The risk of downward mobility in educational attainment

Life Course Research

Sophie Hahn

The Risk of Downward
Mobility in Educational
Attainment
Children of Higher-Educated Parents
in Germany


Life Course Research
Herausgegeben von
Prof. Dr. Steffen Hillmert, Universität Tübingen


Sophie Hahn

The Risk of Downward
Mobility in Educational
Attainment
Children of Higher-Educated Parents
in Germany
With a foreword by Prof. Dr. Steffen Hillmert


Sophie Hahn
München, Germany
This work has been accepted as dissertation thesis at the University of Bamberg in 2015.
The underlying research was funded by the German Federal Ministry of Education and
Research (reference number 01JG1059). The author is solely responsible for the content

of this publication.

OnlinePlus material to this book can be available on
/>Life Course Research
ISBN 978-3-658-14597-2
ISBN 978-3-658-14598-9 (eBook)
DOI 10.1007/978-3-658-14598-9
Library of Congress Control Number: 2016943415
Springer VS
© Springer Fachmedien Wiesbaden 2016
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or
dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained
herein or for any errors or omissions that may have been made.
Printed on acid-free paper
This Springer VS imprint is published by Springer Nature
The registered company is Springer Fachmedien Wiesbaden GmbH


Foreword

In recent years, theoretical models of rational educational decisions have become

popular among researchers who study social inequalities in education. According to
these models, inter-generational maintenance of social status plays a central role for
the considerations of school students and their parents. In particular, the attempt to
avoid the loss of status within the family is perceived as a major driving force of specific educational decisions. Due to the different positions of families in the inequality
structure of society, this rationale tends to lead to origin-specific educational choices
and, consequentially, to social reproduction across generations.
Given its prominence in theory, it is surprising that so far relatively little empirical
research in education and social mobility studies has focused on the phenomenon of
downward mobility. The present work by Sophie Hahn contributes to closing this gap.
It is based on retrospective data from the German National Educational Panel Study
(NEPS) and covers various aspects of the topic including developments along the life
course, access to higher education via the "detour" of vocational training, and drop out
from tertiary education.
The study demonstrates the relevance of social origin for inter-generational and intra-general downward educational mobility. In particular, downward mobility is common in families of higher educated parents. Though expectable from a formal perspective, this is a remarkable substantive finding in light of considerable educational expansion in recent decades. It might also be relevant to discussions about increasing risks of
downward status mobility. Children of higher educated parents also have the highest
level of mobility within their educational career, i.e., social origin is not only an important determinant of educational attainment but also of educational pathways.
The message for inequality research is therefore to look not just at selected transitions but whole educational careers. This conclusion is in line with the paradigmatic
assumptions of life-course research. Therefore, Sophie Hahn's book fits perfectly into
our series, Life Course Research. The series publishes empirical studies – in both English and German – that focus on transitions along the life course in various areas of
life. I hope many readers will benefit from reading this book.
Steffen Hillmert


Preface

This book is a slightly revised version of my dissertation thesis that I handed in at the
University of Bamberg in February 2015. First and foremost, I thank my first supervisor, Thorsten Schneider, for being an outstanding mentor. He invested a large amount
of his time and thought, gave insightful advice and was always available for my questions. Also, I thank Sandra Buchholz for writing an elaborate second expert’s report
and Steffen Schindler for readily agreeing to be additional assessor in the thesis defense. Furthermore, I thank Steffen Hillmert for admitting this book to appear in the
series, Life Course Research. I learned a lot from these four excellent researchers and

their suggestions have improved this book substantially. Last but not least, I am very
grateful for the comments I received from doctoral fellows of the doctoral programme
at the German Institute for International Educational Research (DIPF) and at the colloquia of Thorsten Schneider and Hans-Peter Blossfeld.
The research was funded by the German Federal Ministry of Education and Research (reference number 01JG1059). The author is solely responsible for the content
of this publication.
Sophie Hahn


Contents

Figures and tables ....................................................................................................... 13
1 Introduction: Social mobility perspective and intergenerational downward
mobility in educational attainment ...................................................................... 17
2 Research on intergenerational mobility .............................................................. 23
2.1 Research on social mobility and educational inequality ...................................... 23
2.2 Research on intergenerational downward mobility .............................................. 33
3 Theoretical approaches: The life-course perspective and
rational-choice-based theories of educational decisions .................................... 41
3.1 The life-course perspective................................................................................... 41
3.2 Educational decisions ........................................................................................... 44
3.2.1 Primary and secondary effects ........................................................................... 45
3.2.2 Decision parameters: basic concepts and variations .......................................... 47
3.2.3 Status maintenance ............................................................................................. 52
3.2.4 The Mare model of sequential educational decisions ........................................ 53
3.2.5 Critique of rational choice theory and limitations of rationality ........................ 55
3.3 Combining the life-course approach and rational-choice-based theories of
educational decisions ............................................................................................ 58
4 Data: The sub-study Adult Education and Lifelong Learning of the
National Educational Panel Study ....................................................................... 61
5 Empirical analysis 1: Educational downward mobility over time in

Germany................................................................................................................. 63
5.1 Theoretical considerations and hypotheses .......................................................... 63
5.2 Data ....................................................................................................................... 67
5.3 Variables ............................................................................................................... 68
5.4 Methods ................................................................................................................ 70
5.5 Results .................................................................................................................. 73
5.5.1 Educational pathways and the education of parents........................................... 78
5.5.2 Duration of education by education of parents .................................................. 81
5.5.3 Development of downward mobility over cohorts ............................................. 83
5.6 Summary and conclusion on educational downward mobility over time in
Germany ............................................................................................................... 86


10

Contents

6 Empirical analysis 2: Re-entering the academic pathway after starting
vocational training in Germany ........................................................................... 89
6.1 Social selectivity before and after the tertiary education entrance certificate...... 90
6.2 Theoretical considerations and hypotheses .......................................................... 92
6.2.1 Decreasing effects of social origin at late educational transitions due to
growing independence ........................................................................................ 92
6.2.2 Persisting differences in educational decisions by social origin in adult age .... 93
6.2.3 Path dependence ................................................................................................. 95
6.2.4 Age norms and competing life-course roles ....................................................... 96
6.2.5 Opportunity costs and type of vocational training ............................................. 98
6.2.6 Change in transition probabilities over cohorts................................................ 101
6.3 Data ..................................................................................................................... 102
6.4 Variables ............................................................................................................. 103

6.4.1 Characteristics of the school career .................................................................. 103
6.4.2 Competing life-course roles ............................................................................. 104
6.4.3 Characteristics of the vocational training programme ..................................... 104
6.5 Methods .............................................................................................................. 105
6.6 Results ................................................................................................................ 106
6.6.1 Selectivity of school leavers with tertiary education entrance certificate
choosing vocational training ............................................................................ 106
6.6.2 Descriptive results on activities in the first 5 years after obtaining a tertiary
education entrance certificate ........................................................................... 108
6.6.3 Multivariate results ........................................................................................... 111
6.7 Summary and conclusion on re-entering the academic pathway after
starting vocational training in Germany ............................................................. 116
7 Empirical analysis 3: Access to tertiary education and dropout in
Germany............................................................................................................... 121
7.1 Selection processes prior to tertiary education in the German education
system ................................................................................................................. 122
7.2 Descriptive results on access to tertiary education ............................................. 124
7.3 Theoretical considerations and hypotheses on dropout from tertiary
education ............................................................................................................. 127
7.3.1 Mechanisms linking pre-tertiary pathways and dropout from higher
education........................................................................................................... 127
7.3.2 The role of social origins .................................................................................. 131
7.3.3 The role of time ................................................................................................ 132
7.4 Data ..................................................................................................................... 133
7.5 Sample, variables and methods of the analysis on dropout from tertiary
education ............................................................................................................. 133
7.6 Results on dropout from tertiary education ........................................................ 136
7.6.1 Descriptive results of dropout from tertiary education .................................... 136
7.6.2 Multivariate results ........................................................................................... 140
7.7 Summary and conclusion on access to tertiary education and dropout in

Germany ............................................................................................................. 150


Contents

11

8 Intergenerational downward mobility in educational attainment in
Germany: Summary of the main results and conclusions .............................. 157
8.1 Summary of the main results .............................................................................. 157
8.2 Limitations of the present study ......................................................................... 162
8.3 Conclusions on the utility of studying downward mobility for sociological
research ............................................................................................................... 163
8.4 How do the findings connect to previous and future research? ......................... 166
References .................................................................................................................. 171
Appendix A ................................................................................................................ 189
Appendix B ................................................................................................................ 191
Appendix C ................................................................................................................ 193
Appendix D ................................................................................................................ 195


Figures and Tables

Figure 2.1.
Figure 5.1.
Figure 5.2.
Figure 5.3.
Figure 5.4.
Figure 5.5.
Figure 6.1.

The OED Triangle ................................................................................... 26
Educational attainment by education of the parents ............................... 73
Sequence index plots of educational pathways I .................................... 75
Sequence index plots of educational pathways II ................................... 76
Educational pathways by education of parents ....................................... 79
Mean durations of education by education of parents ............................ 81
Sequence index plots of activities 5 years after obtaining a tertiary
education entrance certificate by parental education ............................ 109
Figure 6.2. Survivor functions for entry into tertiary education after tertiary
education entrance certificate and first vocational training .................. 111
Figure 7.1. Flow chart showing educational pathways with percentages ............... 125
Figure 7.2. Proportion functions of dropout and graduation from higher
education at universities based on survivor functions
(Kaplan–Meier method) ........................................................................ 139
Figure 7.3. Proportion functions of dropout and graduation from higher
education at universities of applied sciences, based on survivor
functions (Kaplan–Meier method) ........................................................ 139
Figure A.1. The German education system .............................................................. 189
Figure D.1. Sequence index plots of pathways of students who enter tertiary
education ............................................................................................... 196
For multi-coloured versions of these figures please visit the website of this book at
www.springer.com.
Table 2.1.
Table 2.2.
Table 4.1.
Table 5.1.
Table 5.2.
Table 5.3.
Table 5.4.

Table 6.1.
Table 6.2.

Ideal typical outflow table from origin class to destination class ........... 24
Ideal typical inflow table from origin class to destination class ............. 24
Overview of samples and survey instruments for NEPS starting
cohort 6 ................................................................................................... 62
Types of educational pathway ................................................................ 78
Percentages of types of pathway by education of parents
and cohorts .............................................................................................. 80
Linear regression of mean duration in education.................................... 83
Logistic regression of reaching a tertiary degree over cohorts
(odds ratios)............................................................................................. 85
Standardized mean grade point averages of the tertiary education
entrance certificate in different categories of post-secondary activities
by education of parents ......................................................................... 108
Post-secondary education of persons with tertiary education
entrance certificate by education of the parents .................................... 110


14
Table 6.3.
Table 7.1.
Table 7.2.
Table 7.3.
Table 7.4.
Table 7.5.
Table 7.6.
Table B.1.
Table C.1.

Table D.1.
Table D.2.
Table D.3.
Table D.4.
Table D.5.
Table D.6.
Table D.7.
Table D.8.
Table D.9.

Figures and Tables

Piecewise-constant exponential model for entry into tertiary
education after tertiary education entrance certificate and first
vocational training................................................................................. 112
Prevalence of pre-tertiary educational pathways among students at
universities and universities of applied sciences .................................. 137
Students at universities and universities of applied sciences by
education of parents .............................................................................. 137
Piecewise constant exponential models on dropout from
universities within 8 years of starting tertiary studies .......................... 141
Piecewise constant exponential models on dropout from universities
within 8 years of starting tertiary studies – including effects of
educational pathways ............................................................................ 144
Piecewise constant exponential models on dropout from
universities of applied sciences within 8 years of starting tertiary
studies.................................................................................................... 146
Piecewise constant exponential models on dropout from
universities of applied sciences within 8 years of starting tertiary
studies – including effects of educational pathways ............................. 148

Distribution over types of education 24 years after school entry ......... 191
Re-entering the academic pathway after starting vocational
training in Germany: Distribution of characteristics in the sample
at certain points in time ......................................................................... 193
Dropout from tertiary education in Germany: Distribution of
characteristics in the sample at certain points in time at
universities ............................................................................................ 197
Distribution of characteristics in the sample at certain points in
time at universities of applied sciences ................................................. 198
Distribution of age at enrolment in universities .................................... 199
Distribution of age at enrolment in universities of applied
sciences ................................................................................................. 199
Piecewise constant exponential models on dropout from
universities within 8 years of starting tertiary studies –
including all covariates ......................................................................... 200
Piecewise constant exponential models on dropout from
universities within 8 years of starting tertiary studies –
including effects of the duration at the Gymnasium ............................. 202
Piecewise constant exponential models on dropout from
universities within 8 years of starting tertiary studies –
including effects of educational pathways ............................................ 204
Piecewise constant exponential models on dropout from
universities within 8 years of starting tertiary studies –
including fields of study........................................................................ 206
Piecewise constant exponential models on dropout from
universities of applied sciences within 8 years of starting tertiary
studies – including all covariates .......................................................... 208


Figures and Tables

15

Table D.10. Piecewise constant exponential models on dropout from
universities of applied sciences within 8 years of starting tertiary
studies – including the effect of the duration at Gymnasium ................ 210
Table D.11. Piecewise constant exponential models on dropout from
universities of applied sciences within 8 years of starting tertiary
tudies – including effects of educational pathways .............................. 212
Table D.12. Piecewise constant exponential models on dropout from
universities of applied sciences within 8 years of starting tertiary
studies – including fields of study ........................................................ 214


1

Introduction: Social mobility perspective and
intergenerational downward mobility in educational
attainment

Introduction

Education is crucially linked to assets such as income, occupational position and social
prestige. These assets are determinants of individual life chances since they affect individual wealth, power, health, and life expectancy. Also, education is unevenly distributed among individuals within a society. Inequalities refer to the access to vertical
levels of educational programmes, degrees, or assessment results and to horizontal differences between types of educational programmes. Both, vertical and horizontal differences may result in unequal life chances.
Due to their crucial consequences on life chances inequalities in education ask for
strong legitimation. In a meritocratic perspective, educational inequalities are legitimate if they are linked to a person’s level of ability and effort. Accordingly, more talented individuals who make stronger efforts should reach higher levels of education
and, in consequence, higher occupational and social positions. From an economical
perspective this assignment of persons to positions is the most efficient way of handling tasks in society and to a higher level of overall wealth.
However, in reality, educational success such as moving to a higher level of education, receiving good grades, or obtaining educational degrees does not solely depend

on talent and effort. There are structural determinants such as the level of differentiation of an education system, the supply of educational programmes in a region, or the
size of a birth cohort that affect individual educational chances above talent and effort.
Furthermore, educational chances depend on political decisions about the expansion of
certain educational programmes. Far from being an exhaustive list, these are some
structural determinants that affect educational success independently from individual
talent and effort.
Structural determinants affect groups in society differently. These differences are
linked to characteristics such as gender, migration background, or social background.
The social mobility perspective on educational inequalities is a particular way of looking at differences in educational outcomes. It links two or more generations by taking
into account the status, or its components, of the parents (and sometimes grandparents)
when analysing differences in educational outcomes. ‘Status’ is a concept in stratifica-

© Springer Fachmedien Wiesbaden 2016
S. Hahn, The Risk of Downward Mobility in Educational Attainment,
Life Course Research, DOI 10.1007/978-3-658-14598-9_1


18

Introduction

tion research that refers to economic and non-economic features of a persons’ standing
in society. In empirical research the most commonly used components are income,
education and occupational prestige. The typical research question from social mobility perspective is: ‘What are the chances of children to obtain a certain educational
degree given their parental level of education, occupation, or social prestige?’ The
concept of social mobility is based on the assumption that educational outcomes, occupational positions, and social prestige can be categorized and ordered hierarchically.
This makes sense because they are linked to different levels of income, wealth, power,
unemployment risks, health, or life expectancy that, indeed, can be aligned in a vertical order. Furthermore, social mobility invokes the notion that the parents’ status defines a starting point (social origin) whereas the achieved status is the point of destination. ‘Mobility’ refers to the trajectory between social origin and social destination.
Education, thereby, is considered as a connecting piece between social status of origin
and occupational and social position of destination. Persons from higher social origin

gain higher levels of education and, thus, higher occupational and social positions. A
link in educational, occupational and social outcomes between two (or more) generations can be expected for several reasons. One reason is that children inherit their parents’ dispositions for intelligence and personality traits that may determine educational
and occupational success. However, genetical inheritance explains the link only partly.
Another explanation is that parents shape their children’s identity by socialization.
Parents significantly affect their children’s ideas, opinions, aspirations and ambitions
by educating and acting as models. It can be assumed that parents usually whish the
best for their children. Thus, it would be reasonable for all parents to try to direct their
children to the highest educational degrees because high educational outcomes are
connected to desirable life chances. However, social mobility theories argue that
higher occupational and social positions come along with a greater stock of resources
that can be invested in the children’s educational and occupational career. Thus, children of higher-status-parents more easily obtain higher degrees and positions themselves. Since on Bourdieu (1982) it has become popular in social mobility research to
think of three different types of resources. First, higher educational degrees and occupational positions bear higher economical resources which are helpful because money
and wealth can be invested in educational fees, learning materials such as books and
living costs during education when the child cannot earn money by itself. Second,
higher status is linked to higher cultural resources such as the parents’ ability to help
their children with homework, the complexity of language spoken at home, familiarity


Introduction

19

with high culture music, art, literature, or having objects at home as books and instruments which lower the thresholds to engage in learning activities. Also, the parents’
knowledge and experience of a higher educational pathway may be a resource that
helps a child to obtain higher educational credentials. Finally, social networks may
further enhance that children obtain educational levels and social positions similar to
those of their parents. Children grow up in the social environment of their parents and
learn how to behave in this social context. Higher educated parents tend to have social
networks with conventions and norms that resemble those of networks in academic
environments so that their children may have less difficulty adjusting to these environments. Also, some theories argue that leaving the parents’ social context by moving

either upward or downward in society involves costs e.g. a feeling of rejection in both,
the new and the former social context. In case of downward moves, persons might suffer from losing the level of wealth they were accustomed to and from feeling that they
have deceived the expectations of their parents. In order to avoid these costs children
may try to obtain an educational level and social position similar to their parents.
Although there are innumerable studies of intergenerational mobility in sociology,
intergenerational downward mobility in educational attainment has rarely been a focus
of analysis. At first sight, this may seem quite natural, because stability in educational
and social attainment from one generation to the next has been and continues to be the
most common pattern in Western societies. Moreover, upward mobility is more frequent than downward mobility. Against this background, one important research question is whether the opening up of higher levels of education to all and the rise in educational attainment has led to more equal chances between social classes. Answering
this question requires analyses of the differences in educational and occupational
chances to be found between classes of social origin. The concepts of intergenerational
upward and downward mobility, however, are rather inadequate for comparisons between classes, because they depend on the place within the class structure: whereas
those at the bottom of the class structure cannot move further downward, those at the
top cannot move further upward. At the same time, chances to achieve high educational levels and high social positions increase from classes at the bottom to classes at
the top.
Nonetheless, there are sound reasons for taking a closer look at intergenerational
downward mobility and placing more emphasis on it in sociological research. First, the
share of persons that are downwardly mobile is quite remarkable in general and in particular in higher classes of origin. For example, Holtmann (2010) states that in West-


20

Introduction

ern Germany only about half of the children with tertiary educated parents of birth cohorts 1920 to 1979 reaches the same education level as their parents. Second, the concept of downward mobility allows theories of educational inequality to be considered
from a different angle. To be sure, the fact that persons do not attain the status of their
parents does not per se contradict a strong link between social origin and destination.
Nevertheless, the focus on those who do not match the odds of maintaining their status
of origin despite favourable preconditions does put theories to the test. Third, there is
strong evidence that parents wish their children to attain a social position at least as

favourable as their own. The study of downward mobility provides an opportunity to
trace parental influence, and thus the effect of social origin, over the life course. On
the one hand, parents exert a strong influence on educational decisions. This is where
resources, effort, and the motive of status maintenance determine the difference in the
educational careers of children who have a different social origin. On the other hand,
with growing age, children become increasingly independent from their parents. Using
the parents’ educational level as a reference for the children’s educational career gives
some idea of the persistence and fading away of social origin effects. Fourth, as posited by the principle of risk aversion, downward mobility is likely to be a strongly
meaningful category for the actors themselves that will have strong consequences for
their behaviour. Research shows that not having attained the parents’ educational degree is an important motivating force for re-entering education and thus a major determinant of later educational careers (Jacob & Tieben, 2009; Jacob & Weiss, 2011).
Hence, preliminary downward mobility is important in light of the development of
social disparities over the life course.
For these reasons this book deals with downward mobility and its effects on educational careers. In the analyses I focus on downward mobility in educational attainment
because this is the core part of the origin–destination relation. Furthermore, I concentrate on persons with the highest educated parents, i.e. persons whose parents have
tertiary degrees, and compare them to persons whose parents have lower than tertiary
degrees. Of course, those whose parents have intermediate educational degrees can be
downwardly mobile, as well. However, this would require a separate, extensive analysis. In the German education system, the divide between educational careers that lead
to tertiary degrees (‘academic careers’) and more vocationally oriented careers is
strong and demarcation to other tracks is clearer than between intermediate types of
education.


Introduction

21

Obtaining an educational degree is a process of several steps. Therefore, exclusively using the highest achieved educational degree would give away the chance to
learn more about the mechanism how social origin affects educational attainment. A
tertiary degree, for example, requires to have completed certain schools and to have
entered university. In Germany, the pathway ideal-typically involves a transition from

primary school to a secondary school where the tertiary education entrance diploma
can be obtained (the ‘Gymnasium’) and afterwards the transition to a university or a
university of applied sciences. However, educational careers often are not continuous
and ideal-typical but involve detours, dropouts and re-enrolments. Indirect as well as
direct pathways may be linked to the motivation to reach the parents’ educational
level. Analysing educational careers might help to find out how typical pathways of
status maintenance and downward mobility look like. Thus, in order to get a more
complete picture of downward mobility it makes sense to look at both, intergenerational moves in the educational level from parents to children and intragenerational progression within educational careers of the children.
This requires longitudinal methods of analysis. Individual educational careers can
be examined as completed total entities or as stepwise sequences of crucial transitions.
Since each approach has blind spots I will apply them both. Looking retrospectively at
the entire career has the advantage that types of educational careers, including downwardly mobile pathways, may be identified. A disadvantage is that only completed
educational careers can be taken into account. Thus, younger persons who may still be
in education have to be excluded. Also, some research questions require taking into
account the stepwise progression of educational careers. For this purpose, it has to be
considered that the original sample of persons who have entered the educational system splits up and partly exits from the educational system. So, the sample of persons
who are eligible to a certain transition differs from the original sample. For example,
when analysing dropout from tertiary education only those persons may be considered
who are enrolled in tertiary education. Thus, additionally to looking at entire educational careers I analyse transition rates at selected points of the educational career.
In order to locate intergenerational downward mobility in educational attainment
within its broader context, I shall first summarize influential research on intergenerational social and educational mobility and present previous research on intergenerational downward mobility in chapter 2. In chapter 3, I shall review the main theoretical
approaches on which the analyses of downward mobility in this book are based. Chapter 4 provides information on the data used in the empirical analyses.


22

Introduction

The analyses themselves are described in chapters 5 to 7. In the first empirical
analysis (chapter 5) I shall look at educational careers as a completed entity in order to

get an idea of the extend of downward mobility, its development over time and different types of careers using the academic pathway as a reference. First, I shall examine
the link between highest educational attainment of parents and children over cohorts.
This tackles the question how far intergenerational downward mobility in educational
attainment exists, and how it has developed during the educational expansion. Second,
I shall identify types of educational careers and compare their frequency in different
groups of social origin.
In the second and third empirical analysis I shall analyse relative risks of selected
transitions in the educational career. I chose one re-enrolment and one dropout transition in order to examine how the parents’ educational degree and the risk of downward
mobility affect educational careers: Are persons who otherwise risk not to reach the
educational level of their parents more inclined to re-enter education? Are they less at
risk to drop out from a path that leads to the educational level of their parents? Both
are late transitions of the educational career. These are particularly interesting in order
to test whether the effect of social origin still affects persons when they are adults.
The analysis in chapter 6 tackles the decision of persons with tertiary education entrance diploma to take up tertiary studies after having started a vocational training.
While tertiary entrance diploma and tertiary studies are parts of the academic pathway
the entrance into vocational training may be considered as a detour from this pathway.
The research question is whether persons who cannot reach their parents’ educational
degree with vocational education have higher inclinations to enter tertiary education
and, thereby, re-enter the academic pathway.
The third empirical analysis (chapter 7) focuses on the effects of social origin on
dropout from tertiary education. The research question is whether social origin prevents from dropping out of tertiary education or whether downward mobility here is a
consequence of late career decisions in which the parental influence has become attenuated.
Chapter 8 comprises a summary of the main results and the conclusion.


2

Research on intergenerational mobility

Research on intergenerational mobility

In this section, I shall review research on intergenerational mobility in order to embed
intergenerational downward mobility in educational attainment within its broader research tradition. First, I shall summarize research on social mobility and educational
inequality. Then, I shall present studies on intergenerational downward mobility.
2.1

Research on social mobility and educational inequality

Social mobility, the ‘relationship between the class position a person occupies and the
class in which he or she was brought up’ (Breen, 2004, p. 1), is a core topic in sociology. In general, high correlations can be observed between the class an individual is
born into, also called the class of origin, and the occupational class position that an
individual achieves, also called the destination class.
Research on social mobility distinguishes between absolute and relative mobility.
The former refers to the observed distribution of persons from different classes of origin over destination classes. This distribution can be illustrated by mobility tables that
cross-tabulate origin class, usually as a row variable, with destination class, usually as
a column variable. Thus, mobility tables show which classes persons reach given the
class they were born into. Beyond this, three main types of information can be read
from mobility tables. First, the diagonal shows the amount of immobility, that is, the
percentage of persons reaching the same class as their parents. Second, the amount of
upward mobility can be computed from the sum of the percentages of persons reaching
higher classes than their parents. Third, the amount of downward mobility can be
computed from the sum of the percentages of persons reaching lower classes than their
parents.
Percentages of mobility between origin and destination classes can be presented in
two ways, either by outflow or inflow tables. Outflow tables show, for each class of
origin, the share of people in any destination class in such a way that the percentages
in each origin class add up to 100 per cent. Inflow tables show, for each destination
class, the share of persons from any class of origin – thus, percentages in each destination class add up to 100 per cent.

© Springer Fachmedien Wiesbaden 2016

S. Hahn, The Risk of Downward Mobility in Educational Attainment,
Life Course Research, DOI 10.1007/978-3-658-14598-9_2


24
Table 2.1.

Research on intergenerational mobility

Ideal typical outflow table from origin class to destination class
Destination class (child's occupation)

Origin class (Highest
occupation of the parents)
Class 1
No.
%
Class 2
No.
%
Class 3
No.
%

Table 2.2.

Class 1
100
50
90

30
20
10

Class 2
80
40
150
50
80
40

Class 3
20
10
60
20
100
50

Total
200
100
300
100
200
100

Ideal typical inflow table from origin class to destination class

Origin class (Highest
occupation of the parents)
Class 1
Class 2
Class 3
Total

Destination class (child's occupation)
Class 1
Class 2
No.
%
No.
%
100
48
80
26
90
43
150
48
20
10
80
26
210
100
310
100

Class 3
No.
20
60
100
180

%
11
33
56
100

As mentioned above, alongside the concept of absolute mobility, research in social
mobility relies on the concept of relative mobility or social fluidity. This refers to the
relative chance that persons in each class of origin have of achieving a certain destination class rather than another. It is measured as the ratio of the odds of reaching one
class destination rather than another among persons of one origin class in comparison
to another (e.g. Breen, 2010, p. 367). Hence, the concept of relative mobility expresses
the differences in chances of access to a destination class between different classes of
origin, and it can be interpreted as ‘the outcome of competition between people from
different class origins to attain more desirable class positions (destinations) and to
avoid less desirable ones’ (Breen & Karlson, 2014, p. 108).
The variation in levels of social mobility over countries and the development of
social mobility over time have attracted particular attention in the study of social mobility. Research is ambiguous regarding whether social fluidity varies strongly between countries and whether there has been an increase in social fluidity over time. In
their analysis of 12 European countries, the USA, Japan, and Australia, Erikson and
Goldthorpe (1992) found few variations between countries and an almost unchanging
level of social fluidity. In contrast, Breen (2004), using data from 11 European countries, reported significant variation between countries and a tendency towards increasing social fluidity in all countries except Great Britain. This is in line with an earlier
study of Ganzeboom, Luijkx, and Treiman (1989) and other cohort-based studies fo-

Research on social mobility and educational inequality

25

cusing on single countries (e.g. Breen & Jonsson, 2007a; Hout, 1988; Mayer & Aisenbrey, 2007; Vallet, 2004). Although there seems to be growing evidence of increasing
social fluidity, social origin remains a strong predictor of class position in all Western
industrialized countries.
Most comparative studies of social mobility that include Germany are limited to
Western Germany for the sake of comparability over time. Eastern Germany has experienced stronger changes in its social structure due to the transition into and out of
the German Democratic Republic (GDR) during which conditions for achieving class
positions had been very different and strongly determined by political loyalty to the
regime of the Socialist Unity Party. In the initial period of the GDR during the 1950s
and 1960s, access to tertiary education of students with working class and farming
parents was promoted strongly through grants and contingents. Afterwards, the opportunities for working class and farmers’ children to obtain higher class positions decreased again. This was the result of a shift in policies towards a stronger emphasis on
performance and economic requirements (Geißler, 1983; Loeffelmeier, 2006; Solga,
1997).
Cross-country comparisons show that Western Germany is one of the most rigid
countries with a strong origin–destination association (Breen & Luijkx, 2004, p. 73;
Erikson & Goldthorpe, 1992, p. 153). This remains the case, although Western Germany has joined the general trend towards increasing social fluidity in cohorts born
since the 1930s (Breen, 2010; Breen & Luijkx, 2007; Mayer & Aisenbrey, 2007;
Müller & Pollak, 2004a). The trend towards increasing social fluidity might be fading
away for cohorts born after the mid-1960s (Breen & Luijkx, 2007; Mayer
& Aisenbrey, 2007), but this is not confirmed by Breen (2010).
In Western Germany, as in other countries, the association between origin and destination class is lower for women (Mayer & Aisenbrey, 2007; Müller & Pollak,
2004a). This is due to women more often attaining a lower class than their fathers.
However, female downward mobility is declining, whereas upward mobility is increasing, so that patterns are becoming increasingly similar for both men and women (Hillmert, 2015; Mayer & Aisenbrey, 2007; Müller & Pollak, 2004a). The reason is that
educational attainment has increased more strongly for women than for men. Also,
against the background of the strongly gender-segregated labour market, the availability of those jobs that women are more likely to choose has increased more strongly
than that of those jobs that men are likely to choose (Müller & Pollak, 2004a, p. 109).

26

Research on intergenerational mobility

Social mobility is closely related to educational attainment. The relationship is often illustrated by the ‘OED triangle’ (Goldthorpe, 2014; see Figure 2.1): social origin
(O) affects educational attainment (E), and the latter affects the social destination (D).
Although there can also be a direct effect of social origin on the access to class positions, education mediates a large part of the origin–destination effect.
Education
E

Origin

Figure 2.1.

O

D

Destination

The OED Triangle

Sociological research is interested in disentangling the direct part of the origin–
destination effect from the part that is mediated by education in order to determine
how far chances of acquiring social positions depend on education in cross-country or
cross-temporal comparisons. Modern societies legitimate the assignment of social positions more strongly by educational attainment. The conclusion that these societies are
more meritocratic, however, should be drawn with caution, because educational attainment itself often depends strongly on social origin, thereby casting doubt on the
meritocratic principle. It is also interesting to disentangle direct and educationmediated effects of social origin in order to examine whether educational expansion

has led to more social fluidity. Educational expansion might affect social fluidity in
two ways (Breen & Jonsson, 2007b). First, because education is the most important
determinant of social positions, educational expansion might equalize access to social
positions if it increases equality in educational attainment. Second, because the education–destination link is stronger at higher levels of education, educational expansion
might equalize access to social positions because a greater share of persons obtain
higher levels of education. Although, as mentioned before, Erikson and Goldthorpe
(1992) found social fluidity to be quite stable over time, other comparative studies on
social mobility have found that educational expansion has indeed increased social fluidity (Breen, 2010; Breen & Luijkx, 2004). This finding also applies to Western Germany, where change in social fluidity can be attributed to educational expansion


Research on social mobility and educational inequality

27

(Breen & Jonsson, 2007b; Müller & Pollak, 2004a). In Western Germany, as in, for
example, France (Vallet, 2004), Sweden, or Great Britain, the origin–destination association tends to be lower at higher levels of education, although it is lowest at the
lower tertiary level rather than at the upper tertiary level (Breen, 2010). This compositional effect of the educational expansion, which stems from the increased share of
persons attaining higher levels of education, is stronger than the effect of equalization
in Western Germany, whereas in Sweden, for example, educational expansion affected
increasing social fluidity more strongly through equalization (Breen, 2010).
Due to the important and increasing role of education in mediating class origins
and destinations (e.g. Breen & Luijkx, 2004; Hillmert, 2015), there is a vast body of
literature concentrating on the link between social origin and educational attainment
(e.g., Breen, Luijkx, Müller, & Pollak, 2009, 2010; Erikson & Jonsson, 1996a; Müller
& Karle Wolfgang, 1993; Pfeffer, 2008; Shavit, Arum, & Gamoran, 2007; Shavit &
Blossfeld, 1993). The starting point of these analyses is the question whether social
inequalities in education have declined in different countries over the course of the
20th century. In favour of decreasing inequalities, one can follow a functionalist view
(Treiman, 1970) and argue that resources are distributed more equally in more industrialized countries, and that education in these societies is more often provided for free.
Thus, through industrialization, education should have become more affordable for the

lower classes, and class differences should decline. According to Breen et al. (2009),
this should apply particularly to the decades after World War II when economic
growth decidedly improved living conditions. During this period, social inequalities
should have been reduced significantly because dispensable income grew strongly
whereas family size declined. As a result, resources available per child rose. The nutrition and health of the lower classes improved and were no longer major drawbacks to
the performance of children with a lower social background. Direct and indirect costs
of education declined. Additional costs of full secondary or tertiary education decreased because compulsory schooling was extended and the proportion of jobs requiring higher levels of education increased (Breen et al., 2009, pp. 1478–1480).
Raftery and Hout (1993), however, found that it is only under certain conditions
that educational inequalities decrease during a period of educational expansion: the
chances of obtaining a certain level of education in one class relative to any other class
remain roughly the same despite the educational expansion as long as the highest class
has not reached a saturation level of nearly 100 per cent. They call this pattern ‘maximally maintained inequality’. Lucas (2001) extended maximally maintained inequality


28

Research on intergenerational mobility

with the concept of effectively maintained inequality. Accordingly, if saturation is
achieved and class differences in the chances of attaining this level decrease, higher
classes find other means to distinguish themselves from lower classes and to ensure
better chances to access higher class positions for their offspring. Given a nearly universal level of schooling, ‘the socioeconomically advantaged seek out whatever qualitative differences there are at that level and use their advantages to secure quantitatively similar but qualitatively better education’ (Lucas, 2001, p. 1652).
Both maximally and effectively maintained inequality are observed patterns rather
than mechanisms that explain the generation and persistence of social inequality in
education. Moreover, functionalist theory, predicting that the education–destination
link will strengthen through industrialization and that, thereby, the origin–destination
link will weaken, is not equipped with micro-social mechanisms. In order to understand how macro phenomena are perceived by individuals and how they trigger certain
actions that have results on the macro level, these macro approaches need to be complemented by micro theories (cf. Goldthorpe, 2014).
One attempt to explain social inequality in education is cultural reproduction theory (e.g. Bourdieu & Passeron, 1964, 1971). This argues that the class structure reproduces itself through the transmission of economic, social, and cultural capital from
parents to their children. Children thus are differently equipped for educational careers

respective to their social origin. Even if economic resources are distributed more
equally, higher classes still find ways to mark their distinction from lower classes
through cultural and social capital. Schools are considered to represent more strongly
the culture of the middle classes, so that students from the lower classes feel less at
ease and have more difficulties in fulfilling their requirements. For example, children
learn strategies from their parents regarding how to interact with teachers in class, and
these affect their success in school. Whereas middle-class parents coach their children
to include their teachers in problem solving, working class parents direct their children
to solve problems on their own without involving the teacher (Calarco, 2014). Second,
in case of difficulties in meeting the educational standards, persons from a higher social origin can rely on their parents’ resources, be it financially by paying for private
lessons, be it socially by using their parents’ contacts, or be it culturally by hiding insufficient skills behind good manners or cultural knowledge.
Other attempts to explain the generation and persistence of educational inequalities
are rational-choice-based theories (e.g. Boudon, 1974; Breen & Goldthorpe, 1997;
Erikson & Jonsson, 1996b; see chapter 3.2 for a more detailed discussion). These as-


Research on social mobility and educational inequality

29

sume that educational inequalities are the result of rational decisions by which actors
maximize utility. Individuals are considered to decide for an educational option based
on their subjective assessments of returns, costs, and probabilities of success. All these
depend on the level of resources related to a certain class position of origin.
Although there is a broad consensus that social inequalities in educational attainment remain strong despite educational expansion, the literature is ambiguous regarding whether social inequalities in education declined during the 20th century. Some
studies find that, despite an impressive and comprehensive uplift across all classes,
differences in educational attainment have not declined in most countries (e.g. Pfeffer,
2008; Shavit & Blossfeld, 1993; see Shavit et al., 2007, regarding post-secondary education). Blossfeld and Shavit (1993) found an equalization of educational attainment in
relation to social background only for Sweden and the Netherlands, whereas social
inequality in education remained stable in the eleven other countries in their analysis.

Moreover, they showed that in some countries, the expansion of secondary education
was accompanied by a growing differentiation into academic and vocational tracks so
that tertiary education remained roughly as exclusive as before.
Focusing on eligibility and access to tertiary education in a study of thirteen countries, Arum, Gamoran, and Shavit (2007) reported that inequality in eligibility to
higher education was stable in nine, increased in one (Italy), and declined in five countries. Regarding inequality in the transition from eligibility to higher education, they
found stable inequality in six countries, an increase in three, and a decline in four.
These findings supported the hypothesis of maximally maintained inequality. In the
case of actual or near saturation (eligibility to tertiary education of 80 to 100 per cent),
inequality in enrolment and access decreased, whereas expansion alone did not lead to
reduced inequality (Shavit et al., 2007, p. 18). The authors classified countries according to the degree of differentiation within the tertiary education system, distinguishing
unitary systems with low differentiation, binary systems with two kinds of tertiary
education institutions, and diversified systems with a variety of different institutions.
Consistent with effectively maintained inequality, they found that expansion is related
to differentiation: Those countries with the highest enrolment rates have diversified
systems that include very heterogeneous institutions and programmes in terms of quality, selectivity, and prestige. However, contrary to expectations, diversified systems
are more inclusive than less differentiated systems. Thus, a higher degree of differentiation does not necessarily lead to greater within inequality. Indeed, a strong tendency
to divert students with a lower social background to the lower tier institutions is found