest school x non‐western 0.09*** 0.09*** 0.17*** (0.026) (0.030) (0.064) HAVO advice track dummy 0.05** (0.023)HAVO/VWO advice track dummy 0.01 (0.027)VWO advice track dummy 0.02 (0.028)CITO test score 0.00 (0.001) Non‐western x CITO test score ‐0.00 (0.000) Constant 0.11*** 0.37*** 0.39*** 0.47*** ‐0.24 (0.011) (0.072) (0.096) (0.108) (0.576) Observations 16,071 16,017 10,993 6,065 3,208 R‐squared 0.00 0.02 0.03 0.04 0.05 Adj. R‐squared 0.000901 0.0189 0.0287 0.0400 0.0436
Notes: Robust clustered standard errors (at neighbourhood level) in parentheses *** p<0.01, ** p<0.05, * p<0.1
![Page 120: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/120.jpg)
94
Table 4‐4 lists the regressions (OLS) on these differences between nearest and
non‐nearest school choosers, and we used again the characteristics of the nearest
school as explanatory variables for the choice for the proximity school or a school
further from the residence. Similar to the summary statistics in Table 4‐3, regressions in
Table 4‐4 demonstrate hardly any dissimilarity between pupils choosing the nearest
school, and those who do not. The key result is a significant effect of interaction
between ethnicity and ethnic school composition: non‐Western pupils are more likely to
choose the nearest school if this school holds more non‐Western pupils. The coefficients
(not statistically significant) we found considering the student's ethnic background are
not sensitive to additional controls. Understandably, urbanicity– a measure for human
activity in a neighbourhood based on the number of addresses– drives the choice for a
more distant school: there are more schools to choose from in densely populated areas,
and, as a consequence, the distance difference between the nearest and the preferred
school tends to be smaller; the same can be said for the number of schools within 5 km.
Recapitulating the above, we have seen that socio‐economic indicators influence
distance to school, but that the comparison between pupils who attend the nearest
school and those who prefer another school does not demonstrate major differences
between these two groups. Next, we introduce measurements of the distance difference
between the nearest and the actual school as a proxy for the selectivity of choice.
4.4.4 Selectivity of choice
Unlike in Table 4‐2, where we measured the absolute distance from home to
school, in Table 4‐5 we consider the distance difference between the nearest and the
actual school, as a measure for the selectivity of choice, assuming that selecting a school
at greater distance implies a more deliberate choice, and has a higher cost. In
accordance with the measurements we presented above, an increase in neighbourhood
SES decreases this distance difference significantly, indicating that pupils living in higher
SES areas, more often choose the nearest school (i.e. more pupils with distance
difference = 0). As we also found earlier (Table 4‐2), the interaction term non‐Western
migrant background x SES neighbourhood reduces the effect of ethnicity considerably.
![Page 121: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/121.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
95
Table 4‐5: Regression results (OLS) for the distance difference (i.e. between the actual and the nearest school), with distance difference = 0 for pupils who choose the nearest school
(1) (2) (3) (4) (5) (6)VARIABLES Dist. Diff Dist. Diff. Dist. Diff. Dist. Diff. Dist. Diff. Dist. Diff.
male 0.08** 0.08** 0.05 0.08* 0.16** 0.20** (0.038) (0.038) (0.047) (0.047) (0.063) (0.080)weighted student funding ‐0.22*** ‐0.20*** ‐0.20*** ‐0.22*** ‐0.30*** ‐0.11 (0.064) (0.055) (0.073) (0.068) (0.105) (0.149)Surinamese/Antillean 0.01 ‐0.10 ‐0.02 ‐0.06 ‐0.10 ‐0.24 (0.126) (0.086) (0.122) (0.124) (0.164) (0.205)Turkish ‐0.20** ‐0.18** ‐0.17 ‐0.26* ‐0.26 ‐0.59*** (0.090) (0.085) (0.147) (0.158) (0.175) (0.218)Moroccan ‐0.43*** ‐0.52*** ‐0.50*** ‐0.65*** ‐0.55*** ‐0.66*** (0.087) (0.070) (0.138) (0.156) (0.158) (0.207)Other Immigrant Background ‐0.07 ‐0.13** ‐0.12 ‐0.19** ‐0.20* ‐0.33** (0.079) (0.062) (0.087) (0.091) (0.119) (0.150)one‐parent household 0.20** 0.13* 0.14* 0.11 0.09 ‐0.03 (0.088) (0.069) (0.082) (0.105) (0.103) (0.133)neighbourhood SES index ‐0.19*** ‐0.11 ‐0.16** ‐0.01 0.07 (0.050) (0.067) (0.076) (0.077) (0.119)Non‐western x SES 0.13*** 0.12** 0.17*** 0.07 0.06 (0.044) (0.057) (0.058) (0.066) (0.096)urbanicity ‐0.32*** ‐0.43*** ‐0.72*** ‐0.20 ‐0.32* (0.112) (0.155) (0.197) (0.151) (0.181)# of relevant schools within 5 km ‐0.03*** ‐0.04*** ‐0.05*** ‐0.05*** ‐0.05*** (0.008) (0.009) (0.011) (0.009) (0.011)distance to nearest relevant school (km) ‐0.40*** ‐0.52*** ‐0.66*** ‐0.55*** ‐0.51*** (0.117) (0.141) (0.162) (0.145) (0.162)relative exam scores nearest school ‐0.30 ‐0.11 ‐0.25 ‐0.21 (0.260) (0.309) (0.252) (0.309)# of tracks offered at nearest school ‐0.01 ‐0.04 ‐0.00 ‐0.08 (0.052) (0.060) (0.055) (0.077)% upward mobility at nearest school ‐0.49 ‐0.57 0.09 ‐0.51 (0.602) (0.711) (0.556) (0.686)average SES index at nearest school ‐0.32*** ‐0.33** ‐0.33** ‐0.55*** (0.114) (0.135) (0.131) (0.167)% non‐western at nearest school ‐0.20 ‐0.19 ‐0.07 ‐0.05 (0.302) (0.332) (0.336) (0.457)% non‐western at nearest school x non‐western ‐0.10 0.18 ‐0.17 ‐0.75* (0.193) (0.207) (0.195) (0.407)Utrecht municipality dummy ‐0.12 0.14 ‐0.03 0.12 0.04 (0.194) (0.247) (0.274) (0.256) (0.299)The Hague municipality dummy ‐0.55*** ‐0.67*** ‐0.85*** ‐0.49*** ‐0.90*** (0.138) (0.183) (0.210) (0.187) (0.254)Rotterdam municipality dummy ‐0.51*** ‐0.53*** ‐0.72*** ‐0.33** ‐0.65*** (0.137) (0.162) (0.185) (0.168) (0.197)HAVO advice track dummy ‐0.85*** (0.106) HAVO/VWO advice track dummy ‐0.90*** (0.122) VWO advice track dummy ‐0.63*** (0.170) CITO test score ‐0.03*** (0.007)Non‐western x CITO test score 0.00* (0.001)Constant 2.36*** 4.84*** 5.58*** 7.62*** 5.10*** 19.03*** (0.082) (0.600) (0.829) (0.955) (0.794) (3.714) Observations 16,071 16,017 10,993 9,504 6,065 3,208R‐squared 0.01 0.05 0.07 0.13 0.09 0.10Adj. R‐squared 0.0100 0.0528 0.0708 0.127 0.0913 0.0949
Notes: Robust clustered standard errors (at neighbourhood level) in parentheses *** p<0.01, ** p<0.05, * p<0.1
![Page 122: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/122.jpg)
96
Table 4‐6: Comparisons of nearest (non‐chosen) school and the actual school, considering non‐nearest school choosers only (14247 individuals, 88.7 % of the data set)
Non‐nearest school choosing Dutch Sur./Ant. Turkish Moroccan Other Imm. Total mean mean mean mean mean meanVARIABLE (SD) (SD) (SD) (SD) (SD) (SD) NEAREST (NON‐CHOSEN) SCHOOL % non‐western at school 54.35 66.81 66.49 63.9 60.74 60.25 (29.18) (27.96) (29.27) (30.) (29.95) (29.75)school level average SES index ‐0.08 ‐0.69 ‐0.73 ‐0.52 ‐0.46 ‐0.38 (1.) (.86) (.85) (.86) (.95) (.97)relative exam school average 0 ‐0.1 ‐0.16 ‐0.16 ‐0.05 ‐0.07 (.28) (.3) (.31) (.32) (.3) (.3) school level average CITO intake 529.23 526.14 524.58 526.5 527.39 527.55 (9.17) (8.96) (9.31) (9.83) (9.69) (9.49)% upward mobility at school level 0.18 0.2 0.23 0.21 0.2 0.2 (.1) (.1) (.11) (.11) (.11) (.11)# of tracks offered at secondary 2.29 2.35 1.95 2.07 2.32 2.23 (1.54) (1.6) (1.75) (1.69) (1.56) (1.61)ACTUAL (CHOSEN) SCHOOL % non‐western at school 32.56 63.24 74.09 75.53 52.82 52.4 (20.64) (25.14) (23.23) (23.2) (28.12) (29.46)school level average SES index 0.56 ‐0.33 ‐0.6 ‐0.58 ‐0.04 0 (.86) (.91) (.82) (.77) (.96) (.99)relative exam school average 0.09 ‐0.03 ‐0.08 ‐0.11 0.03 0.02 (.26) (.26) (.28) (.27) (.3) (.29)school level average CITO intake 535.81 529.37 528.63 528.71 532.95 532.4 (8.89) (9.26) (8.95) (8.85) (9.71) (9.62)% upward mobility at school level 0.15 0.19 0.21 0.22 0.17 0.18 (.09) (.09) (.1) (.1) (.1) (.1) # of tracks offered at secondary 2.75 2.9 2.96 3.06 2.84 2.86 (1.18) (1.33) (1.43) (1.29) (1.32) (1.28)ACTUAL ‐ NEAREST mean mean mean mean mean meanVARIABLE (Pr{|T| > (Pr{|T| > (Pr{|T| > (Pr{|T| > (Pr{|T| > (Pr{|T| > % non‐western at school ‐21.79 ‐3.57 7.6 11.63 ‐7.92 ‐7.85 (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)school level average SES index 0.64 0.36 0.13 ‐0.06 0.42 0.38 (0.000) (0.000) (0.000) (0.002) (0.000) (0.000)relative exam school average 0.09 0.07 0.08 0.05 0.08 0.09 (0.000) (0.000) (0.000) (0.002) (0.000) (0.000)school level average CITO intake 6.58 3.23 4.05 2.21 5.56 4.85 (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)% upward mobility at school level ‐0.03 ‐0.01 ‐0.02 0.01 ‐0.03 ‐0.02 (0.000) (0.004) (0.000) (0.026) (0.000) (0.000)# of tracks offered at secondary 0.46 0.55 1.01 0.99 0.52 0.63 (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) N 5649 1959 1796 2198 2605 14247
Furthermore, a higher SES index of the nearest school reduces the average
distance difference, reflecting that more pupils attend the nearest school when this
school has a higher SES population. Compared with the vocational track as a reference
(VMBO), the distance difference is reduced in the case of pupils enrolled in academic
tracks (HAVO and VWO).30 While overall we found no major differences between
analyses of the absolute distance to school and of the distance difference between the
actual and the nearest school, in the latter case the magnitude seems more sensitive to
controls. As we found repeatedly in the different measurements we present in this
research, the quality of the nearest school as expressed by the relative exam score, does
30 Since the difference in distance must be ≥ 0, we also carried out a Tobit regression. However, we opted for the OLS regression because we found no major discrepancies with the Tobit regression, probably because only around 10 % of individuals choose the nearest school. OLS tables may be easier to interpret, since Tobit regressions require an extra calculation of marginal effects.
![Page 123: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/123.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
97
not demonstrate a significant effect on the distance difference. We added Appendix 4‐2,
to illustrate that the exam score at the proximity school is indeed only marginally
associated with the distance difference, in the case of all ethnic groups.
4.4.5 School choice and segregation
In this paragraph we take a closer look at links between school choice and
segregation. We have already seen that notably Turkish and Moroccan migrant students
travel less far to school. In Table 4‐6 we consider the subset of pupils who do not choose
the nearest school (14247 individuals, 88.7% of the total data set), and compare the
characteristics of their chosen school with the nearest school they did not choose.
Considerable dissimilarities appear between choice patterns of native Dutch pupils,
compared with pupils who have a migrant background. Native Dutch students, relative
to the not‐chosen nearest school, choose a school with lower percentages of migrant
pupils, a higher school average SES , a higher relative exam score, and a higher average
CITO score intake. In contrast, consider, for example students with a Moroccan
background: they choose a school with a higher percentage of pupils with a non‐
Western migrant background, about the same school SES index as the nearest school, a
slightly better relative exam score and more often a broader secondary school, offering
more tracks. We carried out a paired t‐test and found that the characteristics of the
chosen more distant school in all cases differed significantly from those of the nearest
school.
The different choice patterns between native Dutch and migrant pupils
apparently lead to more segregation, and may reveal segregation by choice among
migrant parents. Only in the case of native Dutch students we did find a school quality
induced movement.
We carried out regression analyses for these different choice patterns, and
analysed the difference between the nearest and the actual school for all choosers of
more distant schools, concerning four different indicators at school level. Note that
these four tables are based on different numbers of observations; in some cases, the
information on all school quality indicators was not listed in our data set. We carried out
tests to check for unobserved selection mechanisms, and found that coefficients on
![Page 124: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/124.jpg)
98
socio‐economic variables remained stable between the different subsamples. We
analysed the following school characteristics:
The school's mean exam score (Appendix 4‐3);
The school's percentage of upward mobility ("pupward") to a higher track
(Appendix 4‐4);
The average SES of the school population (Table 4‐7);
The school's percentage of pupils with a non‐western background (Table 4‐8).
We found the most significant coefficients for the difference in school SES (Table
4‐7) and school percentage of pupils with a non‐Western migrant background, as a
further Specification of school SES (Table 4‐8). Tables on relative exam score and upward
mobility have been included in the Appendices.
Looking at the difference in the average school SES between the nearest and the
actual school (Table 4‐7), the intercept in Specification (1) indicates that the reference
group of native Dutch gains from not choosing the nearest school. This outcome
reiterates our finding that native Dutch pupils tend to opt for a school with less migrant
pupils when their nearest school has many migrant pupils. Immigrant students gain less,
and, in fact, Moroccan students do not gain SES status at all. Neighbourhood SES has no
effect on the gain, but the percentage of non‐Western pupils at the nearest school
clearly increases the gain, by some 0.15 SES score (about one‐tenth of a standard
deviation) for a 10 percentage‐point increase. Weighted student funding in the
elementary school also reduces the SES gain between the nearest and the actual school
significantly. Finally, a recommendation by the primary school teacher for one of the
academic secondary tracks also increases the SES difference between the nearest and
the actual school significantly: the mean SES among pupils in academic tracks tends to
be higher than among pupils in vocational tracks. In The Hague and Rotterdam, the gain
is substantially higher than in the other two cities.
![Page 125: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/125.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
99
Table 4‐7: Regression results (OLS) for SES difference (i.e. average SES of chosen school minus average SES of nearest school), considering non‐nearest school choosers only (14197 individuals)
(1) (2) (3) (4) (5)VARIABLES SES diff SES diff SES diff SES diff SES diff Distance difference to nearest school 0.03*** 0.06*** 0.05*** 0.05*** 0.05*** (0.010) (0.013) (0.010) (0.010) (0.013) male 0.01 0.01 0.01 0.01 0.02 (0.013) (0.013) (0.013) (0.016) (0.022) weighted student funding ‐0.16*** ‐0.15*** ‐0.12*** ‐0.16*** ‐0.17*** (0.048) (0.033) (0.029) (0.040) (0.051) Surinamese/Antillean ‐0.20*** ‐0.26*** ‐0.18** ‐0.05 ‐0.06 (0.045) (0.041) (0.073) (0.073) (0.065) Turkish ‐0.41*** ‐0.52*** ‐0.41*** ‐0.25*** ‐0.26*** (0.082) (0.057) (0.083) (0.084) (0.082) Moroccan ‐0.59*** ‐0.58*** ‐0.46*** ‐0.31*** ‐0.28*** (0.091) (0.057) (0.088) (0.091) (0.090) Other Immigrant Background ‐0.16*** ‐0.20*** ‐0.13** ‐0.06 ‐0.04 (0.050) (0.034) (0.050) (0.054) (0.058) one‐parent household ‐0.12** ‐0.12*** ‐0.15*** ‐0.07* ‐0.09* (0.045) (0.030) (0.028) (0.038) (0.048) neighbourhood SES index ‐0.02 0.08* 0.05 0.05 (0.054) (0.047) (0.053) (0.042) Non‐western x SES 0.00 0.01 0.02 0.04 (0.034) (0.031) (0.034) (0.035) urbanicity 0.36*** 0.33*** 0.22*** 0.17* (0.110) (0.075) (0.075) (0.087) # of relevant schools within 5 km ‐0.00 ‐0.00 0.00 ‐0.00 (0.005) (0.004) (0.005) (0.005) distance to nearest relevant school (km) 0.29** 0.17*** 0.15*** 0.13** (0.112) (0.057) (0.056) (0.053) # of tracks offered at nearest school ‐0.02 ‐0.02 ‐0.03 (0.038) (0.039) (0.043) % non‐western at nearest school 1.55*** 1.43*** 1.60*** (0.273) (0.296) (0.296) % non‐western at nearest school x non‐western ‐0.22* ‐0.22 ‐0.18 (0.124) (0.137) (0.226) Utrecht municipality dummy 0.03 0.11 0.10 0.08 (0.166) (0.116) (0.133) (0.127) The Hague municipality dummy 0.66*** 0.56*** 0.70*** 0.73*** (0.166) (0.112) (0.124) (0.125) Rotterdam municipality dummy 0.30** 0.38*** 0.41*** 0.49*** (0.134) (0.134) (0.142) (0.154) HAVO advice track dummy 0.25*** (0.057)HAVO/VWO advice track dummy 0.36*** (0.066)VWO advice track dummy 0.51*** (0.079)CITO test score 0.02*** (0.003) Non‐western x CITO test score ‐0.00 (0.000) Constant 0.56*** ‐1.25*** ‐1.86*** ‐1.70*** ‐11.12*** (0.090) (0.474) (0.409) (0.392) (1.525) Observations 14,197 14,147 14,147 7,912 4,145 R‐squared 0.08 0.18 0.35 0.37 0.43 Adj. R‐Squared 0.0747 0.181 0.344 0.371 0.427
Notes: Robust clustered standard errors (at neighbourhood level) in parentheses *** p<0.01, ** p<0.05, * p<0.1
![Page 126: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/126.jpg)
100
Table 4‐8: Regression results (OLS) for the difference in the percentage of pupils with a non‐Western migrant background (i.e. percentage at the actual school minus the percentage at the nearest school), considering non‐nearest school choosers only (14197 individuals)
(1) (2) (3) (4) (5)VARIABLES % NW diff % NW diff % NW diff % NW diff % NW diff Distance difference to nearest school ‐1.70*** ‐1.99*** ‐1.30*** ‐1.77*** ‐1.03** (0.372) (0.374) (0.289) (0.328) (0.417) male 0.05 0.10 0.39 0.73 0.14 (0.554) (0.529) (0.512) (0.670) (0.914) weighted student funding 8.90*** 8.62*** 9.26*** 10.06*** 7.11** (2.119) (1.882) (1.567) (1.836) (2.998) Surinamese/Antillean 14.19*** 15.06*** 17.64*** 14.83*** 7.45*** (1.837) (1.603) (1.763) (1.835) (2.521) Turkish 24.26*** 26.39*** 26.87*** 22.34*** 17.81*** (3.309) (2.914) (2.698) (2.059) (3.307) Moroccan 27.48*** 28.51*** 27.18*** 22.23*** 16.84*** (3.513) (2.672) (2.524) (2.139) (3.325) Other Immigrant Background 10.44*** 10.66*** 11.15*** 8.94*** 3.89* (1.537) (1.238) (1.288) (1.321) (2.103) one‐parent household 5.38*** 5.20*** 6.67*** 4.74*** 4.61** (1.489) (1.281) (1.103) (1.355) (2.128) neighbourhood SES index 1.27 ‐4.49*** ‐3.13 ‐3.10* (1.713) (1.590) (1.954) (1.871) Non‐western x SES 0.88 0.94 1.52 1.06 (1.450) (1.513) (1.546) (1.656) urbanicity ‐3.29 1.74 4.81 5.68 (4.322) (3.123) (3.523) (3.760) # of relevant schools within 5 km 0.08 0.16 ‐0.01 0.16 (0.264) (0.213) (0.269) (0.235) distance to nearest relevant school (km) ‐8.44* ‐5.05 ‐5.60 ‐3.50 (5.072) (3.818) (3.518) (3.267) # of tracks offered at nearest school ‐3.48** ‐3.89*** ‐4.08*** (1.349) (1.451) (1.519) SES index at nearest school 21.34*** 22.66*** 24.76*** (2.282) (2.652) (2.631) SES index at nearest school x non‐western 1.14 ‐1.18 ‐0.79 (2.062) (1.865) (2.042) Utrecht municipality dummy ‐2.21 ‐15.52** ‐17.86*** ‐19.07*** (8.542) (6.666) (6.415) (5.864) The Hague municipality dummy ‐10.35* ‐8.35* ‐9.30* ‐9.52** (6.164) (4.341) (4.885) (4.524) Rotterdam municipality dummy 0.88 11.92** 11.33** 16.87*** (6.120) (5.076) (5.534) (5.781) HAVO advice track dummy ‐10.91*** (2.381)HAVO/VWO advice track dummy ‐13.43*** (2.948)VWO advice track dummy ‐16.69*** (2.729)CITO test score ‐0.65*** (0.117) Non‐western x CITO test score 0.01*** (0.005) Constant ‐17.57*** 3.92 ‐7.36 ‐4.42 327.73*** (2.775) (17.828) (14.916) (16.965) (63.260) Observations 14,197 14,147 14,147 7,912 4,145 R‐squared 0.14 0.18 0.38 0.41 0.47 Adj. R‐Squared 0.137 0.174 0.375 0.408 0.465
Notes: Robust clustered standard errors (at neighbourhood level) in parentheses *** p<0.01, ** p<0.05, * p<0.1
![Page 127: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/127.jpg)
101
In Table 4‐8 we chose the difference between the percentages of pupils with a
non‐western immigrant background of the actual minus the nearest school as the
outcome variable. As indicated by the intercept in Specification (1), native Dutch
reference students, on average, travel to a school with lower percentages of migrant
pupils. Turkish and Moroccan students travel to schools with higher percentages of non‐
Western immigrants, whereas the difference for the other immigrant groups is smaller.
In the case of Moroccan pupils, the percentage non‐western pupils increases at the
actual school by 9.9 percentage points (‐17.57 + 27.48), minus 1 percentage point per
km increase in distance difference (Specification 5). Since distances to school are
relatively small in the Netherlands (Table 4‐1), we may assume that the different
patterns found in school choice between native Dutch and pupils of Moroccan descent
are indeed the effect of different ethnic preferences. We found comparable patterns of
choice between Turkish and Moroccan students; also Surinamese/Antillean pupils and
those with another migrant background, on average, prefer a school with a higher
percentage of migrants, but to a lesser degree than pupils of Turkish and Moroccan
descent. Introducing the CITO test score (at the cost of substantially lower numbers of
observations) is also associated with a lower percentage of migrant pupils at the chosen
school. These findings confirm our earlier outcomes: migrant pupils and their parents
appear to choose further segregation.
4.4.6 Benefits of school choice?
Upward mobility to a higher track may be a particularly important extra
opportunity for migrant pupils. They are relatively more often the first one in their
family with the option to enrol in academic secondary tracks, and may need extra time
to arrive at this level and later on in higher education. At the start of secondary
education, far higher percentages of migrant pupils enter the lower secondary
vocational track: e.g. 76 % of Moroccan youths, against 42 % of native Dutch youngsters
(Table 4‐1). In fact, this is the width of the educational achievement gap, which indeed
could be reduced by offering migrant students extra opportunities to improve their track
level while in secondary school. The Dutch educational system has two trajectories for
upward mobility: 1) after finishing one track with a diploma, pupils can re‐enrol in a
higher track and obtain a second diploma at this higher level. 2) Pupils can also move up
![Page 128: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/128.jpg)
102
between grades, and finish with a diploma on a higher level than their level of entrance
in Year 1. In Table 4‐9 we consider this second trajectory for upward mobility between
grades. We included only pupils who started on the lower vocational track in Year 1,
since notably this lowest level leaves open the possibility of opportunities for upward
transfer to a higher track. We counted only individuals who chose the more distant
school, in order to be able to assume active choice. 4343 individuals in our data set
started in the vocational track when they entered secondary school. This number was
further reduced to 2730 individuals when we introduced characteristics of the nearest
school (which are not in all cases listed in the data set). Other socio‐economic variables,
however, remained stable in comparison with the total number of individuals in our data
set.
The coefficients in Table 4‐9 show a small, but significant, negative correlation
between a pupils' upward mobility and the distance difference to the nearest school: the
predicted probability for a student to have moved up to a higher track in Year 3 reduces
by 0.01 for every kilometre increase in the distance difference between the actual and
the nearest school. Notably migrant pupils seem to lose opportunities to move up to a
higher track, when they travel further to school. Furthermore, the neighbourhood SES
index is positively associated with upward mobility, indicating that pupils living in a more
affluent residential area also have a higher chance to move up while in secondary
education. The positive neighbourhood effect is reduced for immigrant pupils.
![Page 129: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/129.jpg)
103
Table 4‐9: OLS regressions of individual upward mobility at the entrance of Year 3, considering only pupils who started at the lower vocational secondary level in Year 1 (4343 individuals); non‐nearest school choosers only
Non‐nearest choosing VMBO advice students (1) (2) (3)VARIABLES level‐up in Year 3 level‐up in Year 3 level‐up in Year 3 Distance difference to nearest school ‐0.01*** ‐0.01*** ‐0.01*** (0.002) (0.003) (0.003) male ‐0.03*** ‐0.03*** ‐0.04*** (0.010) (0.010) (0.013) weighted student funding ‐0.06*** ‐0.06*** ‐0.06*** (0.014) (0.014) (0.018) Surinamese/Antillean ‐0.06*** ‐0.05*** ‐0.05** (0.017) (0.017) (0.020) Turkish ‐0.06*** ‐0.04** ‐0.05* (0.019) (0.018) (0.024) Moroccan ‐0.04** ‐0.03 ‐0.03 (0.019) (0.018) (0.024) Other Immigrant Background 0.01 0.02 0.02 (0.018) (0.018) (0.023) one‐parent household ‐0.03* ‐0.02 ‐0.01 (0.015) (0.014) (0.016) neighbourhood SES index 0.05*** 0.05*** (0.010) (0.013) Non‐western x SES ‐0.02** ‐0.03*** (0.010) (0.012) urbanicity ‐0.02 ‐0.01 (0.017) (0.023) # of relevant schools within 5 km 0.00* 0.00** (0.001) (0.001) distance to nearest relevant school (km) ‐0.03** ‐0.04** (0.013) (0.017) relative exam scores nearest school 0.02 (0.033) # of tracks offered at nearest school 0.00 (0.006) % upward mobility at nearest school ‐0.05 (0.086) average SES index at nearest school 0.01 (0.020) % non‐western at nearest school ‐0.00 (0.034) Utrecht municipality dummy ‐0.03 0.00 (0.025) (0.035) The Hague municipality dummy ‐0.04** ‐0.00 (0.019) (0.022) Rotterdam municipality dummy ‐0.04** ‐0.01 (0.019) (0.022) Constant 0.22*** 0.29*** 0.25** (0.020) (0.075) (0.098) Observations 4,343 4,325 2,730 R‐squared 0.03 0.04 0.05 Adj. R‐Squared 0.0253 0.0401 0.0462
Notes: Robust clustered standard errors (at neighbourhood level) in parentheses *** p<0.01, ** p<0.05, * p<0.1
![Page 130: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/130.jpg)
104
4.5 Conclusions and discussion
In the current research we explored 1) how the distance travelled to school is
related to ethnicity and other socio‐economic variables on the individual‐,
neighbourhood‐ and school level; 2) the differences between the group that chooses the
nearest school, and the group that prefers another school; 3) the effects of free school
choice on ethnic segregation; and 4) upward mobility to a higher track as a possible
benefit of school choice.
4.5.1 Who travels further to school?
All over, migrant pupils travel less distance to school than native Dutch pupils,
except for native Dutch pupils in affluent residential areas. Average distance to school is
obviously influenced by population density and the corresponding larger number of
schools in densely populated areas. The number of schools to choose from in the four
Dutch major cities is, for example, 15 schools within a radius of 5 km in the case of
native Dutch students, to 21 schools in the case of students of Turkish descent (Table
4‐1). Since more students with a migrant background at present live in less affluent,
more densely‐populated neighbourhoods, this may partly explain why they travel, on
average, less far to school. However, urbanicity, according to our findings, cannot solely
explain the lower mean distance travelled to school by migrant pupils.
Students of Dutch origin travel the largest mean absolute distance, 3.49 km.
However, the absolute distance to school is inversely correlated with residential SES: in
the poorest residential areas, native Dutch youths travel further than Turkish and
Moroccan youths, while in contrast, the average distance to school is lower among
native Dutch students in affluent areas. This may also explain our findings that the
elementary teacher's recommendation for the academic secondary track is associated
with a lower distance to school: substantially higher percentages of native Dutch
students obtain this advice, and they more often live in an affluent area where the
preferred school may be nearby.
4.5.2 Differential sorting, mobility increases segregation
The literature predicts strong effects on ability sorting of extended school choice
(e.g. Cullen et al., 2000). In the context of unlimited choice in the Netherlands, however,
we found, somewhat to our surprise, virtually no systematic differences between the
![Page 131: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/131.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
105
group of choosers for the nearest school and those who chose a school further from
their home address. We did find, however, differential sorting when we compared the
characteristics of the nearest school and the selected school, that was not induced by
pupils' abilities, but related to the pupils' ethnicity: native Dutch pupils seem to weigh
other aspects when selectively choosing a school at a further distance than is the case
among migrant pupils. When migrant students prefer a school further outside their
residential area, they tend to choose a school with even higher percentages of migrant
pupils than the neighbourhood school. Native Dutch pupils, on the other hand, tend to
make a quality‐driven choice for a school with a higher mean exam score and a higher
mean CITO intake. They prefer, furthermore, a school with a higher SES population and a
lower percentage of migrant students than the nearest school. Remarkably, migrant
students, in contrast, more often choose schools with higher percentages of migrant
students, than the often already high percentages at the nearest school, while at the
same time, they show only a marginal preference for higher school quality. In fact, we
may have revealed here the segregation of migrant pupils by choice. Our findings bring
to mind research by Bunar in Sweden (2010), who learned from interviews with migrant
students in Stockholm and Malmö how important the sense of belonging and
recognition within the pupils' own ethnic group seems to be.
4.5.3 Upward mobility, an important opportunity for migrant students
Finally, although the effect we found is fairly small, we still consider as important
the outcome that the slightly higher odds of migrant students to move up to a higher
track when they attend a school closer to their home. In the domain of academic
research into the achievement gap between the children of low SES and high SES
parents, the question whether disadvantaged pupils should be bussed to better schools,
or better schools should be brought to poorer neighbourhoods, seems centre stage. Our
findings rather support the second strategy.
4.5.4 Policy implications and further research
We plan to follow up on the current research, and investigate more in‐depth why
migrant parents and pupils seem to prefer a school with high percentages of migrant
pupils. For the moment we reflect on two possible explanations: Over the years, schools
with high percentages of migrant students (up to 90%), have largely invested in, for
![Page 132: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/132.jpg)
106
example, the development of tailor‐made strategies to support first‐time academic
learners, effective cooperation with parents, and the encouragement of upward
mobility. There may be, therefore, reasons for migrant parent to choose a school with
large numbers of migrant children; 2) we cannot exclude, however, that migrant pupils
tend to wish to avoid stigmatization in a school with predominantly native Dutch pupils,
and prefer a school where they are the absolute majority. In the Dutch context, as we
described earlier, it is important to keep in mind that no financial concerns limit school
choice in any way; neither do schools with lower SES pupils, in general, have less
qualified teachers, or struggle with a high turnover of teachers. This may enhance
opportunities to identify other motives behind school choice, for example, choices
related to ethnicity, as mentioned above.
Furthermore, we reflected on our finding that the average school exam scores
only marginally drive school choice, even though this information is freely available on
the internet (Inspectorate of Education). Since the average school exam score is strongly
correlated to the average level of schooling of parents, it is complicated for parents to
identify the school's added value which best fits their child. Supported by our finding
that the school's upward mobility does not drive school choice whereas this could be
expectedly an important element in the closing of the achievement gap– we suggest
that the Inspectorate of Education publishes on the Internet both exam results and the
upward mobility of all schools, specified for the separate ethnic groups. More in general,
informing migrant parents is, according to, for example Douglas Archbald (2004), a
requirement for free school choice as a Liberation Model.
We contemplated on the overall marginal associations we found with school
quality indicators. It is imaginable that almost a century of free school choice, the
absence of tuition fees, and the publication of the Inspectorate's assessment reports of
all schools on the Internet have all together resulted in a transparent, fairly
homogeneous school market. Out of about 650 secondary schools nationally, around 30
schools are under intensified supervision of the Inspectorate because of severe
underperformance. Currently only two of these 30 schools are located in the major four
cities (www.onderwijsinspectie.nl). Moreover, underperforming schools typically
improve within one or two years, under intensified supervision of the Inspectorate.
Taking this thought further, in the Netherlands, perhaps less so than in other countries,
![Page 133: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/133.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
107
choosing one school or another may not result in serious divergences in opportunities
for pupils.
Finally, another aspect of school choice that may inspire further research is the
fact that pupils do have an important say themselves in the choice of a secondary school
in the Netherlands. They may, for example, want to go to the same secondary school as
their classmates in elementary education, they may feel inspired by the extra
programmes a secondary school offers, or they may feel attracted to a new well‐
equipped school building. Free school choice and the absence of tuition fees make it
possible for pupils to follow such preferences. How pupil preferences influence patterns
of choice and school results would be an interesting further research question.
We look forward to be able to use the data on final exams in the coming years, as
the BRON database is further developing. We acknowledge that assembling data sets for
the purpose of academic research is time‐consuming, and that the government institute
that collects these data (DUO) does currently not have the extra personnel to carry out
this task. This has deepened our gratitude towards Cees Vermeulen, Erik Smits and Rob
Kerstens, who have already invested their time and support. We do hope that our work
has illustrated the high quality and the possibilities of the BRON data, and its potential to
improve the quality of education at large, through a better understanding of these
complex processes.
4.6 Appendices
In the general overview depicted in Appendix 4‐1, we considered the higher
aggregate level of non‐Western background (includes the vast majority of pupils of
Turkish or Moroccan descent) versus Dutch background. The following maps give a first
impression of the association between the ethnic composition of an area, and the
distance travelled to school. For example, the southern part of the city centre in
Amsterdam has a large share of native Dutch inhabitants (only 0‐20% inhabitants with a
migrant background), and the average distance travelled to school in this area is lower
(lightest shade of blue) than in other districts in Amsterdam.
Some differences between the four cities can be observed. In Rotterdam and The
Hague, neighbourhoods in the city centre more often have a population predominantly
consisting of people with a non‐Western background than is the case for Amsterdam
![Page 134: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/134.jpg)
108
and Utrecht. This relates to differences in city structure: in Amsterdam and Utrecht, top‐
quality residential areas are located in the heart of the city.
Interestingly, in Amsterdam, more so than in the three other cities, schools tend
to be segregated more than neighbourhoods: 20 out of 84 secondary schools in
Amsterdam enrolled between 80 and 100 % pupils with a non‐Western migrant
background, whereas no neighbourhood in Amsterdam has this high percentage of
inhabitants with a migrant background; the degree of segregation of these 20 schools,
exceeds the degree of segregation of the surrounding neighbourhood. While in
Rotterdam the variety in ethnic composition of neighbourhoods reflects the ethnic
composition of schools more closely, also in this city 21 out of 113 schools enrolled
between 80‐ 100% of pupils of non‐ Western migrant descent. In comparison, The Hague
and Utrecht seem to be less segregated, although The Hague is the only city with 6
neighbourhoods in the city‐centre where between 80‐100% of inhabitants have a non‐
Western immigrant background.
We also depicted the schools outside the city boundaries which were attended
by pupils living within the city. There are remarkable differences in the percentage of
pupils attending a school outside the city boundary among the four cities:
Amsterdam: 230/5153=4.4%
Rotterdam: 964/4853=19.8%
The Hague: 791/4000=19.7%
Utrecht: 639/2065=30.9%
In the current chapter we did not investigate further the causes for these
differences.
4.6.1 Explanation of variables
Average SES index school: the average SES of (the parents of) children at a
particular school.
CITO‐score: final test in elementary education; scores range between 500‐ 550,
the national average is 535.
HAVO, HAVO/VWO and VWO advice track dummy: dummy‐coded variable
relative to VMBO (lowest vocational track).
![Page 135: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/135.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
109
Individual upward mobility: pupil who moves up to a higher secondary track.
Municipality dummy‐coded fixed effects of Utrecht, Rotterdam and The Hague,
relative to Amsterdam.
Nearest relevant school: the nearest school to the pupils' residence that offers
the pupils' chosen track.
Non‐Western immigrant: The following definitions by Statistics Netherlands (CBS,
the National Statistics Office, www.cbs.nl) have been used: Western immigrant:
someone originating from a country in Europe (exclusive of Turkey), North
America, Oceania, Indonesia or Japan. Non‐western immigrant: someone
originating from Africa (including Morocco), South America, Asia (exclusive of
Indonesia and Japan) or Turkey.
Other immigrant background: a wide range of nationalities, including highly
educated workers, refugees and people seeking asylum. The variance in SES is
large, accordingly.
Outside municipality border: pupils, who live within city boundaries, but attend a
school outside the city.
Relative exam score: the performance per track level, measured by the mean
exit‐exam score, compared with the mean exit‐exam score of all other schools in
the four major cities offering the same track level.
Relative neighbourhood SES index: Neighbourhood SES score relative to the
average in the four major cities; SES scores have been centred on zero, ‐4 (poor)
to +4 (affluent).
Secondary school advice score: advice by the elementary teacher about which
track to follow in secondary education.
Tracks offered at secondary school: the number of tracks (1 ‐ 4 main tracks) may
differ per school. Comprehensive schools offer tracks on the vocational and
academic levels. Schools may also offer either vocational or academic tracks;
schools may be diversified further and offer only one academic track.
Upward mobility at school level: the percentage of pupils at school level who
move up to a higher secondary track.
![Page 136: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/136.jpg)
110
Urbanicity: a measure for human activity in an area, based on the number of
addresses per km².
Weighted student funding: in primary schools children qualify for extra funding
depending on the educational attainment level of the parents; the pupil weight
for funding is 0 for better‐educated parents (means only standard funding);
children of less well‐educated parents may receive 0.3 to 1.2 extra funding,
additional to standard funding.
Appendix 4‐1: Maps of Amsterdam, Rotterdam, Utrecht and The Hague.
In shades of blue, the average distance travelled to secondary school on the
neighbourhood level; in shades of red, the percentage of non‐Western immigrants at
neighbourhood level. In parentheses, the number of neighbourhoods sharing the same
percentage of non‐Western immigrants, and the number of neighbourhoods sharing the
same average distance travelled to school, respectively. The cities are not depicted to
scale, but have been formatted to fit in this text. The ranking order of size is Amsterdam
(largest), Rotterdam, The Hague, Utrecht.
![Page 137: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/137.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
111
![Page 138: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/138.jpg)
112
![Page 139: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/139.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
113
![Page 140: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/140.jpg)
114
![Page 141: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/141.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
115
Appendix 4‐2: LOWESS regression (Locally Weighted Scatterplot Smoothing; bandwidth = .6) and linear regression (OLS) of the extra distance travelled to the preferred school (as compared with the nearest school) and school's mean final exam score (national standardized exam score)
![Page 142: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/142.jpg)
116
Appendix 4‐3: Regression results (OLS) for difference in the mean relative exam score at school level (i.e. the actual school exam score minus exam score at the nearest school), considering non‐nearest school choosers only
(1) (2) (3) (4) (5)VARIABLES exam diff exam diff exam diff exam diff exam diff Distance difference to nearest school 0.00 0.01* 0.01 0.00 0.00 (0.005) (0.005) (0.005) (0.005) (0.006) male ‐0.01 ‐0.01 ‐0.01 ‐0.02* 0.00 (0.009) (0.009) (0.008) (0.011) (0.014) weighted student funding ‐0.01 ‐0.04* ‐0.03 ‐0.05** ‐0.04 (0.032) (0.022) (0.021) (0.026) (0.040) Surinamese/Antillean ‐0.05 ‐0.08*** ‐0.01 ‐0.04 ‐0.00 (0.033) (0.027) (0.048) (0.042) (0.050) Turkish ‐0.04 ‐0.13*** ‐0.04 ‐0.07 ‐0.08 (0.060) (0.043) (0.065) (0.048) (0.064) Moroccan ‐0.09 ‐0.16*** ‐0.07 ‐0.08* ‐0.06 (0.058) (0.042) (0.067) (0.045) (0.063) Other Immigrant Background ‐0.05* ‐0.06*** ‐0.02 ‐0.03 ‐0.01 (0.027) (0.024) (0.033) (0.031) (0.040) 2nd generation non‐Western Immigrant 0.03 0.04** 0.05*** 0.03 0.02 (0.019) (0.018) (0.017) (0.020) (0.029) one‐parent household ‐0.04 ‐0.05** ‐0.06*** ‐0.03 ‐0.05 (0.028) (0.020) (0.021) (0.020) (0.033) neighbourhood SES index ‐0.06** ‐0.02 ‐0.02 ‐0.01 (0.021) (0.026) (0.025) (0.027) Non‐western x SES ‐0.00 ‐0.01 ‐0.01 0.01 (0.018) (0.018) (0.020) (0.023) urbanicity ‐0.05 ‐0.05 ‐0.03 ‐0.05 (0.052) (0.052) (0.053) (0.062) # of relevant schools within 5 km 0.00* 0.01* 0.00 0.00 (0.003) (0.003) (0.003) (0.003) distance to nearest relevant school (km) 0.01 ‐0.02 ‐0.03 ‐0.03 (0.040) (0.036) (0.034) (0.036) # of tracks offered at nearest school 0.02 0.00 0.00 (0.022) (0.020) (0.027) average SES index at nearest school ‐0.02 ‐0.07 ‐0.06 (0.059) (0.061) (0.071) % non‐western at nearest school 0.40** 0.33** 0.35* (0.159) (0.164) (0.193) % non‐western at nearest school x non‐western ‐0.17** ‐0.09 0.00 (0.073) (0.063) (0.115) Utrecht municipality dummy 0.28*** 0.32*** 0.29*** 0.27*** (0.094) (0.083) (0.082) (0.082) The Hague municipality dummy 0.28*** 0.27*** 0.27*** 0.20*** (0.055) (0.056) (0.059) (0.060) Rotterdam municipality dummy 0.11* 0.13* 0.12* 0.10 (0.067) (0.065) (0.077) HAVO advice track dummy 0.01 (0.032)HAVO/VWO advice track dummy 0.02 (0.034)VWO advice track dummy 0.19*** (0.041)CITO test score 0.00** (0.002) Non‐western x CITO test score ‐0.00 (0.064) (0.000) Constant 0.11*** 0.08 ‐0.18 ‐0.15 ‐2.26** (0.030) (0.236) (0.285) (0.289) (0.985) Observations 8,503 8,471 8,471 4,720 2,524 R‐squared 0.01 0.09 0.16 0.18 0.17 Adj. R‐Squared 0.00420 0.0924 0.153 0.173 0.158
Notes: Robust clustered standard errors (at neighbourhood level) in parentheses *** p<0.01, ** p<0.05, * p<0.1
![Page 143: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/143.jpg)
4 Patterns in secondary school selection in the context of unlimited choice
117
Appendix 4‐4: Regression results (OLS) for difference in school percentage of upward mobility to a higher track (i.e. upward mobility at the actual school minus upward mobility at the nearest school), considering non‐nearest school choosers only
(1) (2) (3) (4) (5)VARIABLES pupward diff pupward diff pupward diff pupward diff pupward diff Distance difference to nearest school 0.16 0.28 0.36* 0.11 0.27 (0.202) (0.207) (0.184) (0.179) (0.221)male 0.27 0.20 0.27 0.20 ‐0.11 (0.326) (0.316) (0.296) (0.358) (0.459)weighted student funding ‐0.78 0.37 0.33 ‐0.60 ‐2.39 (0.886) (0.777) (0.736) (0.917) (1.603)Surinamese/Antillean 1.34 2.81** 1.95 ‐0.49 ‐0.80 (1.186) (1.083) (1.940) (1.750) (1.925)Turkish 1.10 3.19** 2.06 0.14 ‐0.63 (1.937) (1.538) (2.269) (1.738) (2.223)Moroccan 4.07* 6.20*** 4.82** 2.62 1.64 (2.091) (1.571) (2.318) (1.625) (2.067)Other Immigrant Background 0.27 1.15 0.56 ‐0.12 ‐2.11 (0.917) (0.843) (1.189) (1.178) (1.321)2nd generation non‐Western Immigrant 0.40 0.32 0.14 0.02 1.29 (0.704) (0.643) (0.644) (0.828) (1.113)one‐parent household 1.36* 2.11*** 2.44*** 0.33 0.77 (0.735) (0.622) (0.646) (0.628) (0.870)neighbourhood SES index 1.98*** 0.42 1.18 1.47* (0.699) (0.834) (0.829) (0.868)Non‐western x SES ‐0.57 ‐0.33 ‐0.43 ‐0.85 (0.627) (0.639) (0.843) (1.022)urbanicity 0.60 0.80 0.61 0.48 (2.023) (2.185) (2.260) (2.376)# of relevant schools within 5 km 0.00 0.01 ‐0.06 ‐0.09 (0.108) (0.106) (0.118) (0.128)distance to nearest relevant school (km) ‐0.89 ‐0.15 ‐0.14 ‐0.64 (1.976) (1.743) (1.622) (1.613)# of tracks offered at nearest school ‐1.34 ‐1.77** ‐1.27 (0.889) (0.823) (1.060)average SES index at nearest school 2.68 2.21 1.84 (1.962) (2.274) (2.419)% non‐western at nearest school ‐6.55 ‐7.40 ‐12.29* (5.234) (6.366) (7.236)% non‐western at nearest school x non‐western 2.72 5.44*** 8.00* (2.646) (1.974) (4.692)Utrecht municipality dummy ‐2.50 ‐4.16 ‐7.06** ‐6.55** (4.066) (3.373) (3.342) (2.918)The Hague municipality dummy 0.10 0.20 ‐1.24 1.57 (2.178) (2.166) (2.236) (2.630)Rotterdam municipality dummy ‐1.11 ‐0.29 ‐0.59 ‐1.35 (2.415) (2.459) (2.570) (2.787)HAVO advice track dummy ‐3.16** (1.272) HAVO/VWO advice track dummy ‐3.54*** (1.334) VWO advice track dummy ‐8.34*** (1.422) CITO test score ‐0.31*** (0.104)Non‐western x CITO test score ‐0.00 (0.006)Constant ‐3.45*** ‐6.16 0.45 7.96 170.51*** (1.309) (9.372) (10.264) (10.023) (54.278) Observations 9,025 8,997 8,997 5,046 2,709R‐squared 0.01 0.04 0.10 0.14 0.18Adj. R‐Squared 0.00967 0.0396 0.0957 0.134 0.170
Notes: Robust clustered standard errors (at neighbourhood level) in parentheses *** p<0.01, ** p<0.05, * p<0.1
![Page 144: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/144.jpg)
![Page 145: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/145.jpg)
119
5 A tailor made transfer of scientific
knowledge to school practice
But despite the fact that natural knowledge is divine, its
practitioners cannot be called prophets. For other men may
discern and embrace what they teach with as much certainty and
entitlement as they do themselves. They do not just accept it on
faith.
Baruch de Spinoza, 1670
5.1 Introduction
The importance of scientific research for school practice is generally
acknowledged by teachers and school leaders, educational researchers, and policy
makers. Educational policies at the government level, furthermore, seem increasingly
supported by academic research (e.g. the No Child Left Behind Act, U.S. Congress, 2001).
So far, however, the effects of research findings on school outcomes seem disappointing
(Slavin, 2002). The academic literature in the domain of the utilization of knowledge in
the practice of schools reports very limited transfer of knowledge to schools.
This chapter is based on: Van Welie, E.A.A.M., Borghans, L. and De Wolf, I. (2013). Using research outcomes in school practice. Working paper.
![Page 146: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/146.jpg)
120
Ideally, the process to connect school questions with existing research outcomes
should include 1) the articulation of school questions; 2) the search for a match with
research findings; 3) a knowledge based intervention in school processes; and 4) the
analyses of the effects on school outcomes.
The aim of this paper was to investigate what obstacles hinder the process of
knowledge transfer, and how the application of scientific knowledge in schools could be
enhanced. So far the limited use of scientific knowledge in schools appears to be a
stubborn problem. We therefore chose to carry out a detailed study in six secondary
schools
We carried out in‐depth semi‐structured interviews with Principals to identify
and articulate school questions. Next, we carried out a literature review in order to link
these questions to state‐of‐the‐art research publications. Together with the six school
leaders, we concluded this explorative study with an evaluation, again by means of a
structured personal interview, of the applicability in school practice of a tailor‐made
selection of research publications.
In 2011 two reports were published in the Netherlands on the importance of an
enhanced transfer of knowledge from educational research to an effective utilization in
schools (Education Council of the Netherlands, 2011; Commissie Nationaal Plan
Toekomst Onderwijswetenschappen [Committee for the National Plan for the
Educational Sciences], 2011). Inspired by these reports, the present study sets out to
investigate what the nature of school questions is, and whether these questions can be
linked to relevant, high‐quality, existing research.
We carried out our research in secondary schools with a diverse student body,
including at least 30% of pupils with a migrant background. We focused on themes
related to the quality of school outcomes and equal opportunities for all pupils. We
made this choice for diverse schools because the attempt to understand the stubborn
and persistent achievement gap between migrant pupils and native‐born youths, and,
more in general, between low SES (socio‐economic background) children and more
affluent youth, is an ambitious policy goal both in Europe and in the U.S. A wealth of
research outcomes is published yearly on the achievement gap in education, and it may
be assumed that researchers in the field of education implicitly expect that schools could
![Page 147: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/147.jpg)
5 A tailor made transfer of scientific knowledge to school practice
121
benefit from the results of their work, when these results actually would be translated
to school strategies.
Previous research is described in Section 5.2. We describe our methodology in
Section 5.3; the six schools are portrayed in Section 5.4; the findings are listed in Section
5.5 and the conclusions in Section 5.6. The findings support our assumption that schools
need to be understood in their specific context, and that a timely match between
question and answer seems a prerequisite. The outcomes show that school leaders do
indeed have a demand for a knowledge base for their strategic themes; the majority of
queries could effectively be linked to research publications. In five out of six schools, the
selected scientific insights have been used to guide new school strategies, support
existing programmes, or inform discussions with the school board and local (political)
governors. Based on this project, school leaders recommend the formation of an
intermediate function between academe and secondary schools.
5.2 Previous research
Different kinds of knowledge
According to McIntyre (2006), educational research and school practice are "two
sharply contrasting kinds of knowledge…that are at the opposite ends of the spectrum":
researchers abstract ideas from complex realities, while teachers simultaneously work in
many dimensions: the concrete setting of their class, but also the combined levels of
school, parents, local stakeholders, the curriculum, and the many unpredictable daily
events in a large school. McIntyre (ibid.) suggests that the transfer of knowledge vice
versa between these two different worlds might benefit from the work of researchers
who generate general research‐based suggestions for practice in their publications,
discuss possible implications of their research with teachers, and cooperate in
monitoring the effects on outcomes of knowledge‐based interventions.
Landry, Amara and Lamari (2001), instead of comparing the different categories
of knowledge in academe and in schools, focus on the question whether academe raises
barriers, through its culture and organization, and notably in terms of additional
transition costs, against the dissemination of the results of scientific research to
classroom practice. To start with, the authors acknowledge that dissemination efforts at
best play a minor role in the evaluation of the quality of research; this expectedly does
![Page 148: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/148.jpg)
122
not stimulate the transfer of research to classrooms. Based on their research, Landry et
al. (ibid.) describe a list of predictions on how the transmission of knowledge might be
improved. Several of these predictions seem especially relevant for the present
research: 1) educational research projects should be customized to applicability in
practice, and correspond with the needs of teachers; 2) explanatory variables should
reflect the user's context, thereby adding to the credibility of research; 3) the utilization
of research possibly may increase when results are pertinent and reach schools at the
right time, coinciding with current school questions; 4) more easily readable reports
with specific recommendations for practice could improve the applicability of research,
and, perhaps even more important, "factors such as…efforts to focus on variables
amenable to intervention by users", could be expected to bridge the worlds of academe
and practice. Interestingly, Landry et al. (ibid.) introduce the concept of a new feature:
the linkage mechanism (i.e. the translation of research outcomes to practice as a distinct
expertise). With reference to the power of context an important finding of Landry et al.
(ibid.) should be mentioned here: taking into account the user's context, turned out to
be a far stronger variable than the user's needs. The recommendation by Landry et al.
(ibid.) to invest in the further development of skills that are required to identify school
themes, and to explain the applicability of research findings, have inspired the
methodology we describe in the current study.
In the Netherlands, a successful linkage mechanism, albeit in the field of
agriculture, started as early as 1876, when the predecessor of Wageningen University
and Research Centre was founded. Almost from its beginning, in 1877 the first regionally
based testing stations were set up, that brought workers in agriculture and cattle
breeding in contact with scientific researchers. Largely supported by the opportunities of
the Internet, to this day Wageningen is firmly associated with institutions and individual
workers in the domains of food production, the environment, and health. Their website
offers up to date reports and research outcomes considering current themes in the field
(for example http://www.livestockresearch.wur.nl), and, through a web‐based
information centre, workers in the field can submit research questions emerging from
practice, when no financial means for initiating research are available to them.
Admittedly, it is a complicating factor that educational research is carried out in
different scientific disciplines, not organized in one specific university, as is the case in
![Page 149: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/149.jpg)
5 A tailor made transfer of scientific knowledge to school practice
123
Wageningen. We feel, however, that the example of Wageningen University might offer
a valuable example for education, as will be elaborated at the end of this paper.
Furlong (2004) adds to the goal of a more effective validation of knowledge, the
noteworthy caution not to get involved again in "epistemological divisions and paradigm
wars", referring to the constant emphasis in political and policy debates on proven
"what works" research, and the strong priority for controlled randomized trials in
research design. In Furlong's view, researchers should consider the work of Schön (1983)
on the "reflective practitioner", and invest in connecting researchers with professionals
in schools while acknowledging the teacher's own subject knowledge and judgment. The
critical assessment of academic work, furthermore, is regarded to be a necessary specific
skill that should be embedded in on‐the‐job training of teachers in order to equip them
better for research utilization. Furlong (ibid.) also advocates that researchers and
practitioners should work together directly, in order to make research work in schools.
Reciprocal views of educational practitioners and academic researchers
Vanderlinde and Braak (2009) investigated in Flanders the opinions that
teachers, school leaders, researchers and intermediaries hold of one another's roles,
with regard to the gap between research and practice. The outcomes of their focus
group interviews indicate that particularly teachers are sceptical about the value of
research for their classroom practice, since they feel that research does not offer
enough practical applicability. School leaders, however, report that they do, in fact, try
to use research findings, notably for themes concerning the school organization.
Researchers expressed that their first and foremost concern was the assessment of their
work by the academic community, and that they lack the time and incentive to make
their work readable to a wider public, although they recognize that the use of technical
language may form a hindrance for practitioners. Not surprisingly, intermediaries did not
experience the distance between research and practice to be a problem, but rather a
challenge. Arguably, however, the opinion of the researchers in the focus group, to
delegate the transmission of knowledge to intermediaries, could be interpreted either as
an innovative idea or as yet a further seclusion of scholars in academe.
![Page 150: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/150.jpg)
124
Intermediate actors
Nutley, Walter and Davies (2002) take the epistemological discussion on the
preferred design of educational research further, in distinguishing the instrumental from
conceptual use of research. By their nature, results from controlled randomized trials
might more often be instrumental, since preferably specific discrete practices are
compared in these experiments. Bringing into mind, however, the complex and highly
contextualized characteristics of schools, also conceptual insights from research (for
example the role and function of the school amidst societal themes concerning equal
opportunities) might be appropriate and helpful. Nutley et al. (ibid.) refer specifically to
the significance of a dissemination medium for linking research to schools. They
emphasize, furthermore, the importance of the timeliness and accessibility of research
findings, and – not surprisingly– the flexibility, reliability and credibility of research.
Research to support school reform and evaluation
While the current paper wants to elucidate how the transfer of knowledge
between research and individual schools could be improved, the research base for
government policies, however, should also be mentioned here. The assumption is, that
the knowledge underlying government policies should preferably be accessible to
schools, in order to enhance the acceptance and implementation of general policies.
Slavin (2002), when reflecting on the enormously increased technical potential of data
analyses, speaks of a "scientific revolution that has the potential to profoundly
transform policy, practice and research". Yet, he records the fact that large‐scale reform
programmes in the U.S., for example the No Child Left Behind Act (U.S. Congress, 2001)–
albeit such programmes may need a longer period of time to demonstrate outcomes–
have so far failed to demonstrate significantly improved educational outcomes at the
system level. In the Netherlands, heads of school and teachers express doubts on the
knowledge base of government reform programmes, and doubt sometimes whether the
feasibility of translating reform programmes to school practice per se, has been the
subject of research at all. The Education Council of the Netherlands (2011), furthermore,
states that the tendency seems to be that large‐scale educational reforms have not so
demonstrated the expected results. In his inaugural lecture of 2001, Hessel Oosterbeek
already expressed his concern that teachers' motivation might erode, when they have to
![Page 151: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/151.jpg)
5 A tailor made transfer of scientific knowledge to school practice
125
comply with new government policies and then never learn about the effects of such
policies. Slavin (ibid.) critically notices that too often government programmes are not
based on evidence of actual effectiveness in practice. Slavin (ibid.), furthermore,
remarks critically that research can be found to support virtually any point of view given
a specific theme in a school; contradictory research outcomes considering the same
theme, may severely hamper the use of academic knowledge.
Even apart from doubts on an adequate research base for national school
reforms, Morell and Noguera (2011) criticize government policies for neglecting the fact
that some schools, for example, those with many low SES or migrant pupils,
demonstrate excellent outcomes, while others do not. National policies tend, however,
to address all such schools equally as disadvantaged. Worldwide, schools struggle with
context specific obstacles, when trying to implement national reform programmes.
Morell and Noguera (ibid.) convincingly explain how long term meaningful partnerships
between research and practice, leading to joint research agenda setting, research based
interventions in school practice and subsequent monitoring of new strategies, might
enhance the effectiveness of school reform. Research insights may be applied in
different ways in schools. Johnson (1998) distinguishes between instrumental (basis for
action), conceptual (offering insights), process (learning about current processes), and
symbolic (e.g. in political discussions) use of evaluations.
Finally, the Education Council of the Netherlands (2011) recommends three main
strategies for a more effective transmission of research to practice:
1. The formation of networks consisting of schools, universities, and centres for
education, by analogy with medical schools in academe. In addition, at each
school several teachers should develop competences for analysing their results in
a more scientific way, in order to be able to study the effect of interventions.
2. The foundation of a new agency, consisting of representatives from schools,
academe and the government, which coordinates agenda‐setting for
fundamental, practice‐based, and policy‐oriented educational research.
3. Considering the ambition that all schools, in the future, should invest more
systematically in quality improvement, the Council proposes that the Ministry of
Education should allow schools sufficient free rein, in order to develop methods
that fit the individual school best. The government, furthermore, should ensure
![Page 152: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/152.jpg)
126
that school reforms are introduced in a way that can be monitored effectively,
notably by the schools themselves.
Inspired by the academic work described in this Section, we have set out to
investigate in the current study, what obstacles hamper the use of existing academic
knowledge in school practice and how this process could be organized more effectively.
Based on the reviewed literature and our own experiences, we had two assumptions.
First, we assumed that schools are to a considerable extent influenced by their specific
context, and that this context should strictly be taken into account in order to make an
accurate match with relevant research publications; second, we assumed that research
could work in schools when there is a timely match with existing school questions (e.g.
Landry et al., 2001; Morell and Noguera, 2011). The method we designed and tested was
based on these two presuppositions. Through in‐depth interviews we collected main
school questions, and carried out a literature review for every question in search for
state‐of‐the‐art publications to match these school themes. Furthermore, we analysed
the nature of school questions and the different categories of academic knowledge that
supported best school practice. In this way, we set out to achieve a deeper
understanding of concrete intermediary practices that actually link in‐depth explored
school questions with adequate research outcomes. Finally, we took into account that
our method should allow scaling up to larger numbers of schools, within reasonable
costs.
5.3 Methodology
We based our methodology on the following research questions:
1. Do school leaders have significant questions that demand for academic research
outcomes as a guidance?
2. Can school questions be linked to manageable, high quality outcomes of existing
research?
3. What is the nature of school questions, and what categories of academic
knowledge can best be matched with these queries?
4. Can selected research findings subsequently be applied in school practice and
strategies?
![Page 153: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/153.jpg)
5 A tailor made transfer of scientific knowledge to school practice
127
5. Can specific impediments for the transfer of applicable knowledge to schools be
identified?
6. What are the requirements for an intermediate practice that could bridge
schools and academe satisfactory?
Below we will describe the interventions we carried out, and motivate our choice
to focus on the role of the Principal in the transfer and use of knowledge. We will
describe the method of semi‐structured interviews we used, as well as our procedure for
establishing the match between school questions and applicable research findings.
Finally, we will describe the format of the evaluative interview.
5.3.1 Six participating school leaders
We choose to focus on six schools because we wanted: a) a detailed insight in
the present use– or the lack thereof– of scientific knowledge in individual schools; while
b) we wanted to be able to compare schools and find possible patterns; and c) to have a
sufficiently large scale in terms of the number of pupils served by these schools. We
carried out our study in a real‐life context, which was another important reason why we
chose this design (Yin, 2009). In an attempt to make participating schools not too diverse
in too many dimensions, we invited only schools in Amsterdam31 to participate (see
Appendix 5‐1 for a general description of the participating schools).
All six schools offer the two highest tracks in the Dutch system for secondary
education, referred to as "academic tracks" in this chapter, since completion grants
access to higher education.32 Five schools also offer pre‐vocational tracks that qualify for
senior vocational education. We labelled schools A‐F in random order.
31 Calandlyceum, Comenius Lyceum, Hervormd Lyceum West, Open Scholengemeenschap Bijlmer, Spinozalyceum, IJburg College. 32 Official translations of the Dutch educational system in ISCED classification (International Standard Classification of Education by UNESCO, update1997):
VMBO: pre‐vocational secondary education, 4 years, ISCED 2, qualifying for senior secondary vocational education
HAVO: senior general education, 5 years, grade 1‐3 ISCED 2, grade 4‐5 ISCED 3, qualifying for higher education
VWO: pre‐university education, 6 years, grade 1‐3 ISCED 2, grade 4‐6 ISCED 3, qualifying for higher education
MBO: senior secondary vocational education, level 1 ISCED 2, level 2‐4 ISCED 3, level 4 qualifying for higher education
HBO: universities for applied sciences, ISCED 5B
![Page 154: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/154.jpg)
128
In the case of four schools (A, B, D, F), the vast majority of students have a
migrant background; most of these students were themselves born in the Netherlands,
but have at least one parent born abroad. Three of these four schools (A, D, F) mainly
enrol students with a Turkish or Moroccan background. One school (B) enrols mostly
students from Suriname, the Antilles, and Ghana. School E attracts many students from
a well‐educated population in a recently built neighbourhood, as well as students with a
migrant background with parents who had fewer chances to receive good schooling.
School C has a currently changing composition of students and enrols a growing share of
children with middle class parents, while the share of migrant students is decreasing.
The ambition of school C remains, however, to attract students who mirror the diverse
multi‐ethnic population of Amsterdam.
The total number of pupils enrolled in the selected schools (school year 2011‐
2012) is 6335. 1765 pupils are enrolled in the largest school, 353 in the smallest one. The
two smaller schools in our study are growing substantially in student numbers.
In some important aspects the six schools may be characterized as schools that
invest largely in "Systemic Equity" (Scott, 2001; Brown, Benkovitz, Muttillo and Urban,
2011): they have implemented coherent policies to establish equal opportunities for all
students, notably those with a non‐academic family background; they welcome diversity
and are investing in further specific professional development of teachers. All six
participating Principals were committed to the function and place of their school in the
context of societal themes, notably the full and meaningful participation of migrant
pupils in society, and their position on the future labour market.
We offered no cross‐information between participating schools; it may be
assumed, however, that the Principals themselves have frequent collegial contacts.
We chose to base our study on structured interviews with school leaders
because, according, for example, to the report on leadership and student learning,
funded by the Wallace Foundation (Louis, Wahlstrom, Michlin, Gordon, Thomas,
Leithwood, Anderson, Mascall, Strauss and Moore, 2010) "…Leadership explains five to
seven % of the variation in student learning across schools (not to be confused with the
WO: research universities, ISCED 5A
The completion of ISCED level 3 is the internationally agreed initial (or basic) qualification for the labour market. School leavers without an ISCED 3 qualification are regarded as early school leavers or drop outs.
![Page 155: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/155.jpg)
5 A tailor made transfer of scientific knowledge to school practice
129
very large within‐school effects that are likely). Five to seven %, however, is about one
quarter of the total across‐school variation explained by all school‐level variables, after
controlling for student intake or background" (Classroom factors explain more than a
third of the variation). The authors add, however, that it is as yet unclear what
characteristics and practices drive this school leader's effect on outcomes. With the
current study we hope to contribute more insight into the questions, ambitions and
practices of the six involved Principals.
Another reason to choose Principals as our source of information is their
intermediate role in discussing the function of a school in the context of societal themes,
with external parties, other schools and local and national politicians (Sanders, 2009).
Importantly, the six school leaders played an important role in the design of the present
study, where they acted as co‐creators.
5.3.2 Five separate phases of investigation
Phase 1: Semi‐structured in‐depth interviews (Research question 1)
The current research started with a 1.5 hour semi‐structured in‐depth interview
with the Principal. Each interviewee was informed about the overall purpose of the
current research; the participants were also informed about the method and time
schedule of the project. The interviews were (with permission) recorded.
The interview consisted of three parts (see Appendix 5‐2 for the complete list of
questions);
1. Part 1: A reflection on the school's context, in a way that was meaningful and
relevant in the view of the Principal.
2. Part 2: A description of the influence on specific current policies and teaching
practices of distinctive context characteristics, or characteristics of the student
population.
3. Part 3: When reflecting on parts 1 and 2 of the interview, what are important
remaining questions considering further ambitions and developments that are as
yet difficult to answer? What kind of scientific knowledge (practical,
organizational, conceptual) would support the Principal in further ambitions and
the development of school quality, and what are the problems, so far, with
obtaining this type of academic knowledge?
![Page 156: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/156.jpg)
130
A report of the interview, based on the recorded discussion, was returned to the
Principal, in order to verify that the text correctly covered the discussion.
Phase 2: The nature of school questions and the search for a match with scientific literature (Research question 2 and 3).
In every interview, between three and five main, complex questions were
identified, in total 21 questions for the six schools together. In this list different types of
school questions could be distinguished (e.g. conceptual and organizational themes,
long term policy effects, content of the curriculum, the balance between cognitive and
non‐cognitive skills and effective policies for teacher quality) that required
correspondingly different types of applicable academic research outcomes. The
literature search, in general, followed the order of trying to find, in the first place, causal
evidence, based on randomized trials or quasi‐experiments. Secondly, (meta‐) review
studies, carried out by top ranked researchers and institutes, were selected. Thirdly,
when school leaders had questions about good practice in other schools, case studies
carried out by highly ranked researchers were selected. International comparative
studies were selected to match questions on practices in other countries. School leaders
were informed about the nature of every academic research article we sent, following
the above order. Considering the quality of selected publications, the following list of
criteria has been used:
All selected articles had to be published in reviewed journals; books had to be
written or edited by academic authors who published highly ranked articles.
When case studies were used, studies in reviewed journals were selected, or
case studies published on programme websites of top ranking universities or
institutes.
Research outcomes should have clear prospects for translation to school
practices and ambitions.
In the case of research publications based on educational systems in other
countries (this was the case for most of the articles that were selected), the
relevance for the Dutch situation should be obvious (and was explicitly
explained).
No publications before 1999 were used.
![Page 157: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/157.jpg)
5 A tailor made transfer of scientific knowledge to school practice
131
Phase 3: Feedback in the format of a letter
In order to prevent that the Principals from having to read yet another report– a
recurrent complaint expressed during the interviews– the matching academic
publications were sent in the format of a letter, specific for each school, of around five
pages. In this letter the school's questions were summarized and the method used for
selecting publications (described above) was also explained. The links to all the articles
were attached as well. The letter briefly motivated the choice of publications, and how
and why, in our view, these connected in our view to the school's questions. The school
leader could choose to use the short descriptions in the letter, or study the whole of the
attached research papers.
The Principals each received between four and ten references in total. For
practically all school questions an extended list of high quality research papers was
available in academic databases; the idea was, however, to send no more than one or
two state‐of‐the‐art references per question, with the offer that more references could
be sent upon request. We restricted our answer to a maximum of ten references per
letter.
Phase 4: Intermediate check in three of the six schools of the applicability of the research matches that had been sent (Research question 4)
The concise style of the letters raised questions by us on the accessibility of the
texts concerned. In order to test this, with three Principals (A, B and E) an extra meeting
was planned to reflect on the applicability of the recommended research in the letter.
The following interview format was followed:
Did the Principal have the opportunity to read the letter and the attached
scientific articles?
Did the articles, in the Principal's view, match his or her initial questions?
Did the letter raise any further questions, and was extra information needed?
Phase 5: Evaluative interview (Research question 4, 5 and 6)
The evaluative interview took 1.5 hours and was clearly structured around six
evaluating questions on the utilization and manageability of the offered sources of
relevant knowledge. These questions concerned: 1) an impression of the whole
trajectory of the current study; 2) the usefulness of the academic knowledge for
![Page 158: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/158.jpg)
132
discussions with staff; 3) the nature of the use of the presented research outcomes:
instrumental, conceptual, re‐consideration of running processes, use in external debates
(after Johnson, 1998); 4) the influence of the scientific insights on the interpretation of
the school's context; 5) obstacles for the transfer of academic knowledge to practice;
and 6) the options for scaling up the current method (for the full list of questions, see
Appendix 5‐3). The evaluating interview was planned three months after the school
leaders had received the letter with research recommendations. The interviews were
recorded with permission.
The final interview partly mirrored the initial interview: the initial interview
started with the Principal's reflection on the school context and questions that followed
from this specific context, the final interview ended with the question whether the
recommended scientific insights had influenced the school leader's views on the school's
environment.
5.4 Portraits of the participating schools and their main questions
Below we will describe the participating schools in the following order of
discussion:
We start with a characterization of every school, and summarize the number and
type of questions. All questions (numbered Q1, Q2, etc.) and the matches that came out
of the subsequent literature review (numbered accordingly A1, A2, etc.) are listed in
Appendix 5‐4. Table 5‐1 presents all question clustered in themes, plus the matched
publications. Table 5‐2 lists the categories of research publications that have been
selected per thematic cluster of school questions.
Furthermore, we describe an intermediate discussion in the case of Schools A, B
and E, shortly after the Principal had received the scientific matches with his or her
queries.
Finally, we quote the Principal's evaluative remarks on the project and present
an inventory of the use of the suggested research findings, as well as recommendations
for the follow‐up of this project. Table 5‐3 presents the outcomes of the evaluative
interview, and Table 5‐4 the applications of the research outcomes in school practice.
![Page 159: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/159.jpg)
5 A tailor made transfer of scientific knowledge to school practice
133
5.4.1 School A: An enriched idea of the school's added value for students with a migrant family history
School A serves a vast majority of pupils with a Turkish and Moroccan
background; the school offers the two highest secondary tracks, and, therefore all pupils
with a diploma qualify for higher education.
The Principal's most important question concerned the segregated character of
the school, which was situated in most of the pupils' area of residence. After students
have successfully passed their final exam they typically will enrol in higher education,
and may be confronted with a different culture that may be partly unfamiliar to them.
There might be even more challenges, the Principal expects, when his pupils, after
having completed a university degree, enter the higher strata of the labour market. The
Principal had two main questions: Is it an advantage for pupils to be enrolled in a school
that is specialized in upward mobility, albeit segregated‐ or would pupils be better
prepared for a full and meaningful participation in Dutch society, when they would
attend a desegregated school?33 Could pupils, furthermore, benefit from more teachers
who also have a migrant background?
School A raised three main questions, that could each be matched to several
research publications. Two questions were of a conceptual, strategic nature, concerning
segregation of schools; one question concerned policies for hiring teachers. 2 books
were suggested, 1 study based on causal evidence, 1 international comparative study, 1
randomized experiment, 1 website.
The extra discussion, to explore the opportunities for the use of the selected research publications
School leader A was surprised that the research findings in important ways
differed from what he had expected. His expectation had been that desegregated
schools were always the better option, especially considering future chances on the
labour market for his pupils. The Principal, furthermore, valued the presented research
on the importance of developing multiple identities for migrant youth, and said that he
felt encouraged to invest further in specialized teacher competences with regard to
33 In the Netherlands all schools are funded equally by the government and low SES pupils, including many pupils with a migrant background, receive extra funding. Schools are independent of local tax revenues and parents do not pay tuition fee in elementary and secondary education.
![Page 160: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/160.jpg)
134
migrant students. The research findings had been used in debates with local political
leaders and other school leaders in the district.
The evaluative interview
The Principal stated that the project had deepened his thinking about the role
and function of his school in a community largely consisting of people with a migrant
background.
Based on our discussions and the letter with relevant research
publications, I started thinking somewhat differently about my school. I
still think desegregated mixed schools are the preferred option, but I am
more inclined to think now about my school as an accelerator for the full
participation of migrant pupils in society. I especially appreciated the
participation (NB of the researcher) in the debate with political governors
of the Amsterdam district New West. Looking at segregation from
different knowledge‐ based perspectives was helpful; it made me realize
the values of my school, although it is a segregated school.
I valued the scientific answers to my questions and especially that
research acknowledged that schools are deeply contextualized; even a
school in the South East district of Amsterdam may differ in important
aspects from my school in the New West district.34
The content of the letter had not been discussed with colleagues within the
school; school leader A considers the complex aspects of equal opportunities for migrant
children to be his main personal assignment. The literature we had suggested on
improving opportunities for migrant pupils resulted, according to the Principal, in a more
positive view on his school "I have a warmer appreciation of my own school, and have a
higher esteem of our professionalism".
The offered scientific knowledge had influenced views on current school
processes, and was used in strategic discussions with external stakeholders. The
information was not used for direct instrumental actions.
34 All interviews were in Dutch; citations have been faithfully translated literally in English.
![Page 161: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/161.jpg)
5 A tailor made transfer of scientific knowledge to school practice
135
The method developed in the current study, according to the Principal, could be
expected to be of special significance for starting headmasters. The Principal made a
salient final remark: the fact that the interviewer in the current study, unlike a coach or
a consultant, did not charge any fee had strengthened his trust in the project.
5.4.2 School B: the advantage of a professional review of academic literature
School B is situated in a district of Amsterdam with relatively high
unemployment, poverty, and single‐parent families. Around 60% of the pupils have a
Surinamese, Antillean, or Ghanese background. The school has a long history of
extended comprehensive education (the Dutch system typically has early tracking), and
is committed to this principle. The school, however, faces strong competition (notably
regarding talented future students and their parents) from schools in the city centre that
start with early tracking at age 12 at academic levels.
The Principal's main questions refer to the school's choice for extended
comprehensive learning: Is postponed tracking and stimulating upward mobility to a
higher track, chiefly benefitting pupils with a non‐academic family background, or is
longer comprehensive learning benefitting all pupils? Related to the above, the Principal
raised questions on the best balance between cognition on the one hand– and meta‐
cognitive skills (notably academic reasoning) and personal growth on the other, notably
for low SES pupils. Like the five other participating Principals, Principal B emphasized
also the importance of receiving data on the future success of the school's pupils in
higher education.
School B raised three questions that could each be matched to research
publications; the question on the future study success of pupils in higher education after
leaving school could, however, only be answered in general, not for school B specifically.
Two questions referred to both the school's role in the local community and subsequent
choices in current school processes. One question concerned data on later study success
in higher education of former students.
The extra intermediate discussion
Principal B strongly recognized the school's questions in the presented academic
literature. She was particularly interested in research outcomes that prove that schools
which offer a broader array of tracks (both professional and academic tracks) do indeed
![Page 162: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/162.jpg)
136
increase chances for upward mobility (Van Elk et al., 2011). On the other hand, she
recognizes also, however, that this still might not change the current trend that high
potential students tend to choose a school in the city centre that exclusively offers
academic tracks. The selected research publications also support the school's new
programmes on higher order learning skills. The Principal expressed the view that the
intermediate explorative discussion made her more aware of the opportunities for use
of the research findings, and motivated her to discuss the letter with the school's project
team on diversity in education.
The evaluative interview
The school leader pointed out that it is also relevant to learn which questions as
yet cannot unequivocally be answered by research, for example the question: Does a
broad school like hers, providing all tracks (professional and academic), potentially
generate better results for all pupils, or might, in fact, early tracking for gifted students
be more attractive, especially for pupils with more affluent parents? She added that the
current research project came at the right time, since the school was deeply involved in
discussions about continuing its mission of comprehensive learning, with special
emphasis on non‐cognitive skills, or instead making a change to partial earlier tracking.
It is good to be able to use existing knowledge to plan your strategies. We
think too often in education that we cannot select an appropriate
research basis for our school innovations, since research on the same
subject often seems to present contradictory outcomes. Because of this
project we know now which research is relevant in our context, and why.
Immediately upon arrival, the letter was sent through to the management team
and the school's project group for diversity, which under the leadership of a recently
appointed vice‐Principal was reconsidering the school's mission of postponed tracking.
The vice Principal expressed the view that "the presented research inspired our thinking
and gave us a firm base to stand on".
The research findings had been used in different ways, direct and instrumental,
for conceptual thinking, to assess current developments, and in external debates. The
school leadership and the teachers cooperating in the project "diversity", felt more
![Page 163: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/163.jpg)
5 A tailor made transfer of scientific knowledge to school practice
137
reassured after studying the suggested research, to remain faithful– but with new
energy and ambitions– to the school's long history of establishing a balance between
cognitive learning and personal growth, in the context of a diverse community. It was,
furthermore, seen as an advantage to have a source of information from outside the
school.
The Principal expected that the methodology we developed could be interesting
for cooperating groups of schools. Seminars and focused discussions on specific themes
could largely benefit form a tailor‐made selection of academic research. She also
expected herself to have further questions on relevant research outcomes in the future.
5.4.3 School C: The important next step, translating research outcomes to teacher practices
The school's educational philosophy is based on a didactical model that
stimulates pupils to gradually take more responsibility for their own learning process
(known as the Dalton pedagogy, based on the work of Helen Parkhurst). Pupils can
choose, for example, to partly make their own choices by taking extra classes in specific
subjects. The school, moreover, offers ample opportunities for cultural and artistic
development, notably in music. Traditionally, this type of school attracts more middle‐
and upper‐class students, with typically well‐educated parents. The Principal confronts
the school's pedagogical principles with contrasting ideas of strict and ordered school
policies, mainly aimed at cognitive learning, that are in public and political debate often
regarded as being more effective for pupils with a non‐academic background (which is
the case for many migrant students).
The main questions concerned a further knowledge base for the school's
didactical system, an effective organizational model for teachers working with this
system, and research on the effects of grade repetition (a particular concern of the
school leader). three questions were raised; one question could be matched with
research that approached the school issue, but was not an exact match. The other two
questions could be matched directly. 3 reviews were suggested (one of these an
inaugural lecture), 1 report, 1 case study, 1 website.
![Page 164: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/164.jpg)
138
The evaluative interview
The Principal emphasized that it was interesting to learn that the profound
structured discussion on his own school themes could indeed be linked to research
outcomes, although one question could not be matched precisely with the relevant
literature.
It is important to realize that this research project also indicates which
questions still need to be further investigated, because indeed we do not
know the answer to, for example, the question whether my school's
philosophy on learning, results in better learning strategies and planning
later on in higher education.
I felt the interview was an invitation to try to formulate relevant questions
concerning my vision and ambitions, related to actual school practice.
It is important to follow up this research, including the actual
implementation in school practice and policies.
The critical remark was made that the most important part of an adequate
utilization of knowledge, in the view of Principal C, had yet to follow: the translation of
research findings to actual classroom practice; this translation process, in his view,
requires further guidance by case‐study research on good practice. Therefore, the
presented research had not yet been used directly.
The school leader felt that linking school questions to scientific knowledge could
be an interesting starting point for inter‐collegial seminars and discussions.
5.4.4 School D: Taking time to focus on the complex questions we encounter
School D has largely invested in special programmes, like bilingual education
(Dutch and English), top sport tracks, and Technasium (technical and science education).
The school is committed to maximizing opportunities for learning for all pupils, and no
longer thinks along lines of ethnic or SES differences. In their view, they serve the
surrounding community and try to relate to the specific learning requirements of all
young people living in the nearest of the school. They are, furthermore, committed to
postponed tracking, in order to grant pupils more time for exploring their talents and
![Page 165: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/165.jpg)
5 A tailor made transfer of scientific knowledge to school practice
139
capacities. The director's question concerned postponed tracking, extra courses for e‐
learning and its consequences for the organizational model of the school. Considering
the fact that most pupils in school D have a migrant family background, the director was
concerned about the consequences of currently occurring societal prejudice. Finally, the
director inquired after the most effective (evidence based) financial intervention, tailor‐
made for his school and its environment. Five questions were raised, one could be
matched with more general research outcomes, and the other four could be matched
directly.
The evaluative interview
The input for further conceptual thinking on the school's role and function
amidst complex societal themes was felt to be an important advantage of this research
project. The letter with links to relevant research, moreover, turned out to be a valued
source of information for discussions with the school leader's successor.
In the first interview we discussed subjects that are constantly in the back
of my mind; but it is hard to find the time to really organize my thinking. I
appreciated this opportunity. The letter that followed up on the interview
showed that high quality research supports– and in some cases even
proves– our visions on the added value of a broad school like ours that
offers all secondary tracks and extra programmes as well. It sounded like
music in my ears that there is indeed evidence that pupils in professional
tracks have more chances to eventually complete higher education when
they start in a broad school.
I feel supported by this project in our choice to be fully aware of our
ethical, moral and societal responsibility– as opposed to marketing and
competition driven strategies.
In external political debates we feel strengthened in our position.
According to the Principal, especially large complex themes require focused
meetings and time to work out all consequences, time that is always too scarce in the
everyday complexity of school life. It was mentioned specifically that a knowledge‐base
![Page 166: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/166.jpg)
140
for standing practice may be as important as scientific insights for guiding future
projects.
The current method was considered to be worthwhile for cooperating schools,
clustered, for example, around serving a diverse student body.
5.4.5 School E: The pleasure of taking time for reflection
School E is a recently‐founded new school, offering ample opportunities for
choices on school organization and didactical principles, as well as the hiring of new
staff. The Principal is dedicated to making evidence‐based decisions together with the
teachers, and emphasizes the school's ambitions in designing innovative learning for the
21st century. The four raised questions referred to these themes. All questions could be
matched with relevant research outcomes. 3 publications with causal evidence were
used, 4 (meta‐) reviews, 1 case study, 1 report and 1 book.
The extra discussion
Principal E has a background in educational sciences, and is committed to making
knowledge driven decisions for all strategic school themes. He has indeed studied the
research findings in the letter extensively, and feels further supported in his innovative
ambitions and practice. The Principal emphasized that the academic knowledge
provided had inspired him to look at current school developments from new angles,
notably considering the balance between skills and cognition.
The evaluative interview
According to the Principal, reflecting on the overall design of school innovation,
to consider limits and opportunities and to reflect on the concepts underlying one'
thoughts, requires time and concentration. Therefore, one of the main advantages of
the current project was, in the view of the director, the in‐depth discussion on such
concepts.
This was a "gift", to just talk to someone in a certain thoughtful way
about important issues; you do not find the time in the bustle of everyday
work to reflect on such themes; to sit back and think about the larger
design in your way of working, your interventions and your concepts, all
the things that relate to this, what are your impediments, what
![Page 167: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/167.jpg)
5 A tailor made transfer of scientific knowledge to school practice
141
information do you need specifically. You do not easily find the time for
this, at least not in my practice as a school leader. That requires someone
who asks questions in a specific way, who from a comparable background
in educational leadership, and for a specific purpose, continues to ask
further questions about the things that are of concern to you. I really
enjoyed that. The knowledge sources that were offered, you do not find
easily as a Principal. I will certainly do more with this, notably considering
the process of translating requirements for skills for the 21st century to
school practice and policies.
The Principal, furthermore, felt supported in his vision that innovative learning
for the 21st century requires specific skills and further training of teachers. He had
shared the letter with his team and among members of his external network of
innovative schools. The information was used for conceptual thinking, and several
internal and external meetings. While studying the literature on the subject of
innovative learning he had realized once more, how conservative, in his view, education
tends to be. And yet, society is changing at an unprecedented tempo, if not only because
of the phenomenal new digital prospects.
The school leader's recommendation was to use our method to support schools
that focus on the, in his view, under‐studied theme of urban education in the
Netherlands.
5.4.6 School F: Time is the most scarce ingredient
School F serves a student population consisting of more than 90% pupils with a
migrant background, among them many youngsters who are the first in their family to
enrol in secondary tracks that qualify for access to higher education. The school
structurally invests in continuous formative assessment of pupils, with the aim of
enhancing opportunities for upward mobility to higher secondary tracks. The main
questions concern extra programmes (on top of the standard curriculum) for migrant
students, further teacher training, and leadership. Three questions were discussed, that
could all be matched to research publications. 3 matches with publications on causal
evidence were made, 2 reviews, and 1 report.
![Page 168: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/168.jpg)
142
The evaluative interview
School leader F was involved in two large projects the school itself had
commissioned, and because of that had not found the time to study the letter with
research links connected to the school questions.
I realize that I did not find the time to use the information in the letter,
although I think that this might be worthwhile. I would have to study the
literature suggested in the letter, however, outside working hours. I did
send the information to the colleagues in my leadership team, but nothing
came out of that. We are kept busy with all the things that happen every
day in a school.
Furthermore, we commissioned two investigations ourselves, one
quantitative research project into the characteristics of youth that
potentially may choose for our school, and another qualitative research
into the expectations of parents in this part of the city.
We are putting a lot of effort in attracting enough pupils for our academic
tracks, and try to prevent losing many of our talented pupils to schools in
the city centre.
The Principal, moreover, thinks that inspiring cooperation with other schools
might be more effective than the use of scientific research outcomes. When a strategy
has proven to work in another school, the chances are it will work in your school as well.
She also critically questions the further training of research skills for teachers, and rather
emphasizes the importance of peer learning and examples of good practice in other
schools.
5.5 Findings
In this Section we will describe our findings, following the order of the research
questions we presented in Section 5.3.
The rich interviews taught us that school leaders do indeed have important
questions that in most cases had already occupied their thinking for a long time.
![Page 169: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/169.jpg)
5 A tailor made transfer of scientific knowledge to school practice
143
Among a total of 21 questions, in three cases the question could not be linked
exactly to a research publication. In all other cases, a match could be made with one or
several research publications. Considering the three question that could not exactly be
matched, in one case the question was aimed at the long‐term effects of the specific
didactical model of school C; no such research was found, but we nevertheless found
relevant research publications that well‐approached the school leader's question. A
question on the best options for financial investment in school D, would need a much
wider literature study to link options exactly to this particular school; our answer was
more generally directed at the paramount importance of teacher quality. A third
question that could not directly be addressed was asked by all six school leaders, and
concerned the wish for data on the future achievements of former students, in higher
education, and on the labour market. This recommendation will be discussed with the
ministry of education.
The questions could be clustered around five main themes:
1. Segregation of schools
2. Tracking in secondary education
3. The balance between skills and competences on the one hand, and cognitive
learning on the other hand
4. Teacher quality
5. School innovation and leadership
The categories of questions, plus the suggested literature references are
presented in Table 5‐1; for every literature reference a typology of the publication has
been added in this table. Table 5‐2 illustrates the categories of publications that were
used per cluster of questions. Table 5‐3 gives an overview of how the Principals have
experienced our method, how they made use of the new knowledge, and how they
thought about the feasibility for scaling up the current project. Table 5‐4, finally, shows
more in detail how the Principals had made use of the suggested research (conceptual,
process, symbolic, and instrumental).
Taking into account the different context of every individual school, the same‐ or
a largely similar question (e.g. on segregation and diversity), resulted in different best
matches with scientific literature.
![Page 170: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/170.jpg)
144
In five out of six participating schools, the suggested research had mostly been
applied for: 1) reconsidering current strategies; 2) supporting current strategies; 3)
developing new policies; 4) internal and external debates; and 5) making a motivated
choice out of contradictory research outcomes on the same subject (Table 5‐4). In the
case of five schools the suggested research was used to further elaborate the role of the
school in its specific environment; in four schools the suggested research publications
were used to assess current practices, and four school leaders had used the new insights
in external debates. One school leader suggested that the translation of relevant
research to concrete teacher practice could be enhanced by a follow up project. One
school leader felt that illustrative examples in other schools might be more effective
than academic research literature.
We encountered the unanticipated problem, however, that 14 out of the 25
publications were only accessible in the university environment: universities pay
substantial licence fees to allow researchers to study scientific journals. As a
consequence, a considerable part of the publications we selected could not have been
found by school leaders themselves in the first place (apart from the need to possess the
research skills to be able to identify appropriate matches between question and answer)
since schools can never afford these substantial licence fees.
In Schools A, B and E, shortly after the Principals had received our letter with
links to applicable research, an extra intermediate meeting was planned to explore the
options for using this new information. In the case of the other Schools, C, D and F, the
evaluative interview was the first occasion to discuss the letter. In this last set of schools,
the evaluative interview much resembled the extra explorative interview held in the first
three schools.
![Page 171: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/171.jpg)
5 A tailor made transfer of scientific knowledge to school practice
145
Table 5‐1: All school questions clustered in five domains, plus matches with selected scientific publications and the nature of every publication.
Cluster Literature reference School
Segregation; dissimilarity school and community culture‐ culture in higher education and the labour market; multiple identities; civic education.
Banks (2009); book. Bunar (2010), empirical study. Cooper (2011), book. Dobbie and Fryer (2009), causal evidence. Parker, Ninomiya and Cogan (1999), international comparison.
A, D
Tracking in secondary education‐ comprehensive learning. Intergenerational mobility.
Brunello and Checchi (2007), international comparison, evidence. Crul and Schneider (2009), comparative study. Hanushek and Woessman (2006), evidence. OECD (2011), report. Pekkarinen, Uusitalo and Kerr (2009), evidence. Van Elk, van der Steeg and Webbink (2011), evidence.
B, C, D
Balance cognitive learning– non‐cognitive; 21st century skills; Academic skills for pupils from non‐ academic families; Learning for personal growth.
Heckman and Montera (2009), comparative study. Payne (2010), book. Zimmerman (2002), review. Zohar and Dori (2003), empirical study.
B, C, E
Teacher quality; innovative organizational models for teacher cooperation. Teacher ethnicity reflecting pupils' ethnicity.
Boyd, Grossman, Lankford and Wyckoff (2009), evidence. Brown (2007), literature review. Darling Hammond and Youngs (2002), review. Dee (2004), randomized experiment. Jackson and Bruegmann (2009), evidence. Hanushek and Rivkin (2007), evidence. McKinsey (2007), report. Saunders, Goldenberg and Gallimore (2009), quasi‐ experiment. Van der Grift (2010), inaugural lecture.
A, C, D, E, F
Innovative school organization and leadership; digital learning; interactive learning with other actors in society. Extra programmes e.g. sports, music, bilingual curriculum.
Ala‐Mutka, Punie and Redecker (2008), report. Ancess (2000), case study. Cheung and Slavin (2011), meta‐analysis. Nilsson (2008), case study. Notten (2009), descriptive. Sanders and Harvey (2002), case study. Tefera, Frankenberg, Siegel‐Hawley and Chirichigno (2011), practical guide. Välijärvi and Sahlberg (2008), review.
C, D, E, F
Table 5‐2: Categories of selected research publications, per cluster of questions. (Note that the same publication may be suggested to several schools; e.g. 5 meta‐review studies were matched 15 times with a school question).
Cluster of Questions
Category of research
Segregation Tracking Cognitive Learning and
Skills
Teacher Quality
Organization and Leadership
Total
Causal Evidence, Randomized (quasi‐) Experiment
1 4 5 10
(Meta‐) Review 1 2 2 5
Case Study 5 5
(International) Comparative Study
1 1 1 3
Report 1 1 1 1 4
Book 2 1 3
Empirical Study 1 1
![Page 172: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/172.jpg)
146
Table 5‐3: Overview of the outcomes of the evaluative interview. Schools A, B, and E had an extra intermediate exploratory discussion.
School Overall impression
Informationshared with others
How was the suggested research used
Influence on views on the school context
Method interesting for other schools
A Scientific basis strengthens strategies. Knowledgeable sparring partner appreciated.
No. Themes concern my personal goals and struggles.
Conceptual thinking has been influenced.
Alternative views on the added value of school.
Scientific basis strengthens debate. Use method for starting Principals.
B Relevant selection of literature useful, albeit not all school questions could be answered directly.
Yes. Discussed with school board and with teachers.
Recommendations have strengthened the school in the complex redefining of strategies.
Better understanding of complex relationship between context and school.
Research input worthwhile for groups of cooperating schools.
C Connection between research and practice offers important new insights.
Yes. With vice‐Principal and friends.
Scientific basis supports current project to re‐define school didactics.
Research not yet used.
Follow‐up important: translate research to teacher practice.
D Grateful for insight into research base for complex school questions.
Yes. With colleagues school leadership and in meetings with other schools.
Influence on conceptual thinking most important and appreciated.
Supported development of strategies.
Use method for clusters of cooperating schools.
E Valued reflective discussions with researcher who is also experienced in school practice as well.
Yes. In my team, network, and at the Ministry of Education.
Many links to current school innovation. Inspired further reading and conceptual thinking. Support for ambitions.
Supported ambition for learning for future society. Helpful overview of strategic indicators.
Specialize in Urban Education, under‐ studied concept.
F Due to pressing school developments, no time yet to study the letter.
Send to management team and teacher project. Not discussed.
‐ ‐ Illustrative good practices in other schools more relevant.
![Page 173: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/173.jpg)
5 A tailor made transfer of scientific knowledge to school practice
147
Table 5‐4: Application (after Johnson, 1998) of research outcomes by school leaders after three months.
Type of Application Frequency
1. Conceptual 1a: rethinking strategies 3
1b: new projects 1
1c: school's role and societal issues 5
2. Process 2a: support current practices 4
2b: used in discussions with board of overseers 1
2c: support current decision‐making 2
3. Symbolic 3a: use in external debates 4
3b: use in political debates 1
4. Instrumental Direct use in teacher practice 2
5.6 Conclusions and Discussion
In the current explorative study we have made an inventory of existing questions
on the subject of improving educational quality in six urban schools, and tested a
practical method for matching these themes to relevant and manageable outcomes of
existing high‐quality research. We have described both the characteristics of the
questions and the categories of scientific publications that matched these questions. In
the course of the project, we found that taking into account the specific context of
schools and a timely and precise match between questions and answer may add to an
effective use of existing research in schools. Furthermore, we described the ways school
leaders made use of the presented academic knowledge, and mention the obstacles we
encountered concerning the use of academic knowledge in schools. We made an
inventory of the evaluative remarks that the participating Principals made of the current
project and their suggestions for expanding the project to other schools. On the basis of
these suggestions, we reflect on the feasibility of an intermediate function between
educational research and practice, and report on concrete follow‐up.
5.6.1 Principals have questions that can be linked to research outcomes
Considering our first research question, we have found that school leaders do
indeed have important questions, which in their view demand an academic knowledge
basis. As we expected, we found that research publications only rarely reach the school.
Typically, school leaders learn about new educational developments through thematic
reports or publications on government policies. However, in most cases, these
publications and policies do not coincide with current school themes.
![Page 174: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/174.jpg)
148
Out of a total of 21 school questions, all but three questions could indeed be
matched with existing research publications that could be translated to school strategies
and practice (Research question 2, Table 5‐1). For the three questions we could not
match exactly, we found an approximation in two cases, and we will try to put the third
question up to the higher level of the research agenda of the Ministry of Education (the
question addressed long‐term monitoring data on former pupils).
5.6.2 Comparable questions, different approaches
Considering Research question 3 (the nature of questions and answers), the main
questions could be clustered in five domains (Table 5‐2): 1) segregation of schools; 2)
early tracking; 3) the balance between cognition and skills; 4) teacher quality; and 5)
innovation. Although schools may have comparable questions, depending on their
context the most appropriate answer may be different. To illustrate this, we consider
the example of the questions in the cluster "segregation", concerning the best strategy
for serving students from non‐academic families, notably students with a migrant
background. This type of question was raised by all school leaders. In one case we
suggested research that demonstrates the positive effect of close cooperation with
parents, because this particular school was located in a neighbourhood with many
migrant parents who themselves did not have the opportunity to enrol at the level of
academic learning when they were young. In another school, however, we presented
research into the beneficial effects for all students of additional prestigious courses in
academic thinking, because this particular school ran the risk of losing its most talented
pupils to schools with a higher SES population in the city centre. Yet another school,
convinced of the added value of a diverse student body, wanted to attract more first
time academic learners, whereas the school was increasingly attracting middle and
upper class pupils because of its special emphasis on the Arts. In this case, we suggested
research into magnet schools that invest in the arts to enhance opportunities, especially
for children with a non‐academic background.
As Table 5‐2 shows, most of the publications we selected presented causal
evidence (10). The choice for a review or a case study publication was usually motivated
by the fact that the Principal had specifically asked for illustrations of the use of
alternative strategies. Note that the same publications may have been suggested to
![Page 175: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/175.jpg)
5 A tailor made transfer of scientific knowledge to school practice
149
more than one school. They valued, furthermore, publications that offer an overview
and analysis of large numbers of articles on a specific theme, because this offered them
more comparative insights.
5.6.3 Actual applications of academic knowledge and impediments
Our overall finding is that a match with existing school questions can– according
to the interviewed Principals– in important ways contribute: to developing school
strategies; offer support for current policies; inspire communication with the school
board, and strategic discussions with neighbourhood and municipal (political) governors
(Research question 4). After only three months, in the case of 17 out of 21 of the school
questions, five out of six Principals had already used the provided evidence for further
school development and innovative programmes or as a knowledge base for current
processes, notwithstanding the short period of time of three months between receiving
our research suggestions and the final interview.
As Table 5‐4 illustrates, in most cases the new knowledge and insights had been
used to support and inspire conceptual thinking (9 times), especially considering the
school's role in societal issues like segregation. In 7 cases, the research outcomes were
used to guide and support current school processes. In 5 cases the information was used
in debates with external stakeholders. Instrumental, direct use occurred only 2 times.
The main hurdles in knowledge dissemination (Research question 5) we
identified are the following: the Principals indicated that specialized professional skills
are needed to match upcoming questions to relevant literature, skills which are not
always available in schools. Moreover, the participants confirmed that the structured
initial interview had been essential in the articulation of significant themes. However,
we also encountered a more serious problem: we found that about 50% of the papers
we matched to school questions were not accessible outside the network of the
university. Access to scientific papers, in most cases, requires the payment of substantial
licence fees, which schools cannot afford.
5.6.4 Contributing to the use of existing research in school practice
Our study suggests that a bridging function between academe and secondary
schools could enhance the use of academic knowledge in schools, notably since schools,
on the one hand, and research institutes, on the other, obviously have different goals
![Page 176: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/176.jpg)
150
and organizational incentives (Research question 6). The involved school leaders were of
the opinion that it would be worthwhile to expand the method we developed to other
schools. They stressed the importance of our tailor‐made approach, and were in favour
of an intermediate flexible network function, rather than a bridging institute. More
specifically, they suggested the development of courses in knowledge based leadership
for starting school leaders, and seminars for experienced school leaders to study more in
depth main school questions. Furthermore, the involved school leaders suggested
initiating clusters of cooperating schools, in order to create opportunities for
comparative research based experiments. One of the Principals advocated the start of a
cooperative cluster especially for urban schools, since– in his view– urban education
appears to be an understudied theme in the Netherlands. Finally, it was suggested to
follow‐up the current project with research‐based strategies for the translation of
research outcomes from theory to actual classroom practice.
Important steps have been made by the Ministry of Education, Culture and
Science, and by the Netherlands Organization for Scientific Research, to start new
cooperatives between researchers and schools in designing a joint agenda for
educational research. Additionally, the current research contributes to a more effective
use of research in school practice, by linking already existing research publications to
school themes. By using existing research, school questions can be addressed in a
relatively short period of time.
Several outcomes of the current research initiated direct follow up. 1) A cluster
of innovative schools has asked for a programme to be started to link research to school
developments as a structural new way of working; and 2) A leading institute offering
advanced courses in educational development for school leaders and teachers in all
levels of education is interested in scaling up the method we developed. Moreover,
opportunities for advanced courses for school leaders aimed at applying research in
school practice are currently being discussed.
In summary, the method we developed what may be a first step to make
research actually work in schools, when the following requirements for its effective
transfer from theory to practice are taken into account:
Schools should be understood in their specific context in order to match theory
accurately to practice.
![Page 177: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/177.jpg)
5 A tailor made transfer of scientific knowledge to school practice
151
A timely match between actual questions and research answers is a prerequisite.
School leaders expect a motivated selection and a workable amount of research
findings to support their work; neither general reports, nor long lists of
publications to choose from, fit their demands.
The participating Principals were mostly interested in academic knowledge that
guides their conceptual thinking and decisions on school strategies, related to
the school environment– both the surrounding community and societal themes.
According to the Principals, in‐depth discussions between researchers and school
leaders which explore options for making use of academic knowledge can
contribute in major ways to the transfer of research from theory to practice.
The start‐up of a flexible network consisting of school leaders and academics will
be further elaborated
A serious obstacle needs further study: academic articles, in many cases, are not
freely accessible outside the university environment because of the licence fees
required to access scientific journals.
5.6.5 Contributing to research into the valorisation of scientific knowledge
The findings in the current case study may also contribute to research in the
domain of the valorisation and utilization of knowledge. In the current study we tried to
avoid epistemological discussions (except an occasional reflection) on the intricate
differences between the culture of academe and school practice. We focused, more
concretely, on designing an intermediate practice that acknowledges the characteristics
of both worlds, and yet sets out to establish a connection that makes research actually
work in schools. The participating Principals supported the feasibility of expanding our
method to secondary education at large. As to the question what competences
professionals in an intermediate network function should have, principals expressed the
view that working experience in both worlds should be a strict requirement; this may
include both school leaders with research experience and researchers who actually
worked inside schools for a longer period of time. All participants agreed that flexible
tailor‐made networks consisting of scholars, professionals in schools and policy makers
could in important ways add to the effective use of scientific knowledge in educational
practice.
![Page 178: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/178.jpg)
152
5.6.6 Limitations and further research
An obvious limitation of our study is the short period of time (three months)
between sending the response to the questions raised in the first interview, and the final
evaluative interview. The results of knowledge‐based interventions may require
monitoring over a longer period of time. Furthermore, as one of the Principals stated,
the utilization of academic knowledge might benefit from follow up programmes on the
actual translation of research to classroom practice. We consider that further research
into the possible effects on school outcomes of knowledge‐based interventions to be
our most important subject for follow‐up research. Clusters of cooperating schools may
offer interesting opportunities for the set‐up of experiments. Additionally, we consider
our study among six schools as a pilot for a project involving a larger number of schools.
Finally, an interesting future experiment would be to investigate whether
different researchers, who are also experienced in secondary school practice, would
make comparable matches of a given question with existing research publications.
![Page 179: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/179.jpg)
5 A tailor made transfer of scientific knowledge to school practice
153
5.7 Appendices
Appendix 5‐1: General characteristics of the six participating schools.
School Tracks Extra Profile Number of pupils School year 2010/11
Commitment
A Senior general, Pre‐ university
Extra opportunities for migrant pupils in Years 1 and 2; Assisted homework, extra reading and maths
495, substantial growth
Dedicated to supportingpupils who will be the first in their (migrant)family to continue at university level
B Pre‐ vocational, Senior general, Pre‐ university
Combined comprehensive Years 1 and 2, to promote learning between all pupils
1615, stable Balancing personal growth and high learning goals for a very diverse student body.
C Pre‐ vocational, Senior general, Pre‐ university, Gymnasium
Dalton school,extra opportunities in the Arts, Professional theatre and music studios
1205, growing Pupils learn to take own responsibility for learning. Partially tailor‐made curriculum according to pupils' choice.
D Pre‐vocational, Senior general, Pre‐ university, Gymnasium
Sports class (top sport potentials) Art & culture class; Technasium; Bi‐lingual track in English and Dutch
1765, growing To be a strong context‐oriented school and serve all pupils who live around the school.
E Pre‐ vocational, Senior general, Pre‐ university
Innovative school, using evidence‐based research for the design of an interactive learning community
353, newly‐founded school, growing fast
Creating a community of learners in open interaction with the school's context and society at large
F Pre‐ vocational, Senior general, Pre‐ university, Gymnasium
Bi‐lingual track in English and Dutch.; Cooperation with primary schools for high potential pupils
902, growing Committed to stimulatingupward mobility to a higher track for a majority of pupils with a migrant background
Notes: For ISCED qualifications see footnote 3. Gymnasium is a pre‐university track with extra courses in the classical languages Greek and Latin. Dalton is a didactical concept based on the ideas of Helen Parkhurst.
Appendix 5‐2: Format of the first interview
Before the actual interview started, the interviewees were informed about the
purpose of the study: to identify core questions concerning the school's performance,
![Page 180: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/180.jpg)
154
that are regarded to be school or context specific. Secondly it was explained that the
questions would be the starting point for a review of the relevant academic literature;
the quality standards for the selected articles were clarified, notably the requirement
that the research findings could be translated to school practice or strategic
developments. The interviewee was informed that we would provide an answer by
letter to the raised questions consisting of no more than six pages. Finally, the
evaluating interview was announced, three months after the letter had been received. In
Schools A, B and E an extra meeting, shortly after the letter was sent, was planned. All
interviews were recorded with permission. The confidentiality of the discussion was
guaranteed.
Phase 1
For the record, a brief overview of the school's administrative school
characteristics was asked (number of pupils, staff, offered secondary tracks, extra
programmes, composition of the student body).
How would you describe, in ways that are meaningful to you, the school context?
Could you describe where your pupils come from, how many live close the
school, how many pupils does the school attract from other city districts or from
other towns?
Could you describe which context, staff, or pupil characteristics have influenced
school policies?
Phase 2
Could you elucidate how context and pupil characteristics have been translated
to specific, demonstrable school practice and strategies?
What kind of knowledge, data and information supported these decisions?
Were stakeholders (e.g. parents, the governing board, local political government)
in any way involved in these discussions?
Does the school have a special didactical concept, a specific mission, or motto?
Phase 3
Did the characteristics of the school and its context that we discussed, lead to
specific views on your role as a school leader.
![Page 181: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/181.jpg)
5 A tailor made transfer of scientific knowledge to school practice
155
With reference to school quality improvement as a continuous process, what are
the most significant themes, questions or obstacles on your agenda? Could you
select those themes that are within your span of control as a Principal, but that
require, for example, because of their complexity, or because these are highly
specific concerns for your school, additional knowledge and insights. What, in
your view, makes it difficult to acquire appropriate knowledge?
Could you select three to five questions that you consider to be crucial?
The interview was concluded with the question whether the Principal had missed
something in the dialogue, or wanted to add information. At the end of the
interview the time schedule was briefly repeated.
Appendix 5‐3: Questions for the evaluating interview
1. Could you describe how you experienced this research project (associations,
images, metaphors)?
2. Did the matches with academic knowledge influence your thinking or your work?
Could you describe if and how you could apply this knowledge? Did you discuss
the matches of scientific outcomes with the management team, teachers, or
external parties?
3. When you actually did use the information that was offered, and in what way did
you use this? a) Direct practical application, b) conceptual thinking, c) influence
on current school processes, d) used in discussions with external parties.
4. In interview 1 we started with your reflections on the school's context and how
these context characteristics lead to school strategies and practices. In this final
interview we ask the mirror of this initial question: did the presented matches
with research influence your views on the school's context?
5. Would it be worthwhile to scale up this method? If so, do you have critical
remarks and suggestions? Could you describe your views on the characteristics of
an intermediate role between academic research and schools?
![Page 182: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/182.jpg)
156
Appendix 5‐4: Questions and matches per school
School A
Q1 Is it an advantage for pupils to be enrolled in a school that is specialized in upward
mobility, albeit segregated, or would pupils be better prepared for a full and
meaningful participation in Dutch society if they were to attend a desegregated
school.
A1 Banks (2007), Cooper (2011), Dobbie and Fryer (2009), Parker, Ninomiya and Cogan
(1999).
Q2 Should recruitment policies aim at hiring more teachers with the same background
as pupils? Would teachers with the same migrant background enhance the learning
and identity building of migrant pupils?
A2 Dee (2004).
Q3 Which examples could inspire the school in translating specific school
characteristics to class room practice?
A3 The Centre for Multicultural Education of the University of Washington was
suggested as a source for illustrative programmes and academic publications on
multicultural learning, (http://education.washington.edu/cme/). Furthermore, the
Stanford Centre for Opportunity in Education (http://edpolicy.stanford.edu) and the
Dutch "Echo" programme with university students acting as mentor for secondary
school pupils (www.echo‐net.nl/) were advised as examples of interesting networks
and good practice.
School B
Q1 Is postponed tracking chiefly benefitting pupils with a non‐academic family
background, or is comprehensive learning benefitting all pupils? Related to this,
how can the best balance be created between cognition on the one hand‐ and non‐
cognitive skills and personal growth on the other hand, notably for low SES pupils?
A1 Hanushek and Woessman (2006), Payne (2010), Van Elk et al. (2011).
Q2 In the Principal's view, the development of academic learning skills, excellent
language skills, and academic reasoning is an especially important mission for a
secondary school with a large number of pupils with a non‐academic background.
![Page 183: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/183.jpg)
5 A tailor made transfer of scientific knowledge to school practice
157
Does research support this view, has research been carried out on the effects of
different classroom practices for academic learning strategies?
A2 Brunello and Checchi (2007), Crul and Schneider (2009), Zimmerman (2002), Zohar
and Dori (2003).
Q3 This third question was raised by all other interviewees: the school would be very
interested in longitudinal data on study success and access to the labour market of
its own former students; this was considered to be crucial information for
evaluating the quality of teaching, didactical practices, and strategic choice.
A3 Privacy laws and regulation severely limit this type of school based information on
individual former students. Since all six Principals brought up this issue, this
question, obviously, will be followed up by further exploration of options with the
Ministry of Education.
School C
Q1 Has research been carried out that demonstrates that the school's specific
didactical system (Dalton), and the emphasis on the Arts, can potentially reinforce
the emancipation and upward mobility of underrepresented groups in higher
education and the higher strata of the labour market?
A1 In the present study, this turned out to be a difficult question to answer precisely.
We choose to take as a proxy for the school's specific didactical system, the
example of magnet schools (Notten, 2009). Furthermore, the aforementioned
Center for Multicultural education at the University of Washington provides
research findings on the importance of cultural education and the arts. The OECD,
in its publication Education at a Glance (OECD, 2011), finally, argues that the
success of the Finnish system may be explained by, among other indicators, the
combination of comprehensive education until age 16 and flexible ability grouping.
This combination seems related to school C's model.
Q2 The second question is also related to the school's didactical model: Does research
offer insights into organizational development models in which teachers take a
major role in exploring, implementing and monitoring innovative teaching
methods?
A2 Ancess (2000), van der Grift (2010).
![Page 184: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/184.jpg)
158
Q3 Third, the Principal was interested in research on the effects of grade repetition and
examples of successful other practices to respond to the learning tempo of pupils.
A3 (Välijärvi and Sahlberg, 2008).
School D
Q1 The director was particularly interested in a firm research base for postponed
tracking, in the context of learning in diverse groups.
A1 Hanushek and Woessman (2006), Pekkarinen, Uusitalo and Kerr (2009)Van Elk, Van
der Steeg and Webbink (2011).
Q2 Considering that most pupils at School D have a migrant background, the director
was concerned about the effect of prejudice and stigmatizing that pupils may
encounter. He was especially interested in research on the long‐term effects of
stigmatization.
A2 Banks (2007), Bunar (2010).
Q3 Suppose the school had extra financial resources, what would, according to
scientific research, be the most effective intervention for further quality
development?
A3 It turned out to be very complex to answer this question on the allocation of extra
resources specifically for this particular school. Academic literature puts a focus,
however, on the quality of teachers as paramount: Brown (2007), Jackson and
Bruegmann (2009), McKinsey & Company (2007).
Q4 The school is well equipped for computer supported learning: What are the best
research‐based practices that could guide the school in the optimal use of e‐
learning?
A4 Ala‐Mutka, Punie and Redecker (2008), Cheung and Slavin (2011).
Q5 Which studies can further inform the school about effective strategies for
innovating the organization of learning, notably interactive learning with the
school's environment?
A5 Ancess (2000), Nilsson (2008).
![Page 185: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/185.jpg)
5 A tailor made transfer of scientific knowledge to school practice
159
School E
Q1 The main questions concerned the balance between innovative learning, while
maintaining the high cognitive standards of the national final exam, and,
furthermore, a research base for interactive learning with the environment.
A1 Heckman and Montera (2009), Payne (2010).
Q2 Secondly, the Principal was interested in a knowledge base for effective leadership
in innovative schools
A2 Sanders and Harvey (2002).
Q3 Since school E is organized in smaller unities with sub teams of teachers, more
insight is required in best evidence for effective teacher cooperation and training.
A3 Brown (2007), Jackson and Bruegmann (2009), McKinsey (2007), Saunders,
Goldenberg and Gallimore (2009).
Q4 The hiring of new teachers is an important strategic advantage of a newly founded
school with a constantly growing student body; what teacher qualities correlate
positively to learning outcomes?
A4 Boyd, Grossman, Ing, Lankford and Wyckoff (2009), Darling‐Hammond and Youngs
(2002), Hanushek and Rivkin (2007), Van der Grift (2010).
School F
Q1 The Principal inquired after illustrative extra programmes at other schools,
additional to the standard curriculum, aimed at creating extra opportunities for
pupils, like the example of magnet schools in the U.S.
A1 Tefera, Frankenberg, Siegel‐Hawley and Chirichigno (2011).
Q2 Research convincingly demonstrates that no other factor has such a large effect on
learning outcomes as teacher quality. In School F, where teachers constantly work
on the upward mobility of their students, this may be even more important. What
elements of teacher quality matter especially, and what research skills should
teachers develop in order to effectively monitor the effects of their teaching
methods?
A2 Ancess (2000), Darling‐Hammond and Youngs (2002) Hanushek and Rivkin (2007).
![Page 186: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/186.jpg)
160
Q3 Furthermore, a question is raised on the relatively under‐used opportunity in Dutch
education to use the influence and authority of the Principal in selecting personnel,
and his or her role in stimulating effective cooperation among teachers.
A3 Boyd, Grossman, Ing, M., Lankford, H., Wyckoff, J. (2009).
![Page 187: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/187.jpg)
161
6 Conclusions and Discussion
I know that I am human and may have erred. However, I have
taken great pains not to err, and to ensure above all that whatever
I have written should be entirely consistent with the laws of the
land, with piety, and with morality.
Baruch de Spinoza, 1670
6.1 Conclusions
This thesis has explored success in secondary education among pupils with a
migrant background in the four major cities in the Netherlands. This quest commenced
from four different points of departure: ethnic segregation of the pupils' area of
residence, the stability of school careers, patterns of school choice, and the transfer of
relevant and practicable research outcomes to schools with many migrant pupils.
Local and national data sets, different statistical techniques, and case study
research based on in‐depth interviews have been used in an effort to investigate the
achievements of migrant pupils in a multidisciplinary way. In this last chapter, the
findings and conclusions from the previous chapters are combined and discussed as
concluding research findings that may imply suggestions for policies at the national,
![Page 188: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/188.jpg)
162
municipal, and school level aimed at speeding‐up the closing of the educational
achievements gap between migrant pupils and native Dutch pupils.
The association between the percentage of 16 year‐old pupils of Moroccan
descent at zip‐code level and enrolments in secondary academic tracks is negative.
In this first study, to set the scene, we took a closer look at the widespread, yet
not clearly underpinned notion in public parlance and political debate, that
neighbourhood segregation hinders migrant pupils to find their way to higher strata of
secondary education. The literature on the negative effects of segregation, notably
considering large metropolitan areas in the U.S., describes a strong negative effect on
educational outcomes. Other literature, however, reports the beneficial effects of
closely‐knitted migrant communities and neighbourhood schools.
Instead of considering the ethnic composition of the total population at zip‐code
level, we chose to take a different approach: we measured specifically the share of 16
year‐old pupils of Moroccan descent among all 16 year‐olds living in the same area. We
decided on this option because national statistics point at very large increases in average
attainment levels between immigrant generations of grandparents, parents and the
current young generation. For this reason, we wanted to observe, as closely as possible,
this youngest generation separately. We used bivariate linear (OLS) and LOWESS
regressions to consider the percentage of 16 year‐old Moroccan second generation
youth at zip‐code level, and their actual enrolment in academic secondary tracks.
Overall, this association is negative: for every 10% increase of Moroccan 16 year‐olds,
there is a 2% estimated decrease in enrolments in academic tracks. The data set we used
for this study did not cover other socio‐economic variables besides ethnicity; therefore,
we could not interpret further the high dispersion of data points we found that
suggested that factors other than ethnic neighbourhood composition and ethnicity
affect school success.
Switching between secondary schools burdens pupils; it reduces the odds of
obtaining a diploma or move up to a higher track, and increases the odds of later
dropout. Moroccan pupils living in three disadvantaged zip areas switch schools most
frequently.
![Page 189: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/189.jpg)
6 Conclusions and Discussion
163
We started this second project with an exploration of a very detailed municipal
data set (The Amsterdam Social Development Office, DMO), without formulating specific
questions in advance, with the objective to detect any patterns that might point to
phenomena that may have been overlooked so far. The data set covers the educational
history of 16‐22 year‐olds in three disadvantaged zip areas in Amsterdam. Already, a first
visual inspection of the data files raised questions on the remarkably high number of re‐
enrolments in secondary schools for considerable numbers of pupils.
The literature unequivocally illustrates, and presents causal evidence, that all
switching– even for positive reasons– is harmful for pupils, and even more harmful for
low SES pupils. This harmful effect is attributed to the disruptive effect of, for example,
losing classmates, familiar teachers, and a curriculum the pupil is used to. In the first
part of this research we measured switching in straight counts of schools attended per
pupil– measured after they had exited secondary school– and the relationship to school
success (diploma) or dropout. Our findings point at the same negative effect as the
literature predicts: migrant students switch more often, and repeated switching is
progressively linked to a considerable decrease in the percentage of pupils who obtain a
diploma, and increasing dropout. Motivated by these findings, we repeated our
measurements with the national educational data source BRON at a different point in
time in the pupils' school career (secondary Year 3) and on the larger scale of
Amsterdam. In this second part we looked specifically at the connection between
switching and upward or downward track changes. We used a dummy‐coded variable
for switching here, and did not count the number of schools attended. In the three
socio‐economically‐challenged areas all pupils switch more often, except for native
Dutch pupils. We are somewhat puzzled by our findings that in the case of all groups,
but more so among Moroccan pupils, a share of switchers move to another school
within the same cluster; we do not know whether this type of switching is associated
with secondary track change, or that pupils with behavioural problems more often
migrate within a cluster. Additionally, the vast majority of switchers continue in another
school at the same track level. This raises the same question about the motive for school
change. Especially in the case of migrant pupils, upward mobility to a higher secondary
track is an important extra opportunity to reach the level that best fits their capacities.
Staying enrolled in the same school, is convincingly connected to higher percentages of
![Page 190: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/190.jpg)
164
upward mobility; considering pupils with a Moroccan background, almost one in every
five pupils moves up to a higher track in the same school (in Amsterdam), while this is
only 7.6% after switching.
88.7% of pupils living in the four largest cities exercise the right to free school
choice and do not select the nearest school. Migrant pupils on average prefer a school
with more migrant students than the nearest school; native Dutch pupils show the
opposite preference. Unexpectedly, the odds of moving up to a higher track are
slightly lower for Moroccan pupils who select another school than the one closest to
their residence. SES, rather than ethnicity, drives school choice.
In the literature about the unequal opportunities for low SES pupils to find access
to high quality schools, much is expected of extended options for school choice. In the
Netherlands this choice is completely free, there are no financial barriers of any sort,
and the Inspectorate assesses the sufficient quality of all schools, although there may be
quality differences above this basic quality level. This situation provides interesting
opportunities to analyse which factors drive school choice in this freely accessible and
transparent school market (all school quality reports by the Inspectorate are on the
Internet). We were permitted to use the latest updated version of the rich educational
data source BRON, and could merge these data with the socio‐economic characteristics
of neighbourhoods, and with school quality standards by the Inspectorate. As a proxy for
the selectivity of choice, we used two distance measures: the distance from home to
school, and the difference in distance between the nearest and the preferred school. We
used multi‐level regressions analyses (OLS), also in the case of the dummy‐coded
variable nearest/non‐nearest school, and the distance difference (always ≥ 0); we also
carried out both a Logit and Tobit regression, but preferred the OLS regression because
we found no disturbing differences between the two techniques, and decided against
the complex interpretation of odds ratios in the case of the Logit and Tobit regressions.
We found that migrant pupils on average travel smaller distances to school than
native Dutch students; this may correspond, however, with the fact that more migrant
pupils live in poorer, more densely populated areas, with more schools. Neighbourhood
SES has a two‐way effect on distance to school: in low SES areas, migrant pupils travel
shorter distances, but native Dutch pupils tend to escape the area and travel further to
![Page 191: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/191.jpg)
6 Conclusions and Discussion
165
school. In contrast, in high SES areas, native Dutch pupils tend to travel a small average
distance and more often choose the nearest school with a population that mirrors the
neighbourhood population. Whereas we found that a higher school average SES attracts
native Dutch pupils, somewhat unexpectedly, we found hardly any association between
distances travelled and school quality indicators, like the average exam score and the
school's percentage of upward mobility to a higher track. We were surprised by our
finding that migrant pupils tend to prefer a school further away from their area of
residence, which had an even higher percentage of migrant pupils than the nearest
school. Moreover, in their case, a more selective choice to a school at a greater distance
was only slightly associated with a lower percentage of pupil upward mobility to a higher
track. Additionally, our analyses demonstrate that migrant pupils on average attend a
school with a higher mean percentage of upward mobility than the schools chosen by
native Dutch pupils. However, migrant pupils themselves move up less often to an
academic track.
Secondary school Principals have strategic and complex conceptual questions
that can be matched with the high‐quality practicable outcomes of academic research.
Principals express the view that intermediaries who know the reality of schools, and
are also academic researchers, could bridge the gap between secondary schools and
academe.
Overall, the literature on the applicability of scientific knowledge for school
practice demonstrates that this transfer is rather problematic. Complex issues arise,
considering for example the different types of knowledge –academics may prefer to
isolate features in education, while schools consider the complexity of everyday school
life– and the difficulties school leaders may encounter in translating research findings to
concrete classroom practice.
We made an attempt to study in detail, together with six participating school
leaders of secondary schools with a largely diverse student body, the requirements for
an accurate match between existing research and school questions. We used the
method of in‐depth, semi‐structured interviews to identify school questions, searched
for a match with recent, high‐quality, practicable research publications, and carried out
![Page 192: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/192.jpg)
166
a first evaluation of the actual use of the suggested research in practice, after three
months– admittedly a short period of time.
All but three out of 21 questions could be linked to high quality research
outcomes; we found that most questions were of a conceptual nature, considering the
best strategies for the school amidst its surrounding context of a multi‐ethnic society (all
involved schools were located in Amsterdam). The Principals emphasized that they
repeatedly had repeatedly encountered the problem that scientific research may report
conflicting findings related to the same issue, for example on the subject of school
segregation: would desegregating schools be the preferred option? Or would schools
with a majority of migrant pupils be a better choice, since such school can specialize in
curriculum content (e.g. special attention to language proficiency) and the counselling of
migrant pupils (e.g. repeated formative assessments to enhance upward mobility to
higher tracks)? After three months, five out of the six school leaders expressed the view
that they had made use of the suggested research findings, notably considering strategic
decisions, and communications with the School Board and external parties (for example,
the municipality).
We encountered, however, the rather serious technical problem that only a
limited number of research publications are freely accessible via academic search
machines: in many cases a licence fee has to be paid to a scientific journal in order to get
access to full text reading; such fees can only be paid by universities (or other large
research institutions), and not by single schools.
Principals would welcome an intermediate function between schools and
academe, carried out by people who are well‐experienced in both worlds.
6.2 Policy Implications
6.2.1 A different perspective on segregation
The findings presented in this thesis imply considerations for policies at the
government, municipal and school level. Considering the societal discourse on the
segregation of schools and neighbourhoods, our results show a relatively small, but
significantly negative association between the percentage of 16 year‐old migrant pupils
at zip‐code level and their enrolments in the two highest academic secondary tracks
(Chapter 2). However, we also found a similarly small, but significant positive association
![Page 193: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/193.jpg)
6 Conclusions and Discussion
167
between upward mobility to a higher track and choosing a school closer to the home
address (Chapter 4). In combination, these findings do not convincingly support policies
to desegregate schools by centralized school assignments, and suggest instead that
investing in increasing the chances for upward mobility in schools located in migrant
neighbourhoods, may be worth considering. Interestingly, our analyses of patterns of
school choice reveal that migrant parents and pupils prefer to travel further to a school
with an even larger percentage of migrant pupils than the school closest to their home.
Finally, our investigation of the usability of academic knowledge in the case of six
secondary schools with a diverse student body shows that the Principals of these schools
are strongly committed to using the best scientific knowledge basis for the specific
function of their school vis‐à‐vis the migrant communities they serve (Chapter 5).
Together, the outcomes of the different studies in this thesis seem to make a case for
high‐quality, specialized schools that are well connected to migrant communities in the
surrounding residential areas.
6.2.2 Stable school careers
Although we encountered a rather serious problem concerning school switching,
in retrospect we think that our choice to start with a visual observation of the
Amsterdam municipal data set, without any research question in advance, turned out to
be an interesting strategy. We found that school switching, which according to the
literature, is in all cases harmful to students, occurs more often among migrant pupils
who live in socio‐economically disadvantaged neighbourhoods. Furthermore, we found
that switching reduces the odds of upward mobility to a higher track, an important
instrument for the narrowing of the educational achievement gap between migrant and
native Dutch youth. The reduction of switching, contrary to other factors that are
connected to the achievement gap (e.g. the attainment levels of parents) seems a
manageable strategy. It may be considered, at the level of the government and the
Inspectorate, that school attainment levels should be mentioned more prominently in
public reports on the quality of all schools, and inform migrant parents in particular,
about the importance of this indicator. At the municipal level, the reduction of switching
could be linked to large‐scale programmes to prevent dropout, since switching is
![Page 194: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/194.jpg)
168
strongly related to later dropout. Especially those schools that cooperate in clusters
under the same board can relatively easily decide on policies that may reduce switching.
6.2.3 Validation of academic knowledge
A better insight into school questions, making the link to manageable research
findings, and considering usability in school practice, resulted in concrete
recommendations by school leaders for an intermediate function between schools and
academe. Therefore cooperation with an institute for school leadership is already being
currently explored. At the same time, at the national level new initiatives are also being
developed concerning the knowledge base for educational policies. Both developments
may benefit from the mutual exchange of goals and strategies.
6.2.4 The crucial importance of data
Considering that computing power is no longer a limiting factor, we hope that
our work based on the national database BRON, as the Amsterdam municipal database,
may also bring further aspirations to make national, local, and school data sets
compatible. Schools create many interesting data files themselves, but may need
(financial) support to further develop the architecture of their database and invest in
professional skills to align their data to national data analyses. In this way schools could
specifically monitor the performance indicators they consider to be relevant for decision
making.
The BRON data set is being developed in the first place, for the analysis of
national costs and the outcomes of the educational system. The data set offers,
however, enormous new opportunities for academic research that may add to not only
measuring but also understanding the complexity of everyday school processes, with the
goal to enhance equal opportunities for all pupils.
6.3 Limitations and further research
However much I would have liked to think otherwise during the long and at times
almost desperate journey towards a doctorate, finishing a thesis is a mere debut in the
world of academic research, not a completion.
There are obvious limitations to the studies that have been described in this
thesis, and, more important, this work has generated new questions for further
![Page 195: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/195.jpg)
6 Conclusions and Discussion
169
research. First, the scale of the research carried out may have been a limitation: the
study into switching resulted in promising findings that could be scaled up to the four
major cities in the Netherlands. Furthermore, the reduction of switching offers
possibilities for an experimental research design, which would involve comparing
schools with well ‐defined practices to reduce switching and other schools.
Second, the nature of the explored research questions may require follow‐up
research: analyses of distance to school as a proxy for the selectivity of school choice,
have raised further questions , for example concerning the preference of migrant pupils
and their parents for a school with a high percentage of migrant pupils: Do these pupils
expect that schools with a vast majority of migrant pupils might offer specialized
learning opportunities for migrant pupils, or might it be the case that parents prefer for
their children to stay in the same culture as at home, or even avoid stigmatization in a
school where their child might belong to a minority?
And third, in the case of the study into the transfer of scientific research
outcomes to school practice, time was a limiting factor: we evaluated after only three
months whether the Principals had made use of insights from academic research,
whereas the search for the effects on school processes of a more solid knowledge base
for school practice, requires monitoring over far longer periods of time.
![Page 196: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/196.jpg)
![Page 197: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/197.jpg)
171
References
Ala‐Mutka, K., Punie, P., Redecker, C. (2008). Digital Competences for Lifelong Learning
(Report). European Commission.
Allen, R. (2007). Allocating pupils to their nearest secondary school: The consequences
for social and ability stratification. Urban Studies, 44 (4), pp. 751‐770.
Ancess, J. (2000). The reciprocal influence of teacher learning, teaching practice, school
restructuring, and student learning outcomes. Teachers College Record, 102 (3),
pp. 590‐619.
Andersson, E., Malmberg, B., Östh, J. (2012). Travel‐to‐school distances in Sweden 2000‐
2006: changing school geography with quality implications. Journal of Transport
Geography, pp. 35‐43.
Anger, S., Heineck, G. (2009). Do smart parents raise smart children? The
intergenerational transmission of cognitive skills. Journal of Population
Economics 23 (3), pp. 1105‐1132.
Archbald, D.A. (2004). School choice, Magnet schools, and the liberation model: An
empirical study. Sociology of Education, 77 (238).
Bakker, J. (2012). Cultureel‐etnische segregatie in het Onderwijs: achtergronden,
oorzaken en waarom te bestrijden? Landelijk Kenniscentrum Gemengde Scholen.
Banks, J.A. (2007). Educating citizens in a multicultural society. New York: Teachers
College Press.
![Page 198: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/198.jpg)
172
Banks, J.A. (Ed.). (2009). The Routledge International Companion to Multicultural
Education. New York, NY: Routledge.
Bauder, H. (2002). Neighbourhood effects and cultural exclusion. Urban Studies, 39 (1).
Bauer, P., Riphahn, R.T. (2006). Timing of school tracking as a determinant of
intergenerational transmission of education. Economics Letters, 91 (1), pp. 90‐97.
Baysu, G., de Valk, H. (2012). Navigating the school system in Sweden, Belgium, Austria
and Germany: school segregation and second generation school trajectories.
Ethnicities, 20 July 2012. Retrieved from DOI: 10.1177/1468796812450857.
Belfield, C.R., Levin, H.M. (Eds). (2007). The price we pay. Economic and social
consequences of inadequate education. Washington D.C.: Brookings Institution
Press.
Björklund, A., Salvanes, K.G. (2010). Education and family background: Mechanisms and
policies. IZA (Institute for the Study of Labour) Discussion paper 5002.
Boyd, D., Grossman, P., Ing, M., Lankford, H., Wyckoff, J. (2009). The influence of school
administration on teacher retention decisions. Retrieved August 31, 2011 from
http://www.stanford.edu/.
Brown, M. R. (2007). Educating all students: creating culturally responsive teachers,
classrooms, and schools. Intervention in School and Clinic, 43 (1), pp. 57‐62.
Brown, K., Benkovitz, J., Muttillo, A.J., Urban, T. (2011). Leading schools of excellence
and equity: documenting effective strategies in closing the gap. Teachers College
Record, 113, No. 1, pp. 57‐96.
Brunello, G., Checchi, D. (2007). School tracking and equality of opportunity. The Journal
of Economic Policy. pp. 781‐861.
Bunar, N. (2010). Multicultural schools in Sweden and their communities: social
predicaments, the power of stigma, and relational dilemmas. Urban Education,
46 (2), pp. 141‐164.
Bunar, N. (2010). The geographies of education and relationships in a multicultural city:
enrolling in high‐poverty, low‐performing urban schools and choosing to stay
there. Nordic Sociological Association: Acta Sociologica.
Cabrera, A.F., La Nasa, S.M. (2001). On the path to college: Three critical tasks facing
America's disadvantaged. Research in Higher Education, 42 (2), pp. 119‐149.
![Page 199: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/199.jpg)
References
173
Cheung, C.K., Slavin, R.E. (2011). The effectiveness of education technology for
enhancing reading achievement: a meta‐analysis. www.bestevidence.org.
Coleman, J.S., Campbell, E.Q., Hobson, C.J., McPartland, J., Mood, A.M., Weinfeld, F.D.,
York, R.L. (1966). Equality of educational opportunity. Washington DC:
Government Printing Office.
Commissie Nationaal Plan Toekomst Onderwijswetenschappen [Committee for a
National Programme for the Future of Educational Sciences] (2011). Nationaal
Plan Onderwijs/Leerwetenschappen. Report retrieved from www.vsnu.nl
Cooper, C. R. (2011). Bridging multiple worlds. Cultures, identities, and pathways to
college. New York: Oxford University Press.
Crul, M., Holdaway, J. (2009). Children of immigrants in schools in New York and
Amsterdam: The factors shaping attainment. Teachers College Record, 111, No.
6, pp. 1476‐1507.
Crul, M., Schneider, J. (2009). Children of Turkish Immigrants in Germany and the
Netherlands: The impact of differences in vocational and academic tracking
systems. Teachers College Record, 111, No. 6, pp. 1508‐ 1527.
Crul, M., Schneider, J., Lelie, F. (Eds). (2012). The European second generation compared.
Does the integration context matter? Amsterdam: Amsterdam University Press.
Cullen, J.B., Jacob, B., Levitt, S. (2000). The Impact of School choice on Student
Outcomes. An Analysis of the Chicago Public Schools. NBER working paper 7888.
Cambridge, MA: National Bureau of Economic Research.
Darling‐Hammond, L., Youngs, P. (2002). Defining "highly qualified teachers": what does
"scientifically‐based research" actually tell us? Educational Researcher, 31 (9), pp.
13‐25.
De Witte, K. (2010). Dropout from secondary education: All's well that begins well. TIER
working paper series 10/01. Retrieved from http://www.tierweb.nl
Dee, T. S. (2004). Teacher, race, and student achievement in a randomized experiment.
The Review of Economics and Statistics, 86(1): 195‐201.
Dobbie, W., Fryer, R.G. (2009). Are High Quality Schools Enough to Close the
Achievement Gap? Evidence from a Social Experiment in Harlem. (NBER Working
Paper 15473). Retrieved from National Bureau of Economic Research website:
http://www.nber.org.papers/w15473
![Page 200: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/200.jpg)
174
Dronkers, J. (2010). Positive but also negative effects of ethnic diversity in schools on
education performance? An empirical test using cross‐national PISA data.
Inaugural Lecture at Maastricht University. Maastricht, the Netherlands: Océ
Business Services.
Durlauf, S.N. (2004). Neighbourhood Effects. In Henderson, J.V., & Thisse, J.F. (Eds).
(2004). Handbook of regional and Urban Economics, Volume 4. Amsterdam, the
Netherlands: Elsevier.
Education Council of the Netherlands (2011). Ruim baan voor stapsgewijze verbetering
[Make way for gradual improvement]. The Hague: Onderwijsraad.
Elk, R. van, Steeg, M.van der, Webbink, D. (2011). Does the timing of tracking affect
higher education completion? Economics of Education Review, 30, 1009‐1021.
Furlong, J. (2010). BERA at 30. Have we come of age? British Educational research
Journal, 30:3, pp. 343‐358.
Gevers, J.K.M. (1996). Over iets te ingewikkelds. Presidential Lecture at the opening of
the Academic Year 1996 at the University of Amsterdam. Amsterdam:
Vossiuspers AUP.
Gibbons, S., Machin, S., Silva, O. (2008). Choice, competition, and pupil achievement.
Journal of the European Economic Association, 6 (4), pp. 912‐947.
Gijsberts, M., Vervoort, M., Havekes, E. and Dagevos, J. (2010). Maakt de buurt verschil?
(Does neighbourhood make the difference). The Netherlands Institute for Social
Research. The Hague: Textcetera.
Glaeser, E.L., Resseger, M.G., Tobio, K. (2008). Urban Inequality. (NBER Working Paper
14419). Retrieved from National Bureau of Economic Research website:
http://www.nber.org.papers/w14419.
Gunderson, L.G., Holling, C.S. (2002). Panarchy: Understanding transformations in
human and natural systems. Washington D.C: Island Press.
Grift, W.J.C.M. van der (2010). Ontwikkeling in de beroepsvaardigheden van leraren.
Rijksuniversiteit Groningen, Oratie.
Groot, W., Maassen van den Brink, H. (2003). Investeren en terugverdienen. The Hague:
Sectorbestuur Onderwijsarbeidsmarkt.
![Page 201: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/201.jpg)
References
175
Hanushek, E. A., Rivkin, S.G. (2007). Pay, working conditions, and teacher quality. The
Future of Children, 17 (1), pp. 69‐86. Princeton University.
Hanushek, E.A., Kain, J.F., & Rivkin, S.G. (2001). Disruption versus Tiebout improvement:
the costs and benefits of switching schools. (NBER working paper 8479).
Retrieved from National Bureau of Economic Research website:
http://www.nber.org/papers/w8479
Hanushek, E., Woessman, L. (2006). Does educational tracking affect performance and
inequality? Difference‐in‐difference evidence across countries. The Economic
Journal, 116 (March), C63‐C76.
Harris, R., Johnston, R. and Burgess, S. (2007). Neighbourhoods, ethnicity and school
choice: Developing a statistical framework for geodemographic analysis.
Population Research and Policy Review, 26 (6/6), pp. 553‐579.
Hartog, J., Zorlu, A. (2009). Ethnic segregation in the Netherlands: analysis at
neighbourhood level. International Journal of Manpower, 30, (1/2), pp. 15‐25.
Heckman, J. J. (2008). Schools, Skills, and Synapses. (NBER Working Paper 14064).
Retrieved from National Bureau of Economic Research website:
http://www.nber.org.papers/w14064.
Heckman, P.E., Montera V.L. (2009). School reform: the flatworm in a flat world: from
entropy to renewal through indigenous invention. Teachers College Record, 111
(5), pp. 1328‐1351.
Heineck, G., Riphahn, R.T. (2007). Intergenerational transmission of educational
attainment in Germany: The last five decades. Institute for the Study of Labour,
Discussion Paper no. 2985. Retrieved from http://ssrn.com/abstract=1012343.
Herweijer, L. (2009). Making up the gap: Migrant education in the Netherlands. The
Netherlands Institute for Social Research. The Hague: Textcetera.
Heus, M. de, Dronkers, J. (2008). Can Educational Systems Compensate for Societal
Features? The Effects of Educational Systems and Societal Features of both
Countries of Origin and Destination on the Scientific Literacy of Immigrant
Children in Western Countries. Paper presented at the International Conference
on Survey Methods in Multinational, Multiregional, and Multicultural Context.
Berlin 2008.
![Page 202: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/202.jpg)
176
Ho Sui‐Chu, E., Willms, J.D. (1996). Effects of parental involvement on eighth‐grade
achievement. Sociology of Education, 69 (2), pp. 126‐141.
Inspectorate of Education (2011). Onderadvisering van allochtone leerlingen? [under‐
advising of migrant students?]. Report.
Jackson, K. C., Bruegmann, E. (2009). Teaching students and teaching each other: the
importance of peer learning for teachers. NBER Working Paper Series, paper
15202. Cambridge, MA: National Bureau of Economic Research.
Johnson, R. B. (1998). Toward a theoretical model of evaluation utilization. Education
and Programme Planning, 21, pp. 93‐110.
Kerbow, D., Azcoitia, C., Buell, B. (2003). Student mobility and local school improvement
in Chicago. The Journal of Negro Education, 72, No. 1, pp. 158‐ 164.
Konstantopoulos, S., Borman, G. (2011). Family background and school effects on
student achievement: a multilevel analysis of the Coleman data. Teachers College
Record, 113, No. 1, pp. 97‐132.
Ladd, H.F., Fiske, E., Ruijs, N. (2009). Parental choice in the Netherlands: Growing
concerns about segregation. Paper prepared for the National Conference on
School Choice, Vanderbilt University.
Ladd, H.F., Fiske, E.B., Ruijs, N. (2010). Parental Choice in the Netherlands: growing
concerns about segregation. Sanford Working Papers Series (SAN10‐02).
Landry, R., Amara, N., Lamari, M. (2001). Climbing the ladder of research utilization:
evidence from social science research. Science Communication. SAGE
Publications. Retrieved from http://scx.sagepub.com/content/22/4/396
Leigh McGarrigle, J. (2010). Understanding processes of ethnic concentration and
dispersal. South Asian residential preferences in Glasgow. Amsterdam, the
Netherlands: Amsterdam University Press.
Levin, H. M. (2011). The economics of education. Albany Government Law Review, 4,
pp.395‐425.
Louis, K., Wahlstrom, K.L., Michlin, M., Gordon, M., Thomas, E., Leithwood, K., Anderson
S. E., Mascall, B., Strauss, T., Moore, S. (2010). Learning from leadership:
Investigating the links to improved student learning. Final report of research to
the Wallace Foundation. University of Minnesota: Center for Applied Research
and Educational improvement. Retrieved from www.cehd.umn.edu/CAREI/
![Page 203: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/203.jpg)
References
177
Margolis, J. (2010). Why teacher quality is a local issue (and why Race to the Top is a
misguided flop). Teachers College Record, date published June 23, 2010.
Retrieved from http://www.tcrecord.org ID Number: 16023.
Maroulis, S., Gomez, L. M. (2008). Does "Connectedness" matter? Evidence from social
network analysis within small‐school reform. Teachers College Record, 110, pp.
1901‐1929.
Massey, D. (2001). The Prodigal Paradigm Returns: Ecology Comes Back to Sociology. In
Booth, A., Crouter, A.C. (Eds). (2008). Does it take a village? Hove, East Sussex:
Psychology Press.
May, S. (2009). Critical multiculturalism and education. In Banks, J.A. (Ed.). (2009). The
Routledge International Companion to Multicultural Education. New York, NY:
Routledge.
McIntyre, D. (2006). Bridging the gap between research and practice. Cambridge Journal
of Education, 35:3, pp. 357‐382.
McKinsey & Company (2007). How the World's Best‐performing School Systems Come
out on Top (Report).
McKinsey & Company (2009). The economic Impact of the Achievement Gap in America's
Schools. Australia: New Media.
Miller Marsh, M., & Turner‐Vorbeck, T. (Eds), (2010). (Mis)understanding families:
Learning from real families in our schools. New York, NY, Columbia University:
Teachers College Press.
Morrell, E., Noguera, P. (2011). A framework for change: a broader and bolder approach
to school reform. Teachers College Record, ID Number 16503.
Murnane, R.J. (2009). Educating urban children. In Inman, R.P (Ed.). Making cities work.
Prospects for urban America. Princeton NJ: Princeton University Press.
Nakagawa, K., Stafford, E., Fisher, A., Matthews, L. (2002). The "City Migrant" dilemma:
Building community at high‐mobility urban schools. Urban Education 37: 96.
Nilsson, M. E. (2008). University‐school collaboration based on complementary needs.
Journal for Educational Change, 9, pp. 349‐356.
Noguera, P.A. (2004). Social Capital and the Education of Immigrant Students:
Categories and Generalizations. Sociology of Education, 77 (2), pp. 180‐183.
![Page 204: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/204.jpg)
178
Notten, A. (2009). Desegregation in schools– How can it be achieved? Journal of Social
Intervention: Theory and Practice, 18 (4), pp. 83‐93.
Nutley, S., Walter, I., Davies, H. (2002). From knowing to doing: a framework for
understanding the evidence‐into‐practice agenda. Discussion paper. University of
St. Andrews: Research Unit for Research Utilization.
OECD (2010), PISA 2009 Results: What Students Know and Can Do – Student
Performance in Reading, Mathematics and Science (Volume I). Retrieved from
http://dx.doi.org/10.1787/9789264091450‐en
OECD (2010). OECD Reviews of Migrant Education: Netherlands. Paris: OECD
OECD (2010). Reviews of Migrant Education: Country Note for the Netherlands. Paris:
OECD.
OECD (2011). Education at a Glance. Paris: OECD.
Oosterbeek, H. (2001). Voortschrijdend Inzicht [Progressive Insight]. Inaugural lecture.
Amsterdam: Vossius Press.
Opdenakker, M.C., Van Damme, J. (2001). Relationship between school composition and
characteristics of school process and their effect on mathematics achievement.
British Education Journal, 27 (40), pp. 407‐432.
Oreopoulos, Ph. (2007). Neighbourhood Effects in Canada: A Critique. Paper presented
at the 2007 Statistics Canada Socio‐economic Conference.
Palardy, G.J. (2008). Differential effects among low, middle, and high social class
composition schools: a multiple group, multilevel latent growth curve analysis.
School Effectiveness and School Improvement, 19 (1), pp. 21‐49.
Parker, W.C., Ninomiya, A. and Cogan, J. (1999). Educating world citizens: toward
multinational curriculum development. American Educational Research Journal,
36:117.
Payne, C.M. (2010). So much reform, so little change: The persistence of failure in urban
schools. Cambridge, MA: Harvard Education Press.
Pekkarinen, T., Uusitalo, R. Kerr, S. (2009). School tracking and intergenerational income
mobility: evidence from Finnish comprehensive school reform. Journal of Public
Economics, 93, pp. 965‐973.
Rumberger, R. W., Larson, L. A. (1998). Student mobility and the increase risk of high
school dropout. American Journal of Education, 107, (1), pp. 1‐35.
![Page 205: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/205.jpg)
References
179
Rumberger, R.W., Palardy, G.J. (2005). Does segregation still matter? The impact of
student composition on academic achievement in high school. Teachers College
Record, 107, (9), pp. 1999‐2045.
Rumberger, R.W., Thomas, S.L. (2000). The distribution of dropout and turnover rates
among urban and suburban high schools. Sociology of Education, 73, No.1, pp.
39‐67.
Sanders, M. (2009). Collaborating for change: how an urban school district and a
community‐based organization support and sustain school, family, and
community partnerships. Teachers College Record, ID Number 15304.
Sanders, M. G., Harvey, A. (2002). Beyond the school walls: a case study of Principal
leadership for school‐community collaboration. Teachers College Record, 104 (7),
pp. 1345‐1368.
Saunders, W.M., Goldenberg, C. N., Gallimore, R. (2009). Increasing achievement by
focusing grade‐level teams on improving classroom learning: a prospective,
quasi‐experimental study of Title I schools. American Educational Research
Journal, 46 (4), pp. 1006‐1033.
Scheffer, P. (2000). Het culturele drama [The cultural drama]. NRC Handelsblad 29
January 2000.
Schön, D.A. (1983). The reflective practitioner: How professionals think in action. New
York: Basic Books.
Scott, B. (2001). Coming of age. Intercultural Development Research Association
Newsletter, pp. 1‐8.
Shewbridge, C., Kim, M., Wurzburg, G., Hostens, G. (2010). OECD Reviews of Migrant
Education: Netherlands. Paris, France: OECD.
Slavin, R.E. (2002). Evidence‐based education policies: transforming educational practice
and research. Educational Researcher, 31:7, pp. 15‐21.
Spinoza, Benedict de, (1670) and Jonathan Israel, Editor, (2007). Theological‐Political
Treatise. Cambridge, UK: Cambridge University Press.
Stuart Wells, A., Jellison Holme, J., Tijerina Revilla, A. and Korantemaa Atanda, A. (2009).
Both sides now: The story of school desegregation's graduates. Los Angeles, CA:
California Press.
![Page 206: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/206.jpg)
180
Swanson, C. B., Schneider, B. (1999). Students on the move: residential and educational
mobility in America's schools. Sociology of Education, 72, (1), pp. 54‐67.
Tefera, A., Frankenberg, E., Siegel‐Hawley, G., Chirichigno, G. (2011). Integrating
suburban schools: how to benefit from growing diversity and avoid segregation.
University of California in Los Angeles: Civil Rights project.
www.civilrightsproject.ucla.edu.
Thrupp, M., Lauder, H., Robinson, T. (2003). School composition and peer effects.
International Journal of Educational Research, 37, pp. 483‐504.
Tiebout, C.M. (1956). A Pure Theory of Local Expenditures. The Journal of Political
Economy, 64, pp. 416‐424.
Tweede kamer der Staten Generaal, Vergaderjaar 2006‐07, 30 995, nr. 5. Brief van de
minister van Wonen, werken en Integratie (2007). Aanpak Wijken. The Hague:
SDU Uitgeverijen.
Välijärvi, J., Sahlberg, P. (2008). Should "failing" students repeat a grade? Retrospective
response from Finland. Journal of Educational Change, 9 (4), pp. 385‐389.
Vanderlinde, R, Van Braak, J. (2009). The gap between educational research and
practice: views of teachers, school leaders, intermediaries and researchers.
British Educational Research Journal, 36:2, pp. 299‐316.
Vigdor, J., Ludwig, J. (2010). Neighbourhoods and peers in the production of schooling.
In Brewer, D.J. and McEwan, P.J. (Eds.). Economics of Education. Amsterdam, the
Netherlands: Elsevier.
Welie, E.A.A.M. van, Hartog, J. and Cornelisz I. (2013). Free school choice and the
educational achievement gap. Journal of School Choice. Forthcoming.
Wells, R. (2010). Children of immigrants and educational expectations: The roles of
school composition. Teachers College Record, 112 (6), pp. 1679‐1704.
Werfhorst, H.G. van de, Hofstede, S. (2007). Cultural capital or relative risk aversion? The
British Journal of Sociology, 58 (3), pp. 391‐415.
Werfhorst, van de, H.G. and Van Tubergen, F. (2007). Ethnicity, schooling, and merit in
the Netherlands. Ethnicities, 7, pp. 416‐443.
Yin, R.K. (2009). Case study research. Design and methods. Thousand Oaks, CA: SAGE
Publications Inc.
![Page 207: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/207.jpg)
References
181
Zimmerman, B. J. (2002). Becoming a self‐regulated learner: an overview. Theory into
Practice, 41 (2).
Zohar, A., Dori, Y. (2003). Higher order thinking skills and low‐achieving students: are
they mutually exclusive? The Journal of the Learning Sciences, 12 (2), 145‐188.
![Page 208: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/208.jpg)
![Page 209: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/209.jpg)
183
Samenvatting
Dit proefschrift bestaat uit vier studies naar de positie van leerlingen met een
migranten achtergrond in het voortgezet onderwijs, die wonen in de vier grote steden
Amsterdam, Rotterdam, Utrecht en Den Haag. De onderwijsprestaties van migranten
leerlingen, en de segregatie van scholen en buurten, zijn al enkele decennia onderwerp
van verhit debat in Nederland. Crul, Schneider en Lelie (2009) laten zien dat de kansen
voor een 12‐jarige migranten leerling (hun studie gaat voornamelijk over Turkse
leerlingen) om geconfronteerd te worden met schoolproblemen die uiteindelijk kunnen
leiden tot uitval, vrijwel even groot zijn als de kansen om uitzonderlijk succesvol te zijn in
het onderwijs, en het opleidingsniveau van de ouders verre te overtreffen. In het
maatschappelijke debat echter, lijkt deze succesvolle groep vrijwel aan het zicht
onttrokken te zijn. Het is zorgwekkend dat niet zelden in categorische termen wordt
gesproken over problemen van bijvoorbeeld Marokkaanse leerlingen, die zouden
kunnen suggereren dat hun Marokkaanse afkomst de hoofdoorzaak van hun problemen
is; tevens lijkt er geen rekening mee te worden gehouden dat de leerlingen met
schoolproblemen, niet dezelfde individuen zijn als de succesvolle leerlingen. Verder
laten nationale statistieken (Ministerie van Onderwijs, Cultuur en Wetenschappen) een
gestage toename van het percentage migranten leerlingen op HAVO en VWO zien.
![Page 210: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/210.jpg)
184
Het zal duidelijk zijn dat een opleiding die geen recht doet aan het niveau van je
capaciteiten, en die leidt tot een positie op de arbeidsmarkt waarin je je ambities
onvoldoende kunt verwezenlijken, voor iedere jongere een groot persoonlijk drama is.
Daarnaast echter, vormen jongeren met een migranten achtergrond een groot aandeel
van het toekomstige arbeidspotentieel; zeker in een vergrijzende samenleving hebben
wij hen hard nodig. Tenslotte tonen Groot en Maassen van den Brink (2003) en Levin
(2011) overtuigend aan dat geen investering een grotere opbrengst genereert dan
investeringen in het onderwijs, en dat de kosten van falen zeer hoog zijn. Een belangrijke
bijdrage van hun werk is het pleidooi om de opbrengsten van het onderwijs te
berekenen over het totale werkende leven van een individu, en niet slechts als de
kosten per capita per jaar in het funderend onderwijs.
De vier studies in dit proefschrift onderzoeken 1) of kenmerken van de buurt
waar migranten jongeren wonen, samenhangen met hun niveau in het voortgezet
onderwijs, 2) hoe een stabiele school carrière, niet onderbroken door een overstap naar
een andere school, samenhangt met succes, 3) hoe school keuze, gemeten als de
reisafstand naar een school die wordt verkozen boven de dichtstbijzijnde school,
verschilt tussen diverse etnische groepen, en 4) hoe de transfer van toepasbare
wetenschappelijke kennis scholen met veel migranten leerlingen kan ondersteunen in
hun werk.
In hoofdstuk 2 wordt beschreven hoe het percentage 16‐jarige Nederlands‐
Marokkaanse jongeren, gemeten op het niveau van het 4‐cijferige postcodegebied van
hun woonadres, samenhangt met deelname aan HAVO of VWO. Deze metingen zijn
uitgevoerd voor alle postcodegebieden in de vier grote steden, en daarnaast apart voor
postcodegebieden met geaccumuleerde sociaaleconomische problemen. Tenslotte is
onderzocht hoe de deelname aan HAVO of VWO onder Nederlands‐Marokkaanse
leerlingen samenhangt met het totale percentage 16‐jarigen met een
migrantenachtergrond binnen het postcodegebied waar zij wonen. Ten behoeve van dit
onderzoek is een data set samengesteld uit CBS data (Centraal Bureau voor de Statistiek)
en data beschikbaar gesteld door DUO (Dienst Uitvoering Onderwijs, voorheen CFI).
De associatie tussen de etnische compositie op postcodeniveau en deelname aan
HAVO of VWO (als maat voor onderwijssucces) werd geanalyseerd door middel van
LOWESS regressies (Locally Weighted Scatterplot Smoothing) en lineaire regressies
![Page 211: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/211.jpg)
Samenvatting
185
(OLS). De resultaten laten zien dat er over het geheel een negatief verband is tussen
deze twee variabelen, de dispersie van de datapunten is echter substantieel. De
gemiddelde deelname aan HAVO of VWO onder Nederlands Marokkaanse leerlingen
was in 2009 20.5%, tegen 45.1% onder Nederlandse leerlingen (Bron: Ministerie van
Onderwijs, Cultuur en Wetenschappen). Met de kanttekening dat deze nationale
gemiddelden niet zondermeer vergeleken kunnen worden met de deelpopulatie die in
dit onderzoek is onderzocht, valt op dat alleen in postcodegebieden met rond de 40% of
meer Nederlands‐Marokkaanse jongeren als aandeel van de populatie van 16‐jarigen, de
deelname onder dit landelijke gemiddelde komt. De verschillen tussen alle
postcodegebieden‐ en probleemwijken in het bijzonder, zijn marginaal. De bevindingen
in dit hoofdstuk lijken geen stevige onderbouwing te bieden voor een actief
desegregatie beleid rond schoolplaatsing. Om verder te onderzoeken welke andere
factoren, naast etniciteit, schoolkeuze en niveau in het voortgezet onderwijs sturen, is
een breder onderzoek uitgevoerd, dat in hoofdstuk 4 wordt gepresenteerd.
In het maatschappelijke debat over de onderwijs prestaties van migranten
leerlingen, met name Nederlands‐Marokkaanse leerlingen, stonden enkele postcode
gebieden in Amsterdam West in het centrum van de belangstelling. Door de aandacht
voor veel zorgen en problemen in dit stadsdeel, leken ook hier succesvolle migranten
leerlingen geheel buiten beeld te blijven, of werden zij geportretteerd als de
uitzondering die de regel bevestigt. Om een beter beeld te krijgen van jongeren die
wonen in postcode 1061, 1062 en 1063 in Amsterdam West, werd een data set
geanalyseerd die beschikbaar werd gesteld door de Dienst Maatschappelijke
Ontwikkeling (DMO) van de gemeente Amsterdam. Deze data set bevat een
gedetailleerde schoolloopbaan historie, met alle exacte in‐ en uitschrijfdata in het
onderwijs, van alle 16‐22 jarigen woonachtig in de bovengenoemde drie postcodes op
de peildatum 31 juli 2009. Bij de start van dit onderzoek waren er geen
onderzoeksvragen vooraf, maar gingen we op zoek naar eventuele opvallende
fenomenen in de data. Onmiddellijk, al bij eerste beschouwing van de data, viel het
opvallend grote aantal in‐ en uitschrijvingen op een school op, in het geval van
substantiële aantallen individuen; dit inspireerde onze hoofdvraag, hoe switchen naar
een andere school de resultaten van leerlingen beïnvloedt. De tellingen die worden
gepresenteerd in dit hoofdstuk laten verschillende frequenties van switching tussen
![Page 212: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/212.jpg)
186
etnische groepen zien, en een substantieel verschil tussen migranten leerlingen en
Nederlandse leerlingen. In het geval van Marokkaanse leerlingen behaalde 87.4% een
diploma op HAVO/VWO niveau, wanneer zij hun gehele school periode op dezelfde
school bleven. Onder de Marokkaanse leerlingen die hun derde school voor voortgezet
onderwijs verlieten, had nog slechts 39.1% een diploma behaald.
Om switchen in de drie onderzochte postcodes te kunnen vergelijken met
switchen in de rest van Amsterdam, is een tweede serie analyses uitgevoerd op basis
van de BRON data (burger service nummer en onderwijsnummer; DUO, Ministerie
OCW). Deze analyses laten zien dat met name Nederlands Marokkaanse leerlingen die in
de drie postcodegebieden wonen, vaker switchen dan Marokkaanse leerlingen in de rest
van Amsterdam Over het algemeen switchen migranten leerlingen vaker dan leerlingen
van Nederlandse origine, hoewel er verschillen zijn tussen etnische groepen. In
ongeveer een derde van de gevallen switchen leerlingen binnen een groep van
samenwerkende scholen.
De BRON data maken het mogelijk om de mobiliteit naar een hoger of lager
niveau in het voortgezet onderwijs te meten, als de leerling in de derde klas zit.
Substantieel meer leerlingen die opstromen, blijven op dezelfde school. Onder de
switchers neemt de opstroom af, en neemt de afstroom toe. Voor Marokkaanse
leerlingen in Amsterdam, is het verschil in opstroom tussen leerlingen die niet‐ en die
wel switchen, bijna 12 procentpunt. In tegenstelling tot de hoge kosten die gemoeid
gaan met het terugdringen van drop‐out, zou het reduceren van switching vrijwel
kostenneutraal kunnen zijn, omdat het succes vooral afhangt van goede
overeenkomsten tussen scholen, en transparantie over het beleid om leerlingen al dan
niet vrijwillig over te plaatsen naar een andere school.
Hoofdstuk 4 presenteert een studie naar patronen van schoolkeuze in het
voortgezet onderwijs. In Nederland is de schoolkeuze vrij, (behoudens enkele
ontwikkelingen rond loting bij over‐intekening), en zijn er geen financiële
belemmeringen voor ouders, omdat het funderend onderwijs volledig door de overheid
wordt gefinancierd. Ten behoeve van dit onderzoek hebben wij een zeer rijke BRON data
set gebruikt, met de gegevens van 16.071 leerlingen die in 2008 in groep 8 van de
basisschool zaten, en ten tijde van ons onderzoek in de meeste gevallen in de derde klas
van het voortgezet Onderwijs zaten. Voor een deel van de leerlingen waren ook de CITO
![Page 213: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/213.jpg)
Samenvatting
187
score en het advies van de docent van groep 8 in het basis onderwijs beschikbaar. De
BRON data set met uitgebreide (volledig geanonimiseerde) gegevens op het individuele
niveau, werd gekoppeld aan sociaal economische gegevens op buurtniveau van het
Sociaal Cultureel Planbureau (SCP) en het Centraal Bureau voor de Statistiek (CBS) , en
kwaliteitsgegevens op school niveau van de Inspectie van het Onderwijs. Op deze wijze
konden wij schoolkeuze patronen analyseren op basis van regressies met controle
variabelen op het individuele‐, buurt‐ en school niveau.
Wij keken achtereenvolgens naar 1) de associatie tussen de gemiddelde afstand
naar school en individuele‐ en sociaaleconomische variabelen voor verschillende
etnische groepen, 2) mogelijke verschillen tussen de groep die de dichtstbijzijnde school
kiest en de groep die een andere school prefereert, 3) in het geval van leerlingen die niet
de dichtstbijzijnde school kiezen, het verschil in afstand tussen de gekozen en de
dichtstbijzijnde school, 4) het verschil in kenmerken tussen de gekozen en de
dichtstbijzijnde school en 5) de associatie tussen schoolkeuze en opwaartse mobiliteit
naar een hoger niveau in het voortgezet, op individueel niveau.
Onze resultaten laten zien dat migranten leerlingen gemiddeld minder ver naar
school reizen dan Nederlandse leerlingen. Echter, in postcodegebieden met een lage
gemiddelde SES (Sociaal Economische Situatie) reizen migranten leerlingen nog minder,
maar Nederlandse leerlingen juist verder. Daarentegen in rijke buurten, is de
gemiddelde afstand naar school onder Nederlandse leerlingen laag. Enigszins tot onze
verrassing, vonden wij geen systematische verschillen tussen de groepen die de
dichtstbijzijnde‐ of een andere school kozen. Maar er bleek een opmerkelijke verschil in
school preferenties tussen Nederlandse leerlingen en migranten leerlingen die een
school verder weg kozen: Nederlandsen leerlingen kozen gemiddeld een school met een
lager percentage migranten leerlingen, een hogere school SES, en een hoger school
eindexamen gemiddelde‐ in vergelijking met de dichtstbijzijnde school. Bijvoorbeeld
Nederlands‐Marokkaanse leerlingen daarentegen, kozen een school met een nog hoger
percentage migranten leerlingen dan hun dichtstbijzijnde school, zelfs ten koste van een
lagere school SES. Mogelijk hebben wij hier patronen van zelfverkozen segregatie
zichtbaar gemaakt, een inspiratie voor vervolgonderzoek. Hoewel het verschil klein is, is
de kans op opwaartse mobiliteit (vooral voor migranten leerlingen een belangrijk
![Page 214: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/214.jpg)
188
mogelijkheid om de prestatiekloof te versmallen) statistisch significant iets groter
naarmate leerlingen dichter bij huis naar school gaan.
In hoofdstuk 5 tenslotte, wordt een onderzoek gepresenteerd naar de wijze
waarop toepasbare resultaten van bestaand wetenschappelijk onderzoek, gekoppeld
kunnen worden aan actuele vragen van scholen met veel migranten leerlingen. Het idee
voor dit onderzoek is ontstaan tijdens mijn promotie traject: vrijwel dagelijks verschijnen
er wetenschappelijke publicaties over groepen leerlingen die nog onvoldoende kansen
hebben gekregen om op de juiste school, of het juiste niveau terecht te komen, gegeven
hun capaciteiten. Veel van deze inzichten uit wetenschappelijk onderzoek zouden
vertaald kunnen worden naar innovatieve handelingspraktijken in scholen, en zouden
schoolleiders en docenten kunnen ondersteunen in het nemen van strategische
beslissingen. Echter, uitkomsten van wetenschappelijk onderzoek bereiken scholen niet
of nauwelijks.
Samen met 6 schoolleiders in het voortgezet onderwijs in Amsterdam, is een
methode ontworpen en getoetst om schoolvragen te verbinden wetenschappelijke
kennis. 1) Op basis van een semigestructureerd diepte‐ interview met individuele
schoolleiders, werden belangrijke schoolvragen geïdentificeerd; 2) in de
wetenschappelijke literatuur werd gezocht naar een match met deze thema's; 3) deze
wetenschappelijke inzichten werden gemotiveerd per brief gerapporteerd aan de
schoolleiders; 4) na 3 maanden werd nagegaan of deze aangedragen kennis was
toegepast. De resultaten laten zien dat schoolleiders inderdaad belangrijke vragen
hebben, die in vrijwel alle gevallen konden worden gekoppeld aan toepasbare
wetenschappelijke inzichten. De meeste vraagstukken waren van conceptuele of
strategische aard, bijvoorbeeld over de functie van de school in de context van grote
maatschappelijke vraagstukken rond diversiteit; er waren nauwelijks vragen van meer
instrumentele aard, wat mogelijk samen kan hangen met het feit dat alleen
schoolleiders werden geïnterviewd. Schoolleiders pleitten voor een intermediaire
functie tussen het onderwijsveld en de academische wereld, die naar hun inzicht het
beste vervuld zou kunnen worden door mensen die beide werelden van binnenuit
kennen. Verder hechtten verschillende schoolleiders aan vervolgonderzoek naar de
concrete vertaling van wetenschappelijke inzichten naar de lespraktijk in de klas.
![Page 215: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/215.jpg)
Samenvatting
189
Tijdens ons onderzoek kwamen wij een vrij ernstige belemmering voor scholen
op het spoor: veel publicaties in wetenschappelijke tijdschriften zijn slechts beperkt vrij
toegankelijk via het internet, en kunnen alleen binnen de muren van een universiteit‐
waar aanzienlijke licence fees worden betaald‐ worden geraadpleegd. Deze beperking
maakt het moeilijk voor schoolleiders om zelf te zoeken naar relevante
wetenschappelijke publicaties.
![Page 216: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/216.jpg)
![Page 217: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/217.jpg)
191
Biography
Liesbeth van Welie graduated in Biology from Radboud University (Nijmegen, the
Netherlands), with a specialization in Aquatic Ecology.
She started her career as a teacher in Biology and a few years later was
appointed as Principal of an innovative newly‐founded school for secondary education.
Her second appointment as Principal was at a diverse inner‐city school in Amsterdam,
where she devoted her work to educating migrant pupils for successful access to higher
education.
Next, she was invited by the Board of the University of Amsterdam, to start a
large programme for improving the quality of academic teaching and learning, as Senior
Vice‐President for Education. During these years, she engaged in an extended
international network in the field of quality assessment of higher education,
internationalization and organizational change. Her next step was advisor to the Board
at Maastricht University, followed by two years as senior consultant at a consultancy
firm, where her portfolio consisted of assignments at several Dutch universities.
For several years she held the position of Chief Inspector of Secondary and
Higher Education, where after the Secretary General of the Ministry of Education,
Culture and Science, invited her to make an organisational design for the enhancement
of evidence‐ and information‐based policy making at the government level.
![Page 218: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/218.jpg)
![Page 219: THEY WILL GET THERE!prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit](https://reader036.vdocuments.us/reader036/viewer/2022081513/605034045be8e344874f9712/html5/thumbnails/219.jpg)
193
Top Institute for Evidence Based Education Research (TIER)
The Top Institute for Evidence Based Education Research (TIER) is an inter‐university
institute that conducts research to develop evidence based education. The institute has
three partners: the University of Amsterdam, Maastricht University and the University of
Groningen and is located in Amsterdam, Maastricht and Groningen.
TIER contributes to the improvement of the quality of education in the Netherlands by
promoting an evidence based approach as a guiding principle in education policy and
practice. It accomplishes this by developing (cost) effective education interventions that
are grounded in sound scientific research. TIER research is funded by the Ministry of
Education, Culture and Science and the participating universities through NWO (The
Netherlands Organisation for Scientific Research) and complies with the quality
standards and evaluation procedures used by NWO.
The following books recently appeared in the TIER Research Series:
I. C. Haelermans (2012), On the productivity and efficiency of education. The
role of innovations in Dutch secondary education
II. L. van Welie (2013), They Will Get There! Studies on Educational Performance
of Immigrant Youth in the Netherlands