The risk of downward mobility in educational attainment

Children of higher educated parents also have the highest level of mobility within their educational career, i.e., social origin is not only an impor-tant determinant of educational atta

The Risk of Downward Mobility in Educational Attainment

Herausgegeben von

Prof Dr Steffen Hillmert, Universität Tübingen

Life Course Research

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

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

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

http://www.springer-vs.de/978-3-658-14598-9

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 spe-cific 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 tra-general downward educational mobility In particular, downward mobility is com-mon in families of higher educated parents Though expectable from a formal perspec-tive, this is a remarkable substantive finding in light of considerable educational expan-sion 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 impor-tant determinant of educational attainment but also of educational pathways

in-The message for inequality research is therefore to look not just at selected tions 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

transi-our series, Life Ctransi-ourse Research The series publishes empirical studies – in both

Eng-lish 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 supervi-sor, 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 tions 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 de-fense Furthermore, I thank Steffen Hillmert for admitting this book to appear in the

ques-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 quia of Thorsten Schneider and Hans-Peter Blossfeld

collo-The research was funded by the German Federal Ministry of Education and search (reference number 01JG1059) The author is solely responsible for the content

Re-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 The OED Triangle 26

Figure 5.1 Educational attainment by education of the parents 73

Figure 5.2 Sequence index plots of educational pathways I 75

Figure 5.3 Sequence index plots of educational pathways II 76

Figure 5.4 Educational pathways by education of parents 79

Figure 5.5 Mean durations of education by education of parents 81

Figure 6.1 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 Ideal typical outflow table from origin class to destination class 24

Table 2.2 Ideal typical inflow table from origin class to destination class 24

Table 4.1 Overview of samples and survey instruments for NEPS starting cohort 6 62

Table 5.1 Types of educational pathway 78

Table 5.2 Percentages of types of pathway by education of parents and cohorts 80

Table 5.3 Linear regression of mean duration in education 83

Table 5.4 Logistic regression of reaching a tertiary degree over cohorts (odds ratios) 85

Table 6.1 Standardized mean grade point averages of the tertiary education entrance certificate in different categories of post-secondary activities by education of parents 108

Table 6.2 Post-secondary education of persons with tertiary education entrance certificate by education of the parents 110

14 Figures and Tables Table 6.3 Piecewise-constant exponential model for entry into tertiary

education after tertiary education entrance certificate and first

vocational training 112

Table 7.1 Prevalence of pre-tertiary educational pathways among students at

universities and universities of applied sciences 137 Table 7.2 Students at universities and universities of applied sciences by

education of parents 137 Table 7.3 Piecewise constant exponential models on dropout from

universities within 8 years of starting tertiary studies 141

Table 7.4 Piecewise constant exponential models on dropout from universities

within 8 years of starting tertiary studies – including effects of

educational pathways 144

Table 7.5 Piecewise constant exponential models on dropout from

universities of applied sciences within 8 years of starting tertiary studies 146

Table 7.6 Piecewise constant exponential models on dropout from

universities of applied sciences within 8 years of starting tertiary studies – including effects of educational pathways 148 Table B.1 Distribution over types of education 24 years after school entry 191 Table C.1 Re-entering the academic pathway after starting vocational

training in Germany: Distribution of characteristics in the sample

at certain points in time 193 Table D.1 Dropout from tertiary education in Germany: Distribution of

characteristics in the sample at certain points in time at

universities 197 Table D.2 Distribution of characteristics in the sample at certain points in

time at universities of applied sciences 198 Table D.3 Distribution of age at enrolment in universities 199 Table D.4 Distribution of age at enrolment in universities of applied

sciences 199 Table D.5 Piecewise constant exponential models on dropout from

universities within 8 years of starting tertiary studies –

including all covariates 200 Table D.6 Piecewise constant exponential models on dropout from

universities within 8 years of starting tertiary studies –

including effects of the duration at the Gymnasium 202

Table D.7 Piecewise constant exponential models on dropout from

universities within 8 years of starting tertiary studies –

including effects of educational pathways 204 Table D.8 Piecewise constant exponential models on dropout from

universities within 8 years of starting tertiary studies –

including fields of study 206 Table D.9 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 in-dividual wealth, power, health, and life expectancy Also, education is unevenly dis-tributed among individuals within a society Inequalities refer to the access to vertical levels of educational programmes, degrees, or assessment results and to horizontal dif-ferences between types of educational programmes Both, vertical and horizontal dif-ferences 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 legiti-mate if they are linked to a person’s level of ability and effort Accordingly, more tal-ented 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 han-dling tasks in society and to a higher level of overall wealth

However, in reality, educational success such as moving to a higher level of tion, receiving good grades, or obtaining educational degrees does not solely depend

educa-on talent and effort There are structural determinants such as the level of tion 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

differentia-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 look-ing 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 Introductiontion 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 mobil-ity 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, oc-cupational 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 verti-cal order Furthermore, social mobility invokes the notion that the parents’ status de-fines a starting point (social origin) whereas the achieved status is the point of destina-tion ‘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) genera-tions can be expected for several reasons One reason is that children inherit their par-ents’ 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, chil-dren of higher-status-parents more easily obtain higher degrees and positions them-selves 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 occu-pational 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 instru-ments 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 envi-ronments 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 suf-fer 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 fre-quent than downward mobility Against this background, one important research ques-tion is whether the opening up of higher levels of education to all and the rise in edu-cational 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 be-tween 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 educa-tional 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 par-ticular in higher classes of origin For example, Holtmann (2010) states that in West-

20 Introductionern Germany only about half of the children with tertiary educated parents of birth co-horts 1920 to 1979 reaches the same education level as their parents Second, the con-cept 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 pos-ited 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 de-gree is an important motivating force for re-entering education and thus a major de-terminant 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

educa-tional careers In the analyses I focus on downward mobility in educaeduca-tional attainment

because this is the core part of the origin–destination relation Furthermore, I trate 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 analy-sis In the German education system, the divide between educational careers that lead

concen-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, exclu-sively 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, generational moves in the educational level from parents to children and intra-

inter-generational 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 down-wardly 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 sys-tem 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 educa-tional 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 intergenera-tional social and educational mobility and present previous research on intergenera-tional 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 Chap-ter 4 provides information on the data used in the empirical analyses

22 IntroductionThe 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 differ-ent 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 transi-tion 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 en-trance 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 pre-vents 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 at-tenuated

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 re-search 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 sociol-ogy 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 ori-gin 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 destina-tion 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 Research on intergenerational mobilityTable 2.1 Ideal typical outflow table from origin class to destination class

Destination class (child's occupation) Origin class (Highest

Table 2.2 Ideal typical inflow table from origin class to destination class

Destination class (child's occupation) Origin class (Highest

occupation of the parents)

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 mo-bility Research is ambiguous regarding whether social fluidity varies strongly be-tween 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 coun-tries, reported significant variation between countries and a tendency towards increas-ing 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 & Aisen-brey, 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 ex-perienced 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 oppor-tunities for working class and farmers’ children to obtain higher class positions de-creased 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 Ger-many 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 tination 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 increas-ing, so that patterns are becoming increasingly similar for both men and women (Hill-mert, 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 availabil-ity 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)

des-26 Research on intergenerational mobilitySocial mobility is closely related to educational attainment The relationship is of-ten 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 posi-tions, education mediates a large part of the origin–destination effect

Figure 2.1 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 po-sitions more strongly by educational attainment The conclusion that these societies are more meritocratic, however, should be drawn with caution, because educational at-tainment itself often depends strongly on social origin, thereby casting doubt on the meritocratic principle It is also interesting to disentangle direct and education-mediated 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 educa-tion–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 flu-idity (Breen, 2010; Breen & Luijkx, 2004) This finding also applies to Western Ger-many, where change in social fluidity can be attributed to educational expansion

Education

Destination Origin

E

D

O

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 asso-ciation 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 composi-tional 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 indus-trialized 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 nutri-tion 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 creased because compulsory schooling was extended and the proportion of jobs requir-ing higher levels of education increased (Breen et al., 2009, pp 1478–1480)

de-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 ‘maxi-mally maintained inequality’ Lucas (2001) extended maximally maintained inequality

28 Research on intergenerational mobilitywith 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 uni-versal level of schooling, ‘the socioeconomically advantaged seek out whatever quali-tative differences there are at that level and use their advantages to secure quantita-tively 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 under-stand 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 com-plemented by micro theories (cf Goldthorpe, 2014)

One attempt to explain social inequality in education is cultural reproduction ory (e.g Bourdieu & Passeron, 1964, 1971) This argues that the class structure repro-duces 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

the-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 cial 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 in-sufficient skills behind good manners or cultural knowledge

so-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 ment remain strong despite educational expansion, the literature is ambiguous regard-ing 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 edu-cation) 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

attain-Focusing on eligibility and access to tertiary education in a study of thirteen tries, 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 coun-tries 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 accord-ing 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

coun-to differentiation: Those countries with the highest enrolment rates have diversified systems that include very heterogeneous institutions and programmes in terms of qual-ity, selectivity, and prestige However, contrary to expectations, diversified systems are more inclusive than less differentiated systems Thus, a higher degree of differen-tiation 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

30 Research on intergenerational mobility

in the binary systems that differentiate mainly between academic and occupationally oriented higher education

Whereas the aforementioned studies placed more emphasis on the persistence of inequalities in educational attainment, other studies (e.g Breen & Jonsson, 2005; Breen et al., 2009; Henz & Maaz, 1995; Müller & Haun, 1994; Shavit & Westerbeek, 1998) have provided evidence that class inequalities declined during the last century Breen et al (2009) found decreasing inequality in educational attainment for the birth cohorts 1935 to 1954 in all nine European countries included in their comparative analysis This fits in with their expectation that inequalities in education have declined due to improvements in living conditions and economic growth in the post-World War

II decades The decline in inequality seems to be due primarily to the improvement of the educational chances of children from farming and working class origins (2009,

p 1514)

A robust finding in cross-national comparisons of social inequalities in educational attainment is that the effect of social origin tends to decline from earlier to later transi-tions in the educational system (e.g Blossfeld & Shavit, 1993; Jackson & Jonsson, 2013) There are different explanations for this phenomenon According to the differ-ential selection hypothesis (Mare, 1980), students with a lower social background but the same level of ability have fewer chances of proceeding to the next level Thus, those who proceed upwards have, on average, higher levels of ability, motivation, or other characteristics that are rewarded in the educational system than the levels to be found in their counterparts with a higher social origin If it is not possible to control for these characteristics, differences in transition probabilities will therefore decrease at subsequent levels According to this argumentation, decreasing effects of social origin are, therefore, a statistical artefact due to unobserved heterogeneity A second explana-tion is provided by the life-course hypothesis (Blossfeld & Shavit, 1993) This posits that parents are strongly involved in educational decisions when their children are young, but become less involved as their children grow up because of their increasing autonomy Another reason is that uncertainty regarding the future performance of stu-dents is higher at younger ages Therefore, when assessing what is feasible for their children at these ages, parents refer more strongly to their own experience in the edu-cational system At later decisions, in contrast, students have gained their own experi-ence in the education system; and, as a result, decisions are less influenced by parental background This hypothesis postulates a genuine reduction of social background ef-fects over the life course

Research on social mobility and educational inequality 31 Although social inequalities in educational attainment exist in all countries and tend to be stronger at earlier transitions than at later ones, there are considerable cross-country differences These differences relate to differences in the institutional charac-teristics of educational systems The degree of stratification of an educational system

is one of the characteristics that are considered to affect the level of social inequality Stratification is defined by ‘the proportion of a cohort selected to attain the maximum number of school years provided by the system’ (Allmendinger, 1989, pp 234–235) , and this relates to the vertical differentiation introduced through tracks or schools of different quality (Jackson & Jonsson, 2013, p 308) In highly stratified systems, only a small proportion of students reach the highest educational level Countries with more stratified educational systems have been found to have higher levels of social inequal-ity (Pfeffer, 2008)

Jackson and Jonsson (2013) examined how countries with different institutional characteristics differed with regard to social inequality in performance (primary ef-fects) and choice (secondary effects) In addition to stratification, they introduced se-lectivity as a characteristic of educational systems defined as ‘the degree to which track placement is a function of previous school performance rather than the free choices of students and their parents’ (Jackson & Jonsson, 2013, p 308) They argued that stratification and selectivity should affect choice conditional on performance rather than performance in itself Also, they assumed that higher stratification should relate to stronger inequality in the choice of educational options conditional on per-formance, because track choices are more consequential in highly stratified systems and, thus, upper class parents strongly support the choice of more advanced educa-tional options As for selective systems, they assumed the opposite effect: strong selec-tion based on performance should limit the effect of free choice and, thus, reduce the effect of parental background (Jackson & Jonsson, 2013, p 311) However, although they found that some highly selective and strongly stratified systems such as Germany (including Eastern Germany) and the Netherlands do indeed have higher levels of edu-cational inequality at the first transition, whereas countries with weak selection and stratification have lower levels of inequality, the group of countries classified as in-termediate in both categories did not show the expected results (Jackson & Jonsson,

2013, pp 319–320) As the authors acknowledged, the contradicting effects of tion and stratification are difficult to disentangle because both dimensions are interre-lated in the sense that more stratified systems are also more selective (Jackson

selec-& Jonsson, 2013, p 329)

32 Research on intergenerational mobility

In a cross-country comparison, Western Germany shows high social inequalities (e.g Pfeffer, 2008; Shavit & Blossfeld, 1993), but, according to Breen et al (2009), these seem to have declined within the cohorts born between 1925 and 1945 With re-spect to social inequality in educational attainment, Western Germany resembles France, Italy, and Ireland, which had larger class inequalities at the beginning of the 20th century and a considerable decline afterwards, rather than countries with tradi-tionally lower class inequalities such as Great Britain, the Netherlands, or Sweden (Breen et al., 2009, p 1513) The first transition in Germany takes place at 10 to 12 years, which is comparatively early (see Figure A.1 for an overview on the German education system) According to Jackson and Jonsson (2013), social inequality is, therefore, particularly high at the first transition although not as high as in Italy As in other countries, social inequality is higher at the first transition than at the transition to

tertiary education for the population of pupils at the Gymnasium However, secondary

effects (i.e effects of social origin conditional on performance) are higher at the tion to tertiary education than at the first transition These secondary effects at the transition to tertiary education are high in cross-national comparison, surpassed only in Italy and the Netherlands (Jackson & Jonsson, 2013)

transi-Because educational attainment during educational expansion increased more strongly for women than for men, it is interesting to ask whether social inequalities in educational attainment have developed differently according to gender Following Breen and Goldthorpe (1997, p 297), social inequality in education should have in-creased more (or decreased less) for women than for men They argued that women’s access to class positions in earlier decades of the 20th century had relied more strongly

on marriage In the subsequent decades, education has become an increasingly tant determinant of women’s class position Thus, women’s class inequalities in educa-tion should have become more similar to those of men Another explanation for differ-ences in the educational inequalities of men and women is that parents might invest differently in the education of their sons compared to their daughters, especially in the classes of farmers and the self-employed (Müller & Haun, 1994) This is based on the reasoning that men in these classes are more likely to inherit the farm or family busi-ness, making them less dependent on education, whereas the siblings who do not in-herit are more dependent on education

impor-In their analysis of seven European countries, Breen et al (2010) found that trends

in the social inequality of education for women were quite similar to those for men Significant gender differences in the development of social inequalities over cohorts

Research on intergenerational downward mobility 33 were found only in Italy and Poland In these countries, inequalities diminished more strongly for women than for men As expected, the male advantage in educational at-tainment decreased over time in all countries In Western Germany, gender differences

in upper secondary or lower tertiary education declined comparatively late, but almost disappeared in the youngest cohort born 1955 to 1964 The authors also found support for the hypothesis that, relative to parents from other classes, business- and farm-owning parents invested more strongly in the education of daughters compared to sons These class-specific gender differences seemed to be quite stable over time The main findings from comparative research on social and educational mobility reviewed in this chapter can be summarized as follows: studies on social mobility and

on social inequality in education are divided over the question whether there has been

a trend towards more equal chances to obtain educational levels and social class tions However, according to the more recent studies using larger datasets and more advanced methods, evidence seems to lean towards a decrease in inequalities Apart from this dividedness, research agrees that there continue to be large differences in educational and occupational chances based on social origin, that education plays the major role in the link between origin and destination class in modern societies, that earlier transitions in education are influenced more strongly by social origin than later ones, and that female patterns of social mobility and social inequality in education are becoming increasingly similar to male patterns Germany, with its very selective and stratified educational system, has been found to have a comparatively high level of social inequality in education and rather low levels of social mobility

posi-2.2 Research on intergenerational downward mobility

As described above, both theoretical and empirical analyses have invested much effort

in explaining and analysing the stability of educational or occupational chances by social origin or, in the case of changes, their upward mobility However, less attention has been paid to the portion of persons who do not attain the educational level or the class position from which they originate Nonetheless, downward mobility is ad-dressed as a minor topic in some studies, and a few studies even focus on it For ex-ample, studies on social mobility that report absolute mobility rates usually also in-clude rates of downward mobility These give some idea on the extent of downward mobility in relation to upward mobility and stability Thereby, it has to be considered that the extent of downward mobility also depends on the underlying classification

34 Research on intergenerational mobilityscheme of classes and on whether classes are combined for the sake of clarity or be-cause they contain low numbers of cases A widely used classification is the EGP (Erikson–Goldthorpe–Portocarero) class scheme, also referred to in the literature as the Goldthorpe, or CASMIN (Comparative Study of Social Mobility in Industrial Na-tions) class scheme

When describing the intergenerational class mobility of men, Breen and Luijkx (2004, pp 47–48)showed that the mean percentage of downward mobility in 10 Euro-pean countries and Israel ranged from 18.7 per cent in the 1970s to 16.2 per cent in the 1990s Upward mobility was higher, lying between 28.0 and 33.4 per cent Thus, ver-tical mobility is composed of about one-third downward mobility and about two-thirds upward mobility Among the countries considered in this study, Germany had the low-est rates of downward mobility, ranging from 12.4 in the 1970s to 13.0 per cent in the 1990s, whereas its rates of upward mobility were about average and in line with the other countries Hungary started with the highest percentage of downward mobility in the 1970s (26.2 per cent), but this decreased strongly to 17.8 per cent in the 1990s In the same time period, upward mobility increased in Hungary from 26.9 to 35.9 per cent In general, Breen and Luijkx (2004, p 49) reported moderate change over time and high similarity among countries on all measures of mobility

Based on a seven-category version of the Goldthorpe class schema, Goldthorpe and Jackson (2007) reported that male downward mobility in social class in Great Britain attained 26.5 per cent in the cohort born in 1958 and 29.6 per cent in the cohort born in 1970 against upward mobility rates of 45.2 and 42.2 per cent in the same co-horts Among the women in these cohorts, 36.8 per cent and 34.5 per cent were downwardly mobile, whereas 39.0 per cent and 40.8 per cent were upwardly mobile

As for Western Germany, Pollak (2013) reports that the downward mobility of men increased slightly from 15 to 17 per cent for cohorts born 1976–1980, 1981–1990, 1991–1999, and 2000–2010 For women in Western Germany, downward mobility decreased from 33 per cent in the 1976–1980 cohort to 27 per cent in the 2000–2010 cohort Similar to Goldthorpe and Jackson (2007) above, this study showed a slight trend towards increasing similarity between men and women due to decreasing down-ward mobility in women However, the decrease is not very strong and women still show higher rates of downward mobility

Hillmert (2015) finds rather stable downward mobility in occupational positions of parents and their children for the eight cohorts born between 1919 and 1971 of the West German Life History Study Regarding women there is a longterm decrease be-

Research on intergenerational downward mobility 35 tween cohorts born 1929–1931 and those born 1954–1956, however, there is not much variation in younger cohorts born after World War II

As well as class mobility, educational mobility also depends partly on the

underly-ing categories The followunderly-ing study uses the International Standard Classification of Education (ISCED) 1997, collapsing level four (Post-secondary non-tertiary educa-tion) and 5 (First stage of tertiary education) into one category Focusing on Western Germany, an analysis of educational mobility based on data from the German Socio-Economic Panel (SOEP) including persons born from 1938 to 1962 showed that about

15 per cent of persons over all classes attained a lower educational degree than their parents (Fuchs & Sixt, 2007a, p 12) When at least one parent had a tertiary degree, chances of downward mobility even reached 46 per cent (Fuchs & Sixt, 2007a, p 13) Holtmann (2010) using SOEP data including persons born in Western Germany from

1920 to 1979 found that 12.5 per cent of all persons and 47.8 per cent of persons whose parents had a tertiary degree did not reach their parents’ qualification level (see also: Autorengruppe Bildungsberichterstattung, 2012, p 213) This study distinguishes between the categories ‘no qualification’, ‘lower school diploma’, ‘intermediate school diploma’, ‘tertiary education entrance diploma’, and ‘university/university of applied sciences degree’ This has the disadvantage that it generally does not include other professional qualifications apart from tertiary degrees

The OECD study ‘Education at a Glance 2012’ reported that, in Germany, 22 per cent of non-students aged between 25 and 34 years did not attain their parents’ educa-tional degree According to this study, Germany was the only country apart from Esto-nia and Iceland in which downward mobility in educational attainment was higher than upward mobility (OECD, 2012a, 2012b, pp 3–4) Compared to the above-mentioned studies, this percentage is rather high One reason for these differences might be that the OECD study included more recent birth cohorts Another reason, however, might

be the grouping of educational categories used when measuring intergenerational bility These distinguished between ‘low levels of education (ISCED levels 0–2 com-pleted, the person has not completed upper secondary education); mid-levels of educa-tion (ISCED levels 3–4 completed, the person has completed upper secondary or post-secondary, non-tertiary education); and high levels of education (ISCED levels 5–6 completed, the person has completed tertiary education)’ (OECD, 2012c, p 65)

mo-In summary, absolute rates of downward mobility in Western industrialized tries usually seem to be above 10 per cent; they usually make up about one-third of total mobility Downward mobility in women seems to be decreasing, whereas down-

coun-36 Research on intergenerational mobilityward mobility in men seems rather stable Germany, being a country with rather low social mobility, has comparatively low levels of downward class mobility Neverthe-less, even in Germany, based on the above-mentioned percentages (Müller & Pollak, 2004a, p 91), about one in every eight men and one in every five women does not reach the social class of their parents Downward mobility in educational attainment seems to reach similar levels: about one in every seven to eight men or women does not reach the parents’ educational level; regarding persons with tertiary educated par-ents, nearly every second person is downwardly mobile (see the above-mentioned per-centages reported by Fuchs & Sixt, 2007a; Holtmann, 2010)

One of the rare studies on intergenerational downward mobility examined the cational careers of persons in Switzerland with at least one parent who had a tertiary degree (Schmeiser, 2003) Based on 30 qualitative interviews, Schmeiser (2003) dis-tinguished four types of downward mobility careers in education The first type de-scribes careers in which downward mobility is delayed as long as possible but, finally, takes place very abruptly Persons classified to this type take it for granted that they will complete a tertiary degree They keep following this path despite being increas-ingly over-challenged because they fear being cast out of their social environment Institutional barriers bringing about, for example, dropout from tertiary education due

edu-to weak performance finally demarcate a turning point At this point, they can no longer deny their inability to meet demands The second type is characterized by an early breaking away from the career leading to a tertiary degree Cases in this type cope with the cross-pressure situation by completely rejecting the values and norms of their social environment of origin and integrating into an alternative milieu with dif-ferent values and norms Within this alternative social environment, as Schmeiser (2003, p 225) observes, they nonetheless once more wish to prove successful The third type comprises persons who integrate in neither the social environment of their origin nor that of their destination Similarly to persons assigned to the first type, their aspirations are higher than their performance However, in contrast to the first type, this type does not ignore failures to meet the demands of school Failures are typically attributed to external reasons Cases assigned to this type do not accept a lower occu-pational career They typically try to catch up by investing in further education None-theless, they do not feel at ease working in their environment of origin because of feel-ings of inferiority The fourth type summarizes persons whose parents are of two dif-ferent social origins and/or are divorced and who are, consequently, following occupa-tional careers that switch between both classes The parents’ expectations that their

Research on intergenerational downward mobility 37 child should attain a high social status are lower in this type Changes in the career result from changing loyalties to either of the parents The first three types can be con-sidered as different approaches to coping with the cross-pressure situation of high de-mands and low performance in the educational system According to Schmeiser (2003,

pp 228–230), consequences of downward mobility differ between the ideal types In the case of the first type, downward mobility leads to social isolation because of the long period during which the persons conceal their inability to cope with the chosen career In the case of the second type, in contrast, persons are well integrated in an al-ternative community Downward mobility in the third type prevents integration into either the social environment of origin or that of destination In the fourth type, in con-trast, there are no difficulties with social integration In summary, Schmeiser (2003) documented the influence of career expectations related to social origin on the educa-tional and occupational career of persons with highly educated parents He showed that the dissonance between individual ability and expectations of status maintenance shapes life courses in a specific way with, in most cases, strong consequences for so-cial integration

Another study on intergenerational downward mobility concentrated on the ity of higher white-collar to lower white-collar and blue-collar positions using data from the Stockholm Birth Cohort Study (Alm, 2011) Based on the concept of cultural capital and habitus, Alm (2011) argued that upward mobility of parents and lower lev-els of education should increase risks of downward mobility According to cultural capital theory, higher levels of education are related to higher amounts of cultural capi-tal that can be transmitted to children Furthermore, individuals in each social class share a habitus, that is, class-specific dispositions, attitudes, and perceptual schemas (Bourdieu, 1982, pp 278–283)that affect the individual’s thinking and acting The habitus is considered to ease interactions with other persons from the same class and to constitute a barrier between persons of different classes Persons who are upwardly mobile adapt to the new habitus; however, they do this only slowly In line with the Don Quixote or hysteresis effect (Bourdieu, 1987, pp 116–117), upward mobility in the parents should increase the risk of downward mobility in their children, because upwardly mobile parents have not entirely internalized the new habitus and do not possess the same amount of cultural capital as parents who were born into that class Multivariate analyses revealed that the effect of parental upward mobility was not sig-nificant in this study (Alm, 2011) However, the level of education of white-collar par-ents had a strong effect on their children not reaching a white-collar position This ef-

mobil-38 Research on intergenerational mobilityfect even remained after controlling for the children’s academic ability The study also included direct measures of cultural capital (number of books in the household) and habitus (parental attitudes towards theoretical knowledge and their preferences con-cerning the child’s future occupation) Although attitudes to theoretical knowledge and occupational preferences impacted on downward mobility, they did not mediate the influence of parental education Thus, the results did not support the hypothesis that lower risks of downward mobility in the children of better educated parents are due to cultural capital

Based on similar theoretical arguments, Fuchs and Sixt (2007a) used data from the German SOEP to examine the chances of attaining a tertiary education entrance cer-tificate in children with upwardly mobile parents compared to children of immobile parents In contrast to Alm (2011) (regarding Sweden), they found a significant effect

of upward mobility in the expected direction in Western Germany: children of wardly mobile parents have lower and children of downwardly mobile parents have higher chances of attaining a tertiary education entrance certificate than children of immobile parents Cultural, social, and economic capital – as operationalized in this study – affects the chances of the children, but it does not completely explain the ef-fects of parental upward and downward mobility Nonetheless, Fuchs and Sixt (2007a) argue that persons who are upwardly mobile are more likely to choose a partner with a lower, and those who are downwardly mobile a partner with a higher educational level Furthermore, persons who are upwardly mobile gain lower and those who are downwardly mobile gain higher rates of return from their qualification level These are additional determinants that have consequences for the level of resources available for the child and, thus, for that child’s educational chances Indeed, when additionally in-cluding the relative educational level of the partner and the relative prestige of the first occupational position, mobility of the parent ceases to be significant Thus, together with cultural, economic, and social capital, these determinants seem to explain the negative impact of upward mobility on chances of attaining a tertiary education en-trance certificate From the finding that children of upwardly mobile parents have lower educational chances than children of immobile parents, Fuchs and Sixt (2007a) conclude that educational upward mobility induced by educational expansion is only partially sustainable beyond one generation

up-In a reanalysis, however, Becker (2007) challenged this result and its conclusions Using data from the German Life History Study, he found – in contrast to Fuchs and Sixt (2007a) – a positive effect of upward mobility of parents on children’s chances of

Research on intergenerational downward mobility 39 gaining a tertiary education entrance certificate or an intermediate degree The most important difference is that Becker’s (2007) analysis included the educational level of the grandparents instead of the parents He still controlled for the educational level of the parents; however, he did this by distinguishing whether upwardly mobile parents attained an intermediate or a tertiary education entrance certificate As well as showing that chances of attaining a tertiary education entrance certificate are higher if grand-parents have higher educational degrees, Becker (2007) found that chances are higher when parents are upwardly mobile, especially when they have attained a tertiary edu-cation entrance certificate themselves Based on additional analyses, Becker (2007) argued that the contrary results in Fuchs and Sixt were due to multicollinearity be-tween the parental educational level and their upward mobility Unlike Fuchs and Sixt (2007a), Becker based his theoretical argumentation on Boudon (1974), arguing that educational chances depend on the motive to maintain the parents’ social status and on resources related to the social position of parents In line with this argumentation, he found that educational upward mobility increases the children’s educational chances when it is related to upward mobility in class positions

Responding to Becker’s (2007) critique, Fuchs and Sixt (2007b) pointed out that what seem at first glance to be contrary results actually refer to different reference groups and, therefore, do not actually contradict each other Compared to those who stayed at the lower educational level, upwardly mobile parents pass on better educa-tional chances to their children However, they pass on lower chances of educational attainment compared to those who had already reached the higher educational level a generation before (Fuchs & Sixt, 2007b)

The studies summarized above (Becker, 2007; Fuchs & Sixt, 2007a, 2007b) stress the importance of intergenerational mobility and show its consequences for the next generation Alm (2011) and Fuchs and Sixt (2007a, 2007b) base their theoretical ar-gumentation on the idea of the inertia of the habitus and resources connected to social origin that increases or reduces those resources that are related to the educational level

of destination Becker follows Boudon (1974), arguing that educational qualification translates into social status that is related to resources These can be invested in the education of children and are supported by high motivation to maintain this status

In the following, I shall focus on intergenerational downward mobility in tional attainment, that is, on the link between the level of educational qualification a person gains and the educational level of her or his parents Intergenerational educa-tional mobility can be considered as the core of class mobility, because the transfer of

educa-40 Research on intergenerational mobilityeducational attainment from one generation to the next is the most important mecha-nism of intergenerational social mobility (Breen & Jonsson, 2005) Furthermore, I shall concentrate on persons whose parents have the highest educational degrees, that

is, tertiary degrees This group is particularly interesting with regard to downward bility because – although they have the highest relative educational chances due to their social background – they have the highest structural risks of downward mobility due to their need to complete the highest and most demanding educational careers in order to reach the educational level of their parents

mo-There are different ways to measure social origin, and the education of the parents

is only one of them Other common measures of social origin are social class or social prestige However, following Pfeffer (2008, p 544), parental education is a strong de-terminant of parental social class and prestige and can be considered to temporally precede other measures of social origin Furthermore, it usually has the strongest direct effects on children’s educational attainment

3 Theoretical approaches: The life-course perspective and rational-choice-based theories of educational decisions

Theoretical approaches

Having shortly summarized research on social and educational mobility in general and

on intergenerational downward mobility in particular in the last chapter, this chapter introduces micro theories that provide mechanisms to explain the reproduction of edu-cational chances from one generation to the next The theoretical framework of this study relies mainly on two approaches The first is the life-course approach, which provides a broader analytical frame through its longitudinal perspective The second is

a rational-choice-based approach to educational decisions, which provides a useful tool for modelling the generation of educational outcomes centred on the individual as actor The combination of both approaches makes it possible to analyse the social em-bedding of individual educational careers over time and their aggregate outcomes It also makes it possible to set up empirically testable mechanisms The following sec-tions describe the approaches in more detail

3.1 The life-course perspective

Life-course research focuses on the analytical construct of the life course as a quence of socially defined events and roles that the individual enacts over time’ (Giele

‘se-& Elder, 1998, p 22) It is an approach with many intersections of different disciplines such as anthropology, demography, developmental psychology, economics, and soci-ology (Mayer, 2009, p 414) The sociological point of view focuses on ‘the syn-chronic and diachronic distribution of individual persons into social positions across the lifetime’ (Mayer, 2004, p 163) Hence, it addresses the patterns and variation of sequences of socially meaningful transitions that occur over the life time of individu-als Several authors have described principles of life-course research (e.g Elder, 1998,

pp 3–4; Kok, 2007; Mayer, 2009, p 414) They can be summarized into four major characteristics: these are the role of time, the embedding of individual lives in proc-esses at different levels, the interdependency between different spheres of the life course such as family life and work, and the links between the lives of individuals As regards time, different dimensions are important: individual time and historical time (Elder, 1994, pp 5–6) Individual time refers to age Whereas biological age deter-

© 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_3

42 Theoretical approachesmines broad boundaries for some transitions such as having a child, research finds that

a culture-specific consensus on when certain transitions should take place narrows down these boundaries Societies are organized in an age-structured way with norms prescribing age limits for the access to resources, obligations, and rights such as voting

or receiving a pension Age norms, however, can also have an informal character in prescribing which behaviour is considered appropriate for individuals at which age and

at which time along with in which order life events and transitions should occur tersten & Mayer, 1997, p 235) They ‘create predictable, socially recognized turning points that provide roadmaps for human lives and outline life paths’ (Hagestad & Neugarten, 1985, p 35) Age is also related to social and cultural expectations about life experiences and roles (Settersten & Mayer, 1997, p 235) Individuals internalize these normative timetables and age-related expectations so that individual transitions are assessed as being on time, too early, or too late (Hagestad & Neugarten, 1985,

(Set-p 41) For example, there are commonly shared perceived deadlines marking up to when persons should have completed school, married, and had children (Billari et al., 2011; Settersten & Hagestad, 1996) or up to when parents should support their chil-dren (Hartnett, Furstenberg, Birditt, & Fingerman, 2013) These conceptions are flexi-ble norms in the sense that they allow for variation and do not result in strong sanc-tions if they are not followed (Settersten & Hagestad, 1996) Being on time, however, may reduce stress, because scheduled transitions can be rehearsed or eased by antici-patory socialization and because peers can provide support (Hagestad & Neugarten,

1985, p 41) High probabilities of making a transition in a certain age span, however, are not necessarily due to social norms This is obvious, for example, in the transition

to widowhood in which the death of the spouse is not the result of norm-driven iour In other cases, as in delays of transitions over cohorts, it is difficult to disentangle the influences of norms from those of changing opportunities (Hagestad & Neugarten,

behav-1985, p 53)

Historical time is another important dimension of time in life-course research cause historical events, circumstances, and conditions affect peoples’ lives They may affect several age groups within a certain period similarly (period effect) An eco-nomic crisis, for example, affects an entire population However, it also affects persons

be-at different ages in different ways (cohort effect) For example, it increases risks of unemployment more strongly for persons at the beginning of a labour market career than for those who are established in the labour market and protected by labour protec-tion laws The cohort, which can be defined as an ‘aggregate of individuals (within

The life-course perspective 43 some population definition) who experienced the same event within the same time in-terval’ (Ryder, 1965, p 845), is an important concept with which to explain social change over historical time (Ryder, 1965) Often cohorts are defined on the basis of birth, but other events such as the year of school completion, marriage, or entry into the labour market may also serve as the defining event of a cohort The idea behind the concept is that changes in circumstances modify people of different ages in different ways, and that these transformations persist throughout the life course The concept of cohort makes it possible to analyse how a group of people who experience the same event in a similar period of life proceed through the life cycle (intracohort analysis) and to compare successive cohorts regarding specific transitions (intercohort analysis) (Ryder, 1965, p 861) The latter is a useful approach with which to grasp social change

With regard to both individual time and historical time, the life-course perspective stresses path dependence as an important issue Prior resources and experiences deter-mine later choices and opportunities ‘Individuals’ and generations’ pasts facilitate and constrain their futures’ (Mayer, 2004, p 166) For example, Alexander, Entwistle, and Kabbani (2001) have shown how dropout in high school is strongly affected by ex-periences and resources in pre-high school years Or, with a greater emphasis on the cohort level, Mayer and Schulze (2009) have compared the effects of German reunifi-cation on the life courses of persons born 1971 in East and West Germany

Besides its focus on temporality, another characteristic of the life-course tive is to build on the individual as the initial point and unit of analysis It is interested

perspec-in how perspec-individual behaviour is embedded perspec-in social structure and perspec-in its resultperspec-ing terns at an aggregate level Thus, it analyses ‘social structure at the level of individual action’ (Kohli, 2007, p 253) Life-course research takes into account cultural frames and institutional and structural conditions at different levels Institutions can be con-sidered thereby as filters (Blossfeld, Mills, Klijzing, & Kurz, 2005) that channel struc-tural conditions to groups of individuals Also, the life course itself can be considered

pat-as an institution ‘regulating one’s movement through life both in terms of a sequence

of positions and in terms of a set of biographical orientations by which to organize one’s experiences and plans’ (Kohli, 2007, p 255)

A third major characteristic of the life-course perspective is the perception of terdependencies between different spheres of life such as education, work, and family life With respect to differences in the life courses of women and men, for example, three types of interdependencies can be distinguished: sequential, simultaneous, or

in-44 Theoretical approachesparallel and adjacent institutionalization (Krüger & Levy, 2001, pp 159–162) Sequen-tial institutionalization occurs, for example, when gender-specific education choices channel persons into fields of occupations that are linked to specific opportunities and constraints such as possibilities of flexible work arrangements or levels of income These relate to later decisions on the division of labour within the family and thereby account for gender-specific family arrangements Simultaneous institutionalization indicates that work and family life are interdependent For example, the amount of time invested in work is related not only to the time invested in childcare and the fam-ily household but also to the investments of the partner and to the needs of other household members Adjacent institutionalization originates in external institutions that determine opportunities and constraints on family, work, life, and education such

as public transportation, childcare institutions, and public administration Thus, in der to better understand life courses, it is necessary to take into account interactions between different spheres of social life (Krüger & Levy, 2001)

or-The life-course perspective also stresses that individuals are influenced not only by constraints and opportunities in different spheres but also by interdependencies with other individuals to whom they are related by close social ties (‘linked lives’ Elder,

1994, p 6) This implies that the lives of persons in social relationships are often terwoven in the sense of career contingencies, counter-transitions, and shared life chances (Hagestad & Neugarten, 1985, pp 50–52) Parents’ resources and decisions affect the lives of their children Examples are long-term effects of parental divorce on children’s lives, or ‘life cycle squeezes’ that occur if the family income does not in-crease at the same pace as children’s needs (Hagestad & Neugarten, 1985, p 51) Also, children influence their parents’ lives even when they are grown up Adult chil-dren returning to live in their parents’ home, for example, cause changes in their par-ents’ lives These interdependencies concern not only intergenerational but also in-tragenerational ties Couples take decisions together, and the decisions of one partner regarding, for example, the work career affect the other Thus, individual behaviour can often be explained more comprehensively when the behaviour of other individuals

in-is taken into account

3.2 Educational decisions

Although the life-course perspective constructs a frame within which educational reers can be analysed, it does not include a theory of action that can be used to con-