Private Pensions Versus
-
Upload
wolandowski -
Category
Documents
-
view
234 -
download
6
description
Transcript of Private Pensions Versus
-
Private Pensionsversus SocialInclusion?Non-State Provision for Citizens at Risk inEurope
Edited by
Traute Meyer and Paul Bridgen
University of Southampton, UK
Barbara Riedmller
Freie Universitt Berlin, Germany
Edward ElgarCheltenham, UK Northampton, MA, USA
-
The editors and the contributors severally 2007
All rights reserved. No part of this publication may be reproduced, stored ina retrieval system or transmitted in any form or by any means, electronic,mechanical or photocopying, recording, or otherwise without the priorpermission of the publisher.
Published byEdward Elgar Publishing LimitedGlensanda HouseMontpellier ParadeCheltenhamGlos GL50 1UAUK
Edward Elgar Publishing, Inc.William Pratt House9 Dewey CourtNorthamptonMassachusetts 01060USA
A catalogue record for this bookis available from the British Library
Library of Congress Cataloguing in Publication Data
Private pensions versus social inclusion? : non-state provision for citizensat risk in Europe / edited by Traute Meyer, Paul Bridgen, Barbara Riedmller.
p. cm.Includes bibliographical references and index.1. Old age pensionsEuropeCase studies. 2. Social
securityEuropeCase studies. 3. PensionsEuropeFinanceCase studies.4. Older peopleGovernment policyEurope Case studies. 5.PoorEuropeCase studies. 6. Public welfareEuropeCase studies. 7.PrivatizationEurope. I. Meyer, Traute. II. Bridgen, Paul. III.Riedmller, Barbara, 1945
HD7164.5.P69 2007331.252094dc22
2007016041
ISBN 978 1 84720 353 3
Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall
-
Contents
List of gures viList of tables viiiList of contributors xList of abbreviations xiiiAcknowledgements xiv
PART I INTRODUCTION
1 Private pensions versus social inclusion? Citizens at risk and the new pensions orthodoxy 3Paul Bridgen and Traute Meyer
PART II CASE-STUDIES: THE VETERANS
2 The British pension system and social inclusion 47Paul Bridgen and Traute Meyer
3 The Dutch pension system and social inclusion 79Duco Bannink and Bert de Vroom
4 The Swiss pension system and social inclusion 107Fabio Bertozzi and Giuliano Bonoli
PART III CASE-STUDIES: THE NEWCOMERS
5 The German pension system and social inclusion 139Barbara Riedmller and Michaela Willert
6 The Italian pension system and social inclusion 168Michele Raitano
7 The Polish pension system and social inclusion 193Marek Benio and Joanna Ratajczak-Tucholka
PART IV CONCLUSION
8 Private pensions versus social inclusion? Three patterns ofprovision and their impact 223Paul Bridgen and Traute Meyer
Index 253
v
-
Figures
1.1 Lifetime wage as percentage of average wages of risk biographies 14
1.2 Total assets for pension funds and life insurance investments as a percentage of GDP in 2004 in the six countries 29
2.1 Projected real monthly pension levels from rst and second pillar for British men and women in 2050 55
2.2 Third pillar savings from lifetime income required of British men and women to reach social inclusion line of 40% average wage in 2050 61
2.3 Projected real monthly pension levels of British couples in 2050 63
3.1 Projected real monthly pension levels from rst and second pillar for Dutch men and women in 2050 88
3.2 Final salary and average salary scheme outcomes for risk biographies in the Netherlands 92
3.3 Third pillar savings from lifetime income required of Dutch men and women to reach social inclusion line of 40% average wage in 2050 93
3.4 Projected real monthly pension levels of Dutch couples in 2050 95
4.1 Projected real monthly pension levels from rst and second pillar for Swiss men and women in 2050 116
4.2 Projected real monthly pension levels of Swiss couples in 2050 116
4.3 Third pillar savings from lifetime income required of Swiss men and women to reach social inclusion line of 40% average wage in 2050 124
5.1 Projected real pension levels from rst and second pillar for German men and women, 2050 148
5.2 Third pillar savings from lifetime income required of German men and women to reach social inclusion line of 40% average wage in 2050 152
5.3 Projected real monthly pension levels of German couples in 2050 154
vi
-
6.1 Projected real monthly compulsory pension levels from the rst pillar for Italian men and women in 2050 181
6.2 Third pillar savings from lifetime income required of Italian men and women to reach social inclusion line of 40% average wage in 2050 183
6.3 Projected real monthly pension levels of Italian couples in 2050 185
7.1 Projected real monthly pension levels from the notional and funded component of the rst pillar for Polish men and women in 2050 202
7.2 Third pillar savings from lifetime income required of Polish men and women to reach social inclusion line of 40% average wage in 2050 205
7.3 Projected real monthly pension levels of Polish couples in 2050 207
8.1 Projected pension levels from rst and second pillar of all individuals as percentage of social inclusion line, 2050 227
8.2 Projected pension levels from rst and second pillar of all couples as percentage of social inclusion line 230
Figures vii
-
Tables
1.1 A typology of risk biographies 91.2 Employment and inactivity in the six countries 201.3 Employment interruptions in the six countries 221.4 Divorce in the six countries 221.5 Public expenditure on pensions in the six countries 28A1.1 Assumptions for working time and earnings in
relation to average per year, all biographies 36A1.2 Details of assumptions used in the simulations in
all countries 402.1 The three pillars of the British pension system 492.2 Pension income of the unqualied male worker (bio 4b) in
Britain in 2050 by dierent forms of provision 592.3 The eect of divorce on the pensions of female biographies
in Britain, as percentage of social exclusion 642.4 The eect of divorce on the pensions of male biographies
in Britain, as percentage of social exclusion 64A2.1 Periods and type of employer-supported pension coverage
assumed for each biography in the British simulations 763.1 The three pillars of the Dutch pension system 803.2 Percentage of overall pension gained/lost through divorce
in the Netherlands, January 2050 89A3.1 Pension schemes used for simulations 1044.1 The three pillars of the Swiss pension system 1095.1 The three pillars of the German pension system 1405.2 Second pillar schemes used for simulations in Germany 1476.1 The three pillars of the Italian pension system 1706.2 Authorized Italian pension funds in 2005 1766.3 Rates of return of pension funds and TFR in Italy 1787.1 The three pillars of the Polish pension system 1948.1 Comparative country performance in relation to social
inclusion 2298.2 Comparative country performance in relation to social
inclusion by gender 2298.3 Comparative range of pension outcomes for individuals 2308.4 Median required saving rates and current tax rates 237
viii
-
8.5 Comparative country performance of couples 2398.6 Comparative household pension of mother and unqualied
part-time worker as percentage of social inclusion by marital status and partner 241
Tables ix
-
Contributors
Duco Bannink, PhD, works as a lecturer at the University of Twente, Schoolof Management and Governance, Department of Social Risks and SafetyStudies. His teaching and research concerns the implementation andreform of social policies at the EU and national levels. He is currentlyinvolved in an EU Sixth Framework research project on Work Organisationand Restructuring in the Knowledge Society (WORKS). Recent publica-tions are Hidden Change (with Marcel Hoogenboom 2007) and variousarticles on pension reform (with Bert de Vroom).
Marek Benio is a researcher and teacher at the Department of PublicEconomy and Administration, Cracow University of Economics, Poland.His PhD in economics (2002) was devoted to Employees Pension Schemesin Poland. His academic interests encompass social security, labour law andpublic administration. He has participated in a number of internationalresearch projects. Recent publications include: Invisible Privatisation ofSocial Insurance, ZNAE nr 714/2006, Cracow; Pension System Reforms inCentral and Eastern Europe, in: D. Rosati (ed.): New Europe; Report onTransformation, Instytut Wschodni Sorry, Krynica (with Jerzy Hausner).
Fabio Bertozzi is scientic collaborator at the social policy unit of the SwissGraduate School of Public Administration (IDHEAP) in Lausanne,Switzerland. He has taken part in several national and internationalresearch projects in the area of comparative welfare state analysis, with aspecial focus on pensions and labour market policies. His publicationsinclude Swiss worlds of welfare, West European Politics, 1(4) 2004,pp. 2044 (with K. Armingeon and G. Bonoli) and The Swiss welfare state:a changing publicprivate mix? (in D. Bland and B. Gran, eds, SocialPolicy Puzzles, Palgrave Macmillan, forthcoming, with F. Gilardi).
Giuliano Bonoli is Professor of Social Policy at the Swiss Graduate Schoolfor Public Administration (IDHEAP), Lausanne, Switzerland. He receivedhis PhD at the University of Kent at Canterbury for a study on pensionreform in Europe. He has been involved in several national and interna-tional research projects on the process of welfare state transformation. Hisrecent key publications include: The Politics of Pension Reform: Institutions
x
-
and Policy Change in Western Europe, Cambridge University Press, 2000and Ageing and Pension Reform around the World: Evidence from ElevenCountries, Edward Elgar, 2005 (ed. with Toshimitsu Shinkawa).
Paul Bridgen is a senior lecturer in social policy at the University ofSouthampton. He was the joint academic coordinator (with Traute Meyer)of the EU 5th Framework project on which this book is based. He hasseveral publications on developments in British pensions policy from theend of the Second World War to the present day. He has also published onageing and health and social care policy. He is currently working (withTraute Meyer) on a project on occupational pensions in Britain andGermany as part of the Anglo-German Foundations Creating SustainableGrowth in Europe research initiative.
Bert de Vroom is sociologist and head of the Social Risks and SafetyStudies Department, at the University of Twente, the Netherlands. He isand has been involved in various international comparative and nationalresearch projects and has published widely on issues such as changinglabour markets, welfare state policies, new risks, early retirement and pen-sions. He has co-ordinated the European COST A13 research group,Ageing and Work. At the moment he is involved in the OECD project ondisabilities and transitions, and in the Cinefogo Network of Excellence onthe making of European citizenship.
Traute Meyer is a senior lecturer in social policy at the University ofSouthampton. She has participated in international research projects inthe area of pensions, informal and care work. She is the co-editor ofthe European Journal of Social Policy. Publications include: Genderarrangements and pension systems in Britain and Germany, InternationalJournal of Ageing and Later Life (2006, with B. Pfau-Enger); When dobenevolent capitalists change their mind? Explaining the retrenchment ofdened-benet pensions in Britain, Social Policy and Administration(2005); and Class, gender and chance and the social division of welfare,forthcoming 2008, in Ageing and Society (both with P. Bridgen).
Michele Raitano obtained a PhD in economics from University of RomeLa Sapienza in 2004. Currently he is a research fellow at Istituto StudiAnalisi Economica and the University La Sapienza and collaborates withthe Italian government about the pension reform process. His mainresearch topics are welfare state systems, with a special focus on pensions,an area on which he has published many papers. He is a contributing authorto the annual Welfare State Report (Rapporto sullo Stato Sociale) edited by
Contributors xi
-
Professor F.R. Pizzuti and issued by the Department of Public Economyof the University of Rome (La Sapienza) and the Centro Interuniversitariosullo Stato Sociale.
Joanna Ratajczak-Tucholka is a teacher and research assistant at theDepartment of Labour and Social Policy at Poznan University ofEconomics in Poland, where she earned her PhD. Her research focuses onsocial policy and social security, and she is particularly interested in thegender-specic impact of pension systems. Her PhD concentrated on howpension insurance inuenced the situation of German and Polish womenin old age.
Barbara Riedmller is professor at the Political Science Department of theFreie Universitt Berlin. Her main research focus is the German welfarestate, analysed in the context of change in European countries. She carriedout research on social policy and social inclusion, and on pension reformin European countries. She has advised high-level German social policy-makers regarding the consequences of German pension reforms for socialinclusion.
Michaela Willert received her diploma in sociology from the FreieUniversitt Berlin. She works as senior research assistant at the PoliticalScience Department of the FU. Her research interest is in comparativewelfare state research, particularly pension policy. Moreover she examinestrends of Europeanization of social policy.
xii Contributors
-
Abbreviations
AOW Algemene Ouderdomswet (General Old Age Pensions Law,the Netherlands)
AVS/AHV Assurance Vieillesse et Survivants Alters undHinterlassenenversicherung (old age and survivorsinsurance, Switzerland)
BSP Basic State Pension (UK)CDU Christlich Demokratische Union (Christian Democratic
Party)DGB Deutscher GewerkschaftsbundFDC funded dened contributionGDP gross domestic product IKE Indywidualne Konta Emerytalne (Individual Pension
Accounts, Poland)LPP/BVG Prvoyance professionnelle/Beruiche Vorsorge
(occupational pension, Switzerland)NDC notional dened contributionNHS National Health Service (UK)OECD Organisation for Economic Co-operation and DevelopmentPAYG pay-as-you-goPC/EL Prstations complmentaires/Ergnzungsleistungen
(Supplementary benets, Switzerland)PIPs piani pensionistici individuali (personal pensions, Italy)PPE Pracownicze Programy Emerytalne (Employees Pension
Schemes, Poland)TFR trattamento di ne rapporto (deferred wages retained by
rms, Italy)WWB Wet Werk en Bijstand (General Social Assistance)
xiii
-
Acknowledgements
This book is based on the research project Private Pensions and SocialInclusion. This project was funded under the 5th Framework of theEuropean Union, and the authors gratefully acknowledge this support.Research started in January 2003 and the project ocially nished inOctober 2005. During this period, the contributors to this study met regu-larly to ensure a common methodological framework and to compareresults. The country chapters are the outcome of this process. We thankeveryone involved for their unwavering commitment to our common aim,to make comparable the simulation results for six countries, despite com-plexities of the national backgrounds. Moreover, we are very grateful toKaren Anderson and Maria Evandrou for intellectual support during thenal stages of this book.
For their contribution to the preparation of the book for publication wewould also like to thank Graham Baxendale and Jenny Routledge.
Traute Meyer, Paul Bridgen, University of SouthamptonBarbara Riedmller, Free University of Berlin
xiv
-
PART I
Introduction
-
1. Private pensions versus socialinclusion? Citizens at risk and thenew pensions orthodoxyPaul Bridgen and Traute Meyer
Societal ageing is one of the major challenges facing social policymakerstoday. For many European countries it is predicted that in the near futurethere will be more pensioners than people in employment. During the 1990sa consensus amongst Western social policymakers has emerged that welfarestates need more private pension provision in order to master this trend(Haverland 2001, pp. 31011; World Bank 1994). The policy shift that hasfollowed is justied on the basis of concerns about the eect of demographicchanges on the sustainability of existing pensions systems and argumentsabout economic eciency. A multi-pillar system, many believe (EuropeanCommission 2003; World Bank 1994), is better able to cope with the prob-lems created by population ageing, with pre-funded pensions also regardedas superior to pay-as-you-go pensions in terms of rates of return. The basisfor this consensus, which some call the new pensions orthodoxy (Mller1999), has not gone unchallenged (Augusztinovics 2002; Barr 2000) but, nev-ertheless, a shift away from public provision and an increase in private formshas been the predominant development in the pensions systems of most, ifnot all, industrialized countries (Bonoli et al. 2000; OECD 2005a; Pedersen2004; Zaidi et al. 2006).
What do these trends mean for the adequacy of citizens pensions? Arewe about to see pensioner poverty rise after many years in which it hasfallen in most industrialized welfare states (Casey and Yamada 2003)?Most European governments believe we are not; they are condent thatsteps taken towards privatizing national pension regimes have not compro-mised their aim to maintain social inclusion. This view is bolstered bythe European Council and Commission who expressed condence thatpublicprivate regimes are capable of protecting the great majority of citi-zens from social exclusion:
Pension systems, through public earnings-related schemes (rst pillar), privateoccupational schemes (second pillar) and individual retirement provision (third
3
-
pillar), provide good opportunities for most Europeans to maintain their livingstandards after retirement as a result, and in combination with other tax-benet policies for pensioners, older people, in most Member States, generallyachieved a fair, and in some Member States even relatively high, living standard.(European Commission 2003, p. 6; see also European Council 2006)
This position, however entrenched it may be at the present time, is notbased on a systematic theoretical and empirical appraisal of the potential ofprivate provision. Despite the almost universal shift towards non-publicforms of pensions, and even though European societies seem to have reacheda point of no return as far as the likelihood of governmental support forstronger public sectors is concerned, there remain important gaps in ourknowledge about the implications of recent policy developments. The viewamong policymakers and in academia that pension regimes must haveprivate elements to make them nancially sustainable is very strong, but it isnot clear how successful the joint performance of the dierent pillars ofcurrent regimes actually is with regard to social inclusion, particularly for theprotection of citizens without lifelong, full-time employment biographies. Itfollows from this that there is also a lack of knowledge about what type ofpublicprivate mixes best ensure not only nancial sustainability but also aretirement free from social exclusion for citizens at risk. It is the aim of thisbook to address these knowledge gaps.
In the academic realm empirical studies are rare on the distributive impli-cations of the increased use of non-state provision in European pensionsystems and thus the potential of occupational pensions and private savingsto supplement the rst pillar (OECD 2005b, p. 9). This is not to say thatthe eld is untouched. Especially in the three countries analysed here witha long legacy of public and private pension provision the Netherlands,Switzerland and the United Kingdom knowledge is well established aboutthe general level of public and private sector performance. Research on theNetherlands and Switzerland has demonstrated the general inclusiveness ofthe public pillar in conjunction with occupational pensions (Chapters 3 and4), while in Britain academic analyses of pensions have highlighted the highpoverty risks generated by the insucient level of the public pension and theselective nature of occupational schemes (Chapter 2). However given the rolenon-state provision plays in these three veteran countries, there has been lessinterest than one might expect in the heterogeneity of non-state provision.Questions that ask what types of arrangements are more or less successful,or which sectors of the economy do better than others for citizens at risk andwhy, still remain to be answered. Moreover little work has so far been doneon the distributive consequences of recent retrenchments in all threecountries (Bridgen and Meyer 2005; OFS 2004; van Riel et al. 2003). Forobvious reasons, we know even less about private sector performance in the
4 Introduction
-
three other countries of our study that have only recently undergone systemchanges: Germany, Italy and Poland.
Given this decit regarding individual countries it is not surprising thatcomparative assessments are even rarer (OECD 2005b; Social ProtectionCommittee 2006; Zaidi et al. 2006). Few cross-national studies analysepension regimes in terms of their contribution to social inclusion. Insteada substantial proportion of recently published comparative research onpensions or welfare regimes has concentrated on policy transformation andpension politics with much less attention paid to the distributive conse-quences of these changes (Immergut et al. 2007; Bonoli and Shinkawa2005; Clark and Whiteside 2003; Rein and Schmhl 2004; Shalev 1996).Where analyses include the impact of such policies on citizens they havetended not to dierentiate between private and public pension elements(Johnson and Rake 1998), and are generally based on the use of macro-leveldata, such as labour market or income gures (Disney and Johnson 2001).The problem with the latter is that such data give no clear sense of the con-nection between policies and outcomes because they are a product not justof the policy regime in a country but also the economic and social context.Micro-level data are thus essential in the analysis of present policy becausethey allow a rm connection to be made between the regime and thepension outcomes of individual biographies, and on this basis make it pos-sible to assess the comparative performance of dierent pension regimesfor identical individuals (Johnson and Rake 1998). On this basis one candemonstrate which of the national regimes works best for various types ofindividual. So far comparative research that utilizes such data to studyprocesses of social exclusion has not generally considered the role of pen-sions (Aposori and Millar 2005; Barnes et al. 2004; Berthoud and Iacovou2004; Goodin et al. 1999).
However there are a few pioneering comparative evaluations of the out-comes of publicprivate pensions regimes based on this type of approach(OECD 2005b; Social Protection Committee 2006). Yet in terms of evalu-ating the risks of social exclusion they have two important limitations.Firstly they are based on overly simplied biographies of individuals. Inparticular, the life events encountered by many contemporary citizens andthe potential social risks they constitute remain unexplored. Secondly thestudies simulate the income of individuals only. The potentially consider-able impacts of marriage and divorce on social inclusion after retirement,through divorce payments, derived rights through marriage and throughthe pooling of resources in one household, are not considered at all.
Put in broad terms, therefore, this book will explore to what extentthe widespread condence that multi-pillar arrangements pose no majorproblems for levels of social exclusion is justied. We start from the premise
Private pensions versus social inclusion? 5
-
that if this condence were warranted, current pension regimes public, col-lective, employer-related and personal should be able to oer protectionnot just to citizens with incomes on or above the average, but also to thosewith working lives characterized by the complexity of typical contemporarysocial risks. Taking this perspective we examine the performance of sixcurrent national pension regimes in Europe for these risk biographies, indi-viduals and couples. More specically, the six country chapters that followwill concentrate on three main questions:
What is the nature of the publicprivate mix in each pension regime? What pension levels can hypothetical risk biographies on incomes
below the average but above the poverty line expect under currentconditions,1 and what is the contribution of the private and thepublic sector to their pension outcomes?
How do the policymakers responsible for reform aim to address themain shortcomings of their regimes for risk biographies, given thatgenerally a strengthening of the public sector is not considered aviable option?
CONCEPTUAL AND METHODOLOGICAL ISSUES
The research on which this book is based has sought to avoid the main lim-itations in existing work on this area. Firstly public and private types ofinsurance were systematically included in the assessment of each of thecountry regimes in relation to social inclusion. Secondly in establishing theoutcomes of these regimes, the research builds on a more realistic range ofbiographies, whose life courses reect the variety of social risks experi-enced by individuals in post-industrial societies. This approach raised anumber of important conceptual and methodological issues that need to beaddressed at this stage.
Public and Private Pensions
Any assessment of the respective role of dierent types of pension provisionin preventing social exclusion in retirement has rst to address the debateabout the meaning of some basic concepts, such as private pensions andprivatization. Despite their prominence in political and academic debates,these remain shrouded by a good deal of ambiguity with important implica-tions for their operationalization in the analysis of social policy change. Formany economists and the World Bank, private equals pre-funded systems,and they see a superiority of this form over pay-as-you-go programmes in
6 Introduction
-
terms of nancial sustainability. In the broader academic debate, a dichoto-mous publicprivate typology has often been used to categorize pensions,2
with debate hinging on where to draw the line between public and private (forexample Behrendt 2000, p. 26; Esping-Andersen 1990, p. 91; Pedersen 2004).On the basis of this dichotomizing approach privatization involves a simpleshift from the public side to the private side.
While this approach has its merits, there is a growing recognition thatrestrictive sectoral denitions of pension type unnecessarily limit or simplifythe meaning of privatization and hinder our understanding of the distrib-utive consequences of pension regime design (Drakeford 2000; Hyde et al.2003, pp. 18990; Rein and Wadensj 1997; Shalev 1996). In all countriesstudied in this book, pension entitlements are the outcome of complex inter-action regarding regulation, funding and administration, and thereforethe arrangements on which they are based defy a clear-cut publicprivatelabelling (Leisering 2003). It is misleading to speak for example of the successof the private Swiss or Dutch second pillar in terms of social inclusion, ifthe substantial role the state as regulator played for such success is neglected.For this reason, this book heeds the call for a more sophisticated approach(Burchardt 1997, p. 2; Hyde et al. 2003; van Gunsteren and Rein 1985,p. 230) to the categorization of public and private welfare provision; one thatrecognizes the existence of a wide variety of instruments that involvedierent forms of publicprivate interaction. Thus rather than categorizingpension regimes as either public or private, we suggest there is a contin-uum of regimes all of which mix in various ways pensions with public andprivate aspects (see below).
This also means that we understand privatization not merely as a shiftfrom public to private or unfunded to funded schemes; rather privatizationis any process whereby the states role in pension provision and regulation isreduced and the responsibility of a variety of other societal actors employ-ers, insurers, trade unions and individual citizens for retirement provisionis either passively or actively increased (Bonoli et al. 2000). On this basis ouraim is to determine not whether public regimes work better than privateregimes but rather which combinations of public and private engagement inpensions are most consistent with social inclusion.
Constructing Risk Biographies
To help us make this assessment we microsimulated the pension outcomes ofa range of risk biographies. We suggested above that a problem with exist-ing assessments of pension regime performance on this basis was the use ofunrealistically simple, standardized biographies for establishing outcomes.Thus in its study, the OECD only examines the pension entitlements of a
Private pensions versus social inclusion? 7
-
worker who enters the system today and retires after a full career (OECD2005b, p. 40). Similarly the European Commissions Social ProtectionCommittee selects for calculation three full-time, lifelong workers and onevariant of a broken career, dened as a person with an employment biogra-phy of 30 full-time years, split in two phases by ten years outside the labourmarket (Social Protection Committee 2006, pp. 710).3 While the latterapproach incorporates to some extent care responsibility as a risk, the full-time, lifelong worker remains the centre of attention for both studies. Thusthey do not, or to a limited extent only, consider the complex rules whichdetermine pension rights across all pillars in case of life events such as careresponsibilities, divorce, disability, unemployment, training, early retire-ment, change of employer or migration. In contrast the hypothetical indi-viduals presented in this book are modelled according to the complexity ofreal peoples lives. This implies that their wages oscillate throughout theirlives, that employment interruptions may happen more than once, and forreasons other than childcare alone, for example because of unemployment,further training, an industrial accident, or because a person migrates fromanother country or changes employers. Above all, lifes complexity stemsfrom the fact that people have relationships with each other. All of our con-structed biographies, bar one, are married to each other and either live in onehousehold with their spouse when they start drawing their pension, or theyretire as divorcees, or both. The following chapters show that dierentmarital matches, and rematches, have a signicant impact on an individualschances of escaping social exclusion.
More specically all country studies simulate pension entitlements for nineidentical, hypothetical men and women and for variations of these. We callour types risk biographies because they were created on the basis thatduring their adult lives each individual would experience a range of the socialrisks just discussed. Because of these risks their wages oscillate (see Appendix1.1 for details), and their average yearly lifetime incomes are generally notmuch below 40 per cent of the average wage, our social inclusion line, andnot much above the average (Table 1.1 ). All of these individuals started theirworking lives in 2003 and will retire in 2049, unless they retire early. Allcountry teams thus worked with an identical set of hypothetical men, womenand couples on the same wages in relation to the national average.
The following overview of our typology illustrates our types in detail (seealso Figure 1.1 and Appendix 1.1 ).
Our typology consists of three women and ve men, who are marriedto each other. Some remain married until retirement, others divorce andremarry and some stay single after separation. This made it possibleto include in the simulations derived rights through marriage and divorcepayments, and to assess the eect of combined household income on social
8 Introduction
-
9Tab
le 1
.1A
typ
olog
y of
risk
bio
grap
hies
Typ
e of
biog
raph
y Q
uali
cati
onY
ears
of
Ful
l-ti
me
Lif
etim
e T
ype
ofM
arri
age
Hus
band
/N
umbe
ran
d ri
sks
expl
ored
leve
lem
ploy
men
tye
ars
wag
e %
em
ploy
erst
atus
wif
eof
ofav
erag
ech
ildre
n
1) T
he m
othe
r an
d un
qual
ied
par
t-ti
me
wor
ker
in th
e re
tail
sect
or1a
) div
orce
s (2
332
),lo
w39
639
%2
larg
e tw
ice
bio
4a
2re
mar
ries
(35
) co
mpa
nies
mar
ried
or 5
ban
d re
tire
s ea
rly
1b) d
ivor
ces
(23
32),
low
4231
47%
2 la
rge
divo
rced
bio
4a
2ch
ange
s to
ful
l-co
mpa
nies
or 5
bti
me
wor
k af
ter
child
-rea
ring
at 4
01c
) sta
ys m
arri
ed
low
4231
47%
2 la
rge
once
bi
o 4/
b 2
(fro
m 2
3) a
nd
com
pani
esm
arri
edor
5a/
cch
ange
s to
ful
l-ti
me
wor
k at
40
2) T
he m
othe
r an
d qu
ali
ed p
art-
tim
e w
orke
r in
the
wel
fare
sec
tor
2a) d
ivor
ces
(25
35),
med
ium
376
42%
1 la
rge
twic
e bi
o 8b
2re
mar
ries
(37
) co
mpa
nym
arri
edan
d re
tire
s ea
rly
2b) d
ivor
ces
(25
35),
med
ium
4129
54%
1 la
rge
once
bi
o 8b
2ch
ange
s to
ful
l-co
mpa
nym
arri
edti
me
wor
k af
ter
child
-rea
ring
at 4
2
-
10
Tab
le 1
.1(c
onti
nued
)
Typ
e of
biog
raph
y Q
uali
cati
onY
ears
of
Ful
l-ti
me
Lif
etim
e T
ype
ofM
arri
age
Hus
band
/N
umbe
ran
d ri
sks
expl
ored
leve
lem
ploy
men
tye
ars
wag
e %
em
ploy
erst
atus
wif
eof
ofav
erag
ech
ildre
n
2c) s
tays
mar
ried
m
ediu
m37
642
%1
larg
e,on
ce
bio
8a2
(fro
m 2
5),c
hang
es1
smal
l m
arri
ed3a
) typ
e of
empl
oyer
co
mpa
nyan
d re
tire
s ea
rly
3) T
he m
arri
ed c
arer
and
info
rmal
wor
ker
3a) s
tays
mar
ried
lo
w40
4022
%1
smal
l,on
ce
bio
4b,
3(f
rom
23)
and
1
fam
ily
mar
ried
5a/c
or
6ais
dep
ende
nt
busi
ness
(
5)on
par
tner
be
caus
e of
care
ob
ligat
ions
incl
.el
derl
y ca
re3b
) div
orce
s (2
345
),lo
w38
2737
%1
smal
l,1
once
bi
o 5b
2
is d
epen
dent
on
fam
ily b
usin
ess
mar
ried
,or
6b
part
ner
beca
use
of(
5),1
larg
e di
vorc
edca
re o
blig
atio
ns
food
in
cl.e
lder
ly c
are
man
ufac
ture
r
4) T
he u
nqua
lied
wor
ker
in th
e ca
r in
dust
ry4a
) div
orce
s (2
332
),lo
w46
4679
%2
larg
e tw
ice
bio
1a2
rem
arri
es (
35),
com
pani
esm
arri
edsh
ort
spel
l of
unem
ploy
men
t
-
11
4b) s
tays
mar
ried
lo
w37
3765
%2
larg
e on
ce
bio
1c
2(f
rom
23)
,co
mpa
nies
,1
mar
ried
or 3
ach
ange
s em
ploy
ersm
all b
usin
ess
and
reti
res
(5)
earl
y,af
ter
long
er
spel
l of
unem
ploy
men
t
5) T
he in
term
itte
nt w
orke
r in
the
cons
truc
tion
indu
stry
5a) s
tays
mar
ried
m
ediu
m41
4189
%1
med
ium
,1
once
bi
o 1c
2/
3(f
rom
23)
,sm
all (
5)
mar
ried
or 3
a5b
) em
ploy
men
t ga
ps,
com
pany
,sel
f-ch
ange
of
empl
oyer
empl
oyed
type
,fur
ther
tr
aini
ng5b
) div
orce
s (2
332
),m
ediu
m41
4189
%1
med
ium
,1
divo
rced
bio
1a
2em
ploy
men
t sm
all (
5)
or 3
bga
ps,c
hang
e of
com
pany
,sel
f-em
ploy
er t
ype,
empl
oyed
furt
her
trai
ning
5c) s
tays
mar
ried
m
ediu
m31
3162
%1
med
ium
,1
once
bi
o 1c
2/
3(f
rom
23)
,sm
all (
5)
mar
ried
or 3
aem
ploy
men
t ga
ps,
com
pany
,sel
f-ch
ange
of
empl
oyed
empl
oyer
typ
e,se
lf-e
mpl
oym
ent,
disa
bled
at
the
age
of55
-
12
Tab
le 1
.1(c
onti
nued
)
Typ
e of
biog
raph
y Q
uali
cati
onY
ears
of
Ful
l-ti
me
Lif
etim
e T
ype
ofM
arri
age
Hus
band
/N
umbe
ran
d ri
sks
expl
ored
leve
lem
ploy
men
tye
ars
wag
e %
em
ploy
erst
atus
wif
eof
ofav
erag
ech
ildre
n
6) T
he s
mal
l bus
ines
s en
trep
rene
ur6a
) sta
ys m
arri
ed
med
ium
4646
84%
1 sm
all (
5)
on
ce
bio
3a3
(fro
m 2
3)bu
sine
ss o
fm
arri
edw
hich
he
beco
mes
ow
ner
6b) d
ivor
ces
(23
45)
med
ium
4646
84%
1 sm
all (
5)
di
vorc
edbi
o 3b
3bu
sine
ss o
fw
hich
he
beco
mes
ow
ner
7) T
he d
ivor
ced
prov
ider
in th
e ch
emic
al in
dust
ry6b
) div
orce
d (2
332
;m
ediu
m45
4511
3%2
larg
e di
vorc
ed
bio
1a
235
52)
,nee
ded
com
pani
estw
ice
from
or 3
bto
exp
lore
1a
,onc
e im
pact
of
divo
rce
from
3b
on m
en
8) T
he m
iddl
e m
anag
er in
na
ncia
l ser
vice
s8a
) sta
ys m
arri
ed
med
ium
4141
131%
2 la
rge
mar
ried
bio
2c2
(fro
m 2
5),r
etir
es
com
pani
esea
rly
8b) d
ivor
ces
(25
35),
med
ium
4141
131%
2 la
rge
divo
rced
bio
2a/2
b2
reti
res
earl
yco
mpa
nies
-
13
9) T
he in
com
plet
e lo
w29
2949
%2
larg
e Si
ngle
n/a
0re
side
nt in
the
com
pani
esel
ectr
ical
indu
stry
Not
e:T
he in
divi
dual
pen
sion
for
7 is
cal
cula
ted
usin
g ty
pes
1a a
nd 3
b as
the
bio
grap
hica
l tem
plat
e fo
r hi
s tw
o m
arri
ages
and
div
orce
s.H
owev
er,
as in
divi
dual
bio
grap
hies
1a
and
3b o
nly
divo
rce
once
.The
indi
vidu
al p
ensi
on fo
r 8b
is c
alcu
late
d on
the
bas
is t
hat
he is
sin
gle.
How
ever
bio
8b
isal
so u
sed
as t
he b
iogr
aphi
cal t
empl
ate
for
the
seco
nd m
arri
age
ofbi
o2a.
Hen
ce t
he c
oupl
e 2a
and
8b.
-
inclusion. In addition we included one individual who never married. Ourfemale types include one unqualied and one medium-qualied worker andmother, and one traditional housewife and carer.
The mother and unqualied part-time worker in the retail sector (bio 1)allows us to explore the impact of a combination of risks: the instability ofmarriage, a fairly low lifetime income because of low qualications andpart-time work, as well as employment gaps related to care responsibilitiesfor children. The variations in this category have dierent periods of part-time and full-time work. They are married to an unqualied worker (bio 4)or a worker with an intermittent career (bio 5). Two of these marriages endin divorce.
The risk prole of our second type, the mother and qualied part-timeworker in the welfare sector (bio 2), also enables us to assess the role of anemployment career that changes to meet the needs of children; two ofour variants work full-time only for a few years, one much longer. All aremarried, but one divorces and remarries. The main dierence from the rsttype is that the level of qualication of the women in this category is higher;they therefore have higher wages. The same applies to their husbands,middle managers in the nancial services (bio 8).
The third type, the married carer and informal worker (bio 3), has anindependent income below 40 per cent of the average; for protection sherelies mostly on her husband, who for part of his working life owns a smallbusiness (bio 5 or 6). Her level of formal qualication is low, and eventhough she works no less than the rst two types, she does so mainly as aninformal worker in a small family business. One variant has three children,the other has only two. The latter divorces and for the last spell of herworking life she is single and employed by a large food manufacturer. Bothleave the labour force at the age of 57 to care for their elderly mother.
14 Introduction
Figure 1.1 Lifetime wage as percentage of average wages of riskbiographies
22%
37% 39% 42% 42%47% 49%
54%62% 65%
79% 84%89%
113%
131%
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
3a 3b 1a 2a 2c 1b/c 9 2b 5c 4b 4a 6 5a/b 7 8a/b
-
Our types 4 to 8 reect male biographies. They too were chosen torepresent low and medium levels of qualication, and where the womenexperience care-related employment gaps, some of the men are tempo-rarily out of paid work because of unemployment, illness or training.Notwithstanding this gendered ascription, the results for our men andwomen are applicable to all individuals whose biographies have similarfeatures, irrespective of their gender.
With our fourth type, the unqualied worker in the car industry (bio 4),we aimed to capture the lives of Fordist workers and breadwinners. His levelof qualication is low, yet the rst variant has a lifelong career as a manualworker in large companies, interrupted only by one year of unemploymentbetween two employers. The second variant is more at risk. He was maderedundant in his late thirties, and moved to a small company after that. Inhis late fties he was redundant again and he retired early, after ve years ofunemployment. Both variants are married to retail workers (bio 1), and onehas to share pension rights with his divorced spouse.
Our fth type, the intermittent worker in the construction industry(bio 5), has a medium-level qualication and therefore his wage is higherthan the previous types. His main social risk stems from the fact that heworks for three types of smaller company, including his own as a self-employed worker. Moreover, the third variant becomes disabled after anaccident in his mid-fties, which ends his employment career. All aremarried to either the unqualied worker (bio 1) or the carer (bio 3).
The small business entrepreneur is our sixth type (bio 6). This person likethe previous one has a medium level of qualication, yet his lifetime incomeis lower because for most of his life he runs his own small business. He ismarried to the carer (bio 3).
Type 7, the divorced provider in the chemical industry (bio 7), was con-structed with the main aim of exploring the extent to which pension-sharingobligations through divorce would increase his social exclusion risk. He ismarried to the unqualied worker (bio 1) or the carer (bio 3). His employ-ment career is long and uninterrupted, and his income is above the average.
The middle manager in nancial services has an even higher salary thanthe divorced provider and an uninterrupted employment career; we there-fore did not expect him to be at risk. We included him as an eighth type fortwo reasons.
Firstly, this individual illustrates how the better-o do in the nationalregimes. Secondly, based on theories of social distance (Bottero 2004), weassumed that our qualied part-time worker and mother (bio 2) wouldmarry a partner with a matching level of qualication. In order to calcu-late her household income, as a married and as a divorced pensioner, wehad to include him.
Private pensions versus social inclusion? 15
-
Our nal type is the incomplete resident. This individual has a full-timeemployment career on half the average lifetime wages, but this is shortenedbecause she or he came to the respective country in their mid-thirties.
SIMULATING PENSION ENTITLEMENTS
This book assesses pension income from the public pillar, from occupa-tional and from personal pension schemes (detailed assumptions inAppendix 1.2). To this end, the authors of each country study determinedwhat type of public and collective private pension each individual wouldtypically have access to under their national regime.
This was easiest for the public pillar because the conditions for rightsaccrual are transparent and systematic for all citizens, although it is impor-tant to note that we excluded means-tested benets from the simulations,even those designed specically for retired citizens. This was for tworeasons. On philosophical grounds the group accepted Townsends argu-ment that means testing by itself can operate as an exclusionary processbecause it risks stigmatizing those it seeks to assist (Townsend 1979,pp. 87982). Thus we wanted to focus on non-stigmatizing benets becausethey are more consistent with social inclusion. The second reason is thatwhere means testing has been applied it has in most cases encounteredproblems of take-up, such that signicant proportions of those entitled toa particular income level do not actually receive it (see for exampleEvandrou and Falkingham 2005).
Decisions on non-state provision were more dicult, given that the degreeof compulsion varies signicantly. In Switzerland coverage by occupationalschemes is compulsory and it is quasi-compulsory in the Netherlands.However in Britain, Germany, Italy and Poland occupational schemes arevoluntary and coverage is patchier as a result. Given this diculty we couldhave included the compulsory part of the pension regime exclusively, whichwas largely the approach taken by the authors of the study by the OECD(2005b) discussed above. The advantage of this perspective is that only pen-sions guaranteed to citizens are assessed. However this inevitably means theexisting potential of private pensions in countries with voluntarist regimesto contribute to social inclusion would remain unexplored, despite theirsignicance in many cases. Thus we would have had to exclude the veryestablished British and the growing German occupational sector. Given thatour study is focused precisely on the potential of the private sector to oerprotection, we instead chose to include second pillar schemes in those caseswhere our individuals would have a prospect to have access to them, whilemaking sure that we also explained what would happen if they did not. As
16 Introduction
-
a consequence of this methodological choice the reader will nd simulationsfor second pillar coverage for Germany and Britain, but not for Italy, whereoccupational pensions are still very underdeveloped.
On this basis, country teams calculated for all risk biographies their indi-vidual projected gross pension level for 2050, at rst only dierentiated bypublic pensions and by collective plans oered through employers. Whereindividuals divorce, we calculated entitlements according to pension-sharing legislation in each country; in practice this often means that one ofthe ex-spouses receives entitlements from the other while the other has toshare them.
To assess the social inclusiveness of these outcomes, we compare thehypothetical pension of every individual, irrespective of marriage status,with three gradually increasing social inclusion thresholds. The rst onemeasures poverty only. It captures the amount of means-tested social assis-tance retired citizens with inadequate pension rights would be entitled toclaim in each country. This amount also shows what each society has deter-mined as necessary minimum threshold below which no person should fall.National generosity in this regard diers quite strongly between our coun-tries, as the relationship of social assistance and national wage levels show.In four of our cases the poverty line is much lower than 40 per cent of theaverage wage, a measure that we use as the social inclusion line. Social assis-tance for the elderly is only near the social inclusion line in Switzerland and,to a slightly lesser extent, in the Netherlands. Finally we use 50 per cent ofaverage wages as a comfortable social inclusion line (see Appendix 1.2 forfurther explanations). The results give an overview of the combined per-formance of public pensions and collective plans oered through employ-ers in each country for individual citizens at risk.
Personal savings calculations are added for those individuals not pro-tected suciently by public and employer-related schemes. This may bebecause entitlements are too low, or because citizens do not have access topublic or occupational schemes in the rst place, such as the self-employedand some carers.
The calculations just described evaluate the inclusiveness of each nationalregime for individuals at risk. To assess the position of the 11 retired coupleswe compare their joint pension income with the same gradually increasinginclusion thresholds, adjusted for couples (see Appendix 1.2).
A JUSTIFICATION OF OUR APPROACH
So far we have made an argument for simulating entitlements of complexrather than simplied biographies, because they are closer to the lives
Private pensions versus social inclusion? 17
-
people lead, and they are more likely to show where the weaknesses ofpublic and private pension regimes are with regard to social inclusion. Thisgeneral justication does not however explain how exactly we arrived at thetypes summarized above. More specically it needs to be explained why wedid not develop types that closely represent real lives, and opted insteadfor a looser construction of hypothetical people.
The rst answer to this question is that using risk proles representativeof real lives would have been impossible. To evaluate the inclusiveness ofpresent pension regimes, and thus of recent reforms this studys mainaim can only be done by simulating future entitlements, and this involvesmaking assumptions about how peoples lives will develop. If we hadwanted to measure the pension entitlements of real people, we would havehad to conduct an analysis of the employment histories of individualsnearing retirement. Yet because pension systems undergo constant change,such an analysis would only partially show the impact of the presentregime, and in addition echoes of decisions taken at the age of 18 as wellas the inuence of policies long since reformed (Rake 1999, p. 223). Giventhis inevitable eect, the only way to avoid evaluating the impact of cumu-lative reforms is to freeze programmes and regulations at a set point in time,and to model the consequences (Johnson and Rake 1998, pp. 2656; Mot1999; OECD 2005a, pp. 406).
However why did we not attempt to ground our types more rmly indetailed data about the life courses of younger citizens, to make the typol-ogy more representative, rather than illustrative? The development ofbiographies representative of a range of individuals is very dicult, evenfor one country, not least because of the data hunger of sophisticatedmicrosimulation (Johnson and Rake 1998, pp. 2656). This appetite wouldhave been voracious in a large comparative project. However even if thisproblem could have been overcome it would have left us with six nationaltypologies. These would have been useful for national evaluations of theinclusiveness of each system from the perspective of typical British, Dutch,German, Italian, Polish and Swiss citizens, but the comparative aims of theproject would have been undermined (Johnson and Rake 1998, pp. 2656).To gauge the comparative performance of the six regimes it was vital to usea single typology of individuals for all countries.
Based on these considerations our types could not be too specicallygrounded in national data. Instead the risks we selected for our individualshad to be typical for the broad social and economic trends in post-industrialsocieties: the transformation of the male breadwinner model, expressed byan increase in mothers employment, mostly on a part-time basis, and par-ticularly in the expanding service sector, which is apparent also througha destabilization of marriage as expressed by rising divorce rates and an
18 Introduction
-
increasing rate of births outside marriage and single parents. Moreover weobserve a demise of the standard worker with citizens experiencing greaterexibility in the workplace. This increases the need for further training,makes interrupted employment careers more likely and pushes increasingnumbers of people towards self-employment. Last but not least, with moreexible borders citizens who move into countries from outside Europe maynd themselves with incomplete social rights. These trends are taken fromacademic analyses (see Armingeon and Bonoli 2006; Castles 2003; Hantrais1999; Pierson 2001). In addition they reect many policymakers expecta-tions of the future, as the following quote from the European Commissionsjoint report on social protection and social inclusion illustrates:
Demographic changes are compounded by wider changes in cultural values,social relationships, the organisation of families and the nature of work . . . Theincreased feminisation of the labour force is reected in new social demands (forexample, for care facilities for children, elderly and dependent people) and in anew basis for the division of responsibilities within the household. Notablechanges in family structures have taken place, in terms both of a reduction ofhousehold sizes and increasing diversity driven by lower rates of partnershipformation and higher rates of dissolution. As a consequence family links areweakening and, in cases of need, more people have to rely on alternative formsof support. Finally the working environment has become far more volatile andheterogeneous. (European Commission 2005a, p. 4)
Some manifestation of these trends is evident in all six countries, yetthey also dier. Thus each country has seen an increased feminisation ofthe labour force, although in 2005 employment rates varied substantially.Table 1.2 shows that at least 60 per cent of adult women are in paid work,mostly in the service sector, in Switzerland, the United Kingdom, theNetherlands and Germany. For Dutch and Swiss women this is normallypart-time work, while the majority of British and German female employ-ees work full-time. In contrast non-employment is the norm in Italy andPoland, where less than half of adult women have a job, and where self-employment is more important than in the other countries.
Full-time adult carers continue to play a role in all countries too, andhigh labour force participation of women in a country does not rule outtheir existence, just as low female economic activity does not automaticallyincrease their number. It is true that the share of women between 25 and 54who stay outside the labour force in order to look after their families ishighest in Italy, where labour force participation is lowest, but it is not muchlower in Switzerland where female employment is highest and rather low inPoland, despite high inactivity.
These gures thus show a diverse country spectrum, but at the same timethey demonstrate that each of our female biographies has current relevance
Private pensions versus social inclusion? 19
-
20 Introduction
Table 1.2 Employment and inactivity in the six countries
UK NL CH GER IT PL
WomenEmployment rates, 2005 65.9 66.4 70.4 59.6 45.3 46.8Part-time (%) of female 43.1 75.3 58.8 44.3 25.7 14.2
employment, 2005Employment in services 91.6 90.1 85.1 84.3 79.6 65.2*
(%) of female employment,2004
Employment in industry 7.9 7.8* 12.1 14.1 17 16.5*(%) of female employment,2004
Self-employment (%) of female 7.6 11.2 8 7.9 20 26.1employment, 2004
Average exit age from the labour 61 61 62 61 61 56force, 2005**
Inactivity rates (2554 yrs) 13 9 15 12 17 10because of personal/familyresponsibilities, 2004
MenEmployment rates, 2005 77.6 79.9 83.9 71.2 69.9 58.9Part-time (%) of male 10.6 22.6 11.8 7.7 4.5 7.7
employment, 2005Employment in services 72 68 62.7 60.2 58 44.2*
(%) of male employment,2004
Employment in industry 26.7 27.7* 32.7 37 36.9 35.7*(%) of male employment,2004
Self-employment (%) of male 17.2 16.4 10.6 13.3 28.6 31.4employment, 2004
Average exit age of men 63 61 64 61 61 60from the labour force,2005**
Inactivity rates (2554 yrs) 0 0 0 0 0 0because of personal/familyresponsibilities, 2004
Notes:* Figures for 2003** Exit age for Italy, Switzerland: 2003
Sources: http://epp.eurostat.cec.eu.int; European Commission (2005b); CH: Swiss FederalStatistical Oce (2005a)
-
for each of our countries. By the same token with regard to men despitesome variation a high rate of part-time work in the Netherlands and avery low rate in Italy, very low participation rates in Poland and very highin Switzerland, as well as very high levels of self-employment in Poland andItaly the male norm, reected by our typology, is still to be a full-timeworker in either the service or the manufacturing sector who does not taketime o to take on family responsibilities.
The majority of our hypothetical individuals retire at the age of 65; earlyretirement, at 60, is an exception. Those with lower qualication levels enterthe labour market at 18, those with medium-level qualications at 20. Whilethese assumptions do not reect the decisions of some governments to raisethe pension age above 65, they are higher than the current real average exitage from the labour force in all our countries. The protracted nature of theprocess of raising the real retirement age is illustrated by the fact that the EU-wide target set by the Stockholm European Council in 2001 to raise employ-ment rates of older citizens between 55 and 64 by at least 50 per cent by 2010so far has only been met by Britain (56 per cent) and Switzerland (65 percent) and that the Commission is concerned the target will not be reached bymany unless countries increase their eorts (European Commission 2005a,pp. 10, 34; European Commission 2006, pp. 289; OECD 2005a).
The growing importance of more volatile working environments iscaptured in our typology by periods of unemployment and further train-ing, which several of our male types encounter. This reects currenttrends; Table 1.3 shows that there is a chance either event will be experi-enced in any of our countries, yet the likelihood varies quite a lot. Lifelongeducation is signicant in Switzerland, Britain and the Netherlands, whileunemployment is low. Conversely unemployment is very high in Polandand substantial in Germany and Italy, while adults in these three coun-tries have fewer opportunities to be in training programmes. It is notice-able that female unemployment rates are higher than those of men inall countries but Britain. This increased social risk that women bear isreected in our typology by the greater employment gaps our hypotheti-cal women are subjected to.
An indicator for the occurrence of our ninth type, the incomplete resi-dent, is also included in Table 1.3. It shows that the highest share of non-nationals of the population by far live in Switzerland and the lowest inPoland. These gures include non-nationals from EU countries whosepension rights are protected by EU legislation, but more dierentiated datais not available.
Finally divorce plays an important role for our hypothetical biographies.This is reected by real current divorce gures shown in Table 1.4. Againthey vary a lot; with marriage more stable in Italy and Poland and most
Private pensions versus social inclusion? 21
-
fragile in Germany and the UK, but the gures show that in all countries,except for the Netherlands, divorce has been on the rise over the last 20years.
The discussion above shows that all types resonate with the situation inthe six countries. Complete mismatches do not exist, but it is important tokeep in mind that certain life course events are currently less likely in somecountries than in others. This issue will be further discussed as it arisesin the individual country case-studies. With regard to the books mainquestions the country characterizations indicate that, were the conditionsof 2004/05 to continue into the future, Poland, Italy and Germany wouldprobably face greater problems insuring their populations against social
22 Introduction
Table 1.4 Divorce in the six countries
Divorces per 100 marriages 1985 1995 2002
United Kingdom 44.6 52.8 50.5Netherlands 41.1 41.9 39.5Switzerland 29.4 38.5 40.8Germany 36.1 39.4 52.1Italy 5.2 9.3 15.4Poland 18.4 18.4 23.7
Source: Eurostat_ http://epp.eurostat.cec.eu, UK gures for 2002 are from 2000
Table 1.3 Employment interruptions in the six countries
UK NL CH GER IT PL
Men% of 2564-year-olds in 24 17 30 8 6 5
education and training, 2004% of 25-year-olds and over 3.5 3.8 3.2 7.8 4.8 13.8
unemployed, 2005
Women% of 2564-year-olds in 34 18 27 7 7 6
education and training, 2004% of 25-year-olds and over 3 4.4 4.5 9.6 8.4 16.6
unemployed, 2005Non-nationals of national 4.7 4.3 20 8.9 3.4 1.8
population around 2004
Sources: http://epp.eurostat.cec.eu.int; CH: Swiss Federal Statistical Oce (2005b)
-
risks in retirement than Switzerland, the Netherlands and the UnitedKingdom, simply because a larger share of their populations women andmen are economically inactive or unemployed. The country studies willshow whether and to what extent this is the case and what type of systemsare in place to deal with these trends.
WHY THESE SIX COUNTRIES?
Private Pension Veterans
The countries included in this book allow us to study the relationshipbetween private pensions and social inclusion against diverse institutionalbackgrounds. They can be grouped according to either their Beveridgeanor Bismarckian legacies, which determined whether they developed multi-pillar systems fairly early during the second half of the twentieth centuryor whether they were dominated by one public programme (Bonoli 2003,p. 400).
The Swiss, British and Dutch pension systems have historically beenbased on a rst, revenue-funded part, paid as at rate benet, which coversat most basic needs (Bonoli 2003, pp. 400402). However despite this struc-tural similarity there are important dierences between the rst pillararrangements in each country. In Switzerland a public pay-as-you-go(PAYG) pension to meet basic needs was introduced in 1946 and was mod-erately earnings-related, and universal. In contrast the Dutch public PAYGpension, introduced in 1959, is a basic at rate benet for all seniors basedon residence. The British Basic State Pension, introduced in 1946, isdierent again: it is at rate like the Dutch pension but like Swiss state pro-vision is social insurance based. It has never risen above subsistence level(Hannah 1986, pp. 54, 59), although in 1975 a serious attempt was made tocreate a more inclusive state pension through the introduction of the StateEarnings Related Pension Scheme.4 In the changed political climate of the1980s this system was cut down again before it could mature, but a newState Second Pension is still likely to provide an important supplementaryincome to future retirees.
Partly because of the relatively low level of state provision in eachcountry, substantial occupational pension systems have developed, initiallyon a voluntary basis. Because the level of the rst pillar was regarded asinsucient by higher earners, interest grew in pensions above this level. Thisvoluntary development of occupational pension is also likely to have beenthe result of employers, particularly the larger and thus richer ones, usingoccupational provision as a tool to ensure the recruitment and retention of
Private pensions versus social inclusion? 23
-
workers, particularly skilled ones, the early retirement of older surpluslabour, and to manage the relationship with trade unions (Clark et al. 2007,p. 19; Green et al. 1984; Hannah 1986, pp. 1830; Hart 1984, pp. 45, 5051;Hawkesworth 1977; Mares 2001, pp. 1956; Myles and Pierson 2001; Rein1996, p. 40; Sass 1997, pp. 1837; Whiteside 2003, p. 33, 2006). In eachcountry government intervention has further encouraged occupational pro-vision through scal incentives (Hannah 1986, p. 38; Sineld 2000, p. 141;Titmuss 1958).5
However although the early development of occupational pensionsystems in all of our veteran countries was initially voluntary they nowdier considerably regarding the degree of compulsion imposed onemployers and employees. In Switzerland, in 1985, occupational pensionsbecame compulsory for employees above a xed earnings threshold, with aminimum standard set for benets. This reform followed strong pressurefrom left-wing parties expressed through the Swiss federal referendumsystem. They wanted the income replacement benets of occupational pen-sions to be distributed on a more universal basis (Bonoli 2004). Howeverthe system does not cover the self-employed and is less inclusive with regardto women, because they are more likely to fall below the earnings thresh-old (Bonoli 2007; Chapter 4).
In the Netherlands, moves towards more general coverage of occupa-tional provision occurred much earlier as part of broader corporatistarrangements (Whiteside 2006). Legislation enacted in 1947 prescribedthat, where the Ministry of Social Aairs adjudged it appropriate, collec-tive agreements about pensions in one company had a binding eect for therespective sector as a whole. This became particularly relevant during the1980s and 1990s when the second Dutch pillar expanded greatly, takingpressure o the public system (see also Clark and Whiteside 2003, p. 145;Haverland 2001, pp. 309, 314). Yet while the state sets the regulatoryframework for occupational schemes, the specic conditions of sector-widepension plans are negotiated between employers and trade unions(Anderson 2007; van Riel et al. 2003, p. 67). As in Switzerland the systemgenerally does not cover the self-employed but it is more open for part-timeworkers than the Swiss and therefore more inclusive overall. In 2005 itcovered more than 90 per cent of the workforce (OECD 2005a, p. 152).
The British system is the most voluntarist regime in this group (Pierson1994, pp. 5373). Coverage has always been selective because occupationalpension provision is not compulsory (Clark 2000). It has been providedmainly, but not exclusively, for those with full-time earnings, stable employ-ment careers, likely to work in private companies or the public sector(Baldwin 1990, pp. 23940, 243; GAD 2003, p. 18; Ginn 2003, pp. 1314;Meyer and Bridgen forthcoming 2008) and has traditionally excluded those
24 Introduction
-
in small businesses, or with less stable contracts. Until the 1990s part-timeworkers, too, were much more likely not to be members (Groves 1983, p. 40;Groves 1987, pp. 209, 211), but as in the Netherlands, this has begun tochange through improved equal pay and sex discrimination legislation(Mazey 1998, pp. 13944).
Thus Switzerland, the Netherlands and the United Kingdom all have amature occupational pension system in addition to the rst pillar, yet thedierences are also signicant. Firstly while the Swiss and the Dutch statepensions aimed to prevent poverty, the British has been below the povertyline, creating greater dependency on other sources of income, such as occu-pational or private provision or other means-tested benets. Secondly cov-erage through occupational schemes in the Netherlands and Switzerlandhas been much higher than in Britain, because of a greater degree of com-pulsion in the former two countries which nevertheless allowed for bothregimes to exclude the self-employed. Swiss and Dutch second pillars areenforced through legislation, with the dierence that Swiss businesses haveto meet minimum standards, which they can exceed if they wish, whileDutch employers and trade unions have greater scope to negotiate schemedetails. Britain in contrast allows us to study the impact of an essentiallyvoluntarist regime.
Private Pension Newcomers
Until the late 1990s Germany, Italy and Poland tted into the classicBismarckian mould. After extending their scope and generosity from the late1950s, by the end of the 1960s state-regulated and compulsory pay-as-you-gopension systems existed for all employed citizens in the three countries. Thesewere largely funded through social insurance contributions from employersand employees in Germany and Italy, and through taxes in Poland. Pensionentitlements were dependent on length of employment career and weredesigned to preserve pre-retirement standards of life for full-time workers,although the link between earnings and benets in Poland and Italy was notalways transparent with some workers privileged over others, for examplePolish miners and steelworkers. In the 1970s the Italian public pensionreplaced 80 per cent of nal earnings for a full-time worker after 40 years,while this rate was 70 per cent in Germany and 51 per cent in Poland. In addi-tion means-tested social assistance benets had been introduced in order toprotect those without sucient insurance from poverty (Chapter 5, 6, 7 in thisbook; Ferrera and Jessoula 2007; GUS 2003, p. 447; Mller 1999; Stroinski1998, p. 30). Funded systems as well as voluntary arrangements throughcompanies or individuals only played a small role, if any. They were crowdedout by the strong public sector (Bonoli 2003, p. 400; Haverland 2001, p. 312).
Private pensions versus social inclusion? 25
-
During the 1990s reform pressures mounted because all three countriessuered from chronic high unemployment and low activity rates, driving upthe social insurance contribution for the economically active populationand for employers (Table 1.3). It is fair to say that Polands as well asGermanys economic problems were inuenced too by the collapse of theEastern Bloc, leading to a fundamental change in both societies, but par-ticularly in Poland and the East German New Lnder (Schulze and Jochem2007). To make matters worse, Germany, Italy and Poland are aected par-ticularly strongly by ageing populations; they are amongst the countrieswith the lowest fertility rates in Europe and thus their predicted depend-ency ratios for the elderly population were comparatively high too, increas-ing the need for pension reform. Against this background governments inall three countries felt compelled to implement fundamental changesduring the 1990s. These are detailed in Chapters 5, 6 and 7. However, brieyput, they involved the sweeping away of the principle of the rst pillar asguarantor of the customary standard of living for the long-term employed.They also led to a lowering of public pensions and to the strengthening ofactuarial principles in the rst pillar in Italy and Poland, which now grantrelatively insignicant compensation for citizens with care responsibilitiesor for part-time workers. Thus German, Italian and Polish citizens are nowexpected to depend more strongly on the less redistributive second andthird pillars in order to ll the gap that decreased public pension levels arelikely to leave. So far in all countries reliance has been placed on the vol-untarism of non-state actors. Incentives in the form of tax and subsidieshave been created for citizens as well as for corporate actors to invest vol-untarily in occupational or personal schemes in order to make up for thepublic shortfall.
This brief overview demonstrates that our three countries share aBismarckian legacy and that all have recently made a sudden private turn.However it also became apparent that privatization looks rather dierentfor each of them. In moving away from predominantly statist systems theyhave each chosen a dierent mix of public and private provision. Our studywill show how these dierent approaches aect the way these three coun-tries fare in producing pension outcomes consistent with social inclusion.
In conclusion our group comprises three countries that already had sub-stantial funded components, and therefore have been able to adjust tocurrent circumstances by more incrementally reforming their existingprogrammes. The other three countries have recently implemented sub-stantial changes to their pension systems, introducing funded elements intomature pay-as-you-go schemes. As a result of these developments compar-ing the six pension regimes as they exist today, the systematic publicprivate dierences between the two groups disappear on the level of pro-
26 Introduction
-
grammes and legislation, and a continuum of regimes takes their placewhich mix in various ways pensions with public and private aspects. Wheredierent countries might lie on this continuum depends on the principlechosen to organize it. In this regard one important issue is the role of thestate either as provider or regulator. At one end of a continuum organizedon this basis, we would nd Switzerland and the Netherlands where stateprovision is inclusive and redistributive and where the coverage of occupa-tional schemes is broad in scope because of legislation, but where employ-ers and trade unions still have some freedom to use occupational schemesto further their own interests. At the other end of this type of continuumare Germany and the United Kingdom. Here state provision is lower andthe regulation of occupational and personal schemes is based on settingstandards and oering incentives, leaving it to other societal actors todecide whether occupational provision is supplied. In the middle, we ndItaly and Poland, which are similar in that the role of the state in pensionprovision remains dominant, but where dierences exist in the statesengagement with non-state providers, Italy having moved more in thisdirection than Poland. However if we chose instead to organize our contin-uum in relation to the size and generosity of the state pay-as-you-go system,and thus the role left to non-state provision in preventing social exclusion,our order would be Italy, Switzerland, the Netherlands, Germany and theUK, followed by Poland as the country with the smallest pay-as-you-gopension in our group. In terms of assessing the combination(s) of publicand private approaches most consistent with the aim of social inclusion inretirement, such distinctions are, as will be seen, extremely important.
We can therefore say that by selecting these six countries we have chosenexamples from two distinct institutional legacies which today show a broadrange of combinations of private and public elements. Our aim is to seewhich combinations are most consistent with social inclusion.
SPENDING PATTERNS AND AGE PROFILE
One would expect that Beveridgean and Bismarckian trajectories wouldalso generate two distinct spending patterns, with higher public expendituredisplayed by the newcomers and greater engagement in privately ownedpension funds in the veteran group.
However as Table 1.5 indicates, countries with mature three pillarsystems do not necessarily spend less money on public pensions than thosewith predominantly public pillars.6 The dierences between some of theveteran and newcomer group were small in 2003 but they becomeinsignicant when we take dependency rates into account. Italy and Poland
Private pensions versus social inclusion? 27
-
are certainly top spenders, but the former also has the highest dependencyrates in the group. German expenditure is almost the same as two veterancountries, the Netherlands and Switzerland, but on the basis of a higherdependency ratio than either of them. This suggests that the Dutch andSwiss public systems are more generous than the German system andalmost equal in generosity to the Italian system, despite the fact that theyalso have large private pillars. Thus the only veteran country that doesindeed display low expenditure on public pensions is Britain, but even herethe dierence is not great. We therefore conclude that while a three pillarregime may be accompanied by low public spending on pensions undercertain circumstances, there are no reliable systematic dierences in publicexpenditure between the two groups.
When we consider the role of funded pensions in each country throughthe amount invested by pension funds and life insurance companies theanticipated dierence between veterans and newcomers is conrmed moreclearly (Figure 1.2).7 Due to the mature funded systems of Switzerland,the Netherlands and the United Kingdom investments were worth morethan these countries GDP in 2004, and far above the OECD average(OECD 2005d, p. 3). In fact Switzerland and the Netherlands, togetherwith Iceland, are the countries with the strongest-funded pension marketsglobally (OECD 2005d). Britain is somewhat less developed, probablybecause second pillar coverage is patchier, due to its voluntarist nature, andtherefore less money is invested than in the quasi-compulsory occupationalpensions sectors of Switzerland and the Netherlands (OECD 2005b).However the fact that British life insurance investment is higher than any-where else suggests that individuals have taken the initiative to make up forthe shortcomings of the public and occupational sectors; as we will arguein Chapter 2, these are unlikely to be citizens on lower incomes, though.
In summary it is true that the three countries with a long history offunded pensions invest large amounts of their national wealth in thesesystems; this is also true under the conditions of British voluntarism, albeit
28 Introduction
Table 1.5 Public expenditure on pensions in the six countries
UK NL CH GER IT PL
Public expenditure on pensions, 11.0 12.6 13.2 13.4 15.1 14.3current prices (% of GDP), 2003
Old age dependency ratio, 2004 24.3 20.5 25 26.8 28.9 18.6
Sources: Expenditure: Eurostat_ http://epp.eurostat.cec.eu; Old Age Dependency Ratio(Population 65 and over as percentage of population 1565): EPC (2006) Annex;Switzerland: gures for 2000 from Bundesamt fr Sozialversicherungen (2006): T14.
-
to a lesser extent. Funded schemes are growing fast in the newcomer group,too (OECD 2005d), yet levels are still comparatively low.
Overview
As we have just shown, in spite of very dierent legacies, in 2006, thepension systems of the privatization veterans and the privatization new-comers look less dierent. This is because the latter group has followed thetrend for multi-pillar systems so dominant in highly developed countries.Against this background the main aim of this study is to explore what riskof social exclusion publicprivate pension regimes pose for citizens wholead complex lives, which include risks typical for post-industrial society. Inthis introduction we have sought to explain the rationale behind this ques-tion and to justify to the reader why, to examine it, we use microsimulationof retirement income of illustrative individuals.
In the following, each of the case-studies is structured in the same way.First they characterize all pillars of their current pension regime, consider-ing in particular how occupational schemes and private savings programmesare regulated by the state and/or the trade unions and to what extent suchnon-state actors can engage in this area on a voluntary basis. In a second stepeach country study calculates how socially inclusive their national regimesare on the basis of the microsimulation methodology outlined above. Finallyconsideration is given to the extent to which the weaknesses in each countrysperformance are being addressed in the current policy debate. On the basisof the responses provided to these questions in the country chapters, the nal
Private pensions versus social inclusion? 29
Notes: CH/NL 2003 data used for 2004; CH: 2003 data preliminary estimates,http//www.oecd.org/document/46/0,2340,en_2649_34853_36091822_1_1_1_1,00.html,accessed August 2006
Figure 1.2 Total assets for pension funds and life insurance investments asa percentage of GDP in 2004 in the six countries
153140
116
3222
11
109112 106
65
4 3 7
84
Switzerland Netherlands United Kingdom Germany Italy Poland Total OECD
Inve
stm
ents
com
pare
d wi
th G
DPin
per
cen
t
Pension Funds and Life insurance Pension Funds only
-
chapter assesses whether it is possible to identify publicprivate pensionmixes which successfully provide socially inclusive outcomes.
NOTES
1. This research generally relates to the pension systems in each country as they operated in2003, when this research started. However in some countries (for example Germany andItaly) the impact of ongoing reform processes on system design was anticipated. This isdetailed in the country chapters.
2. This has occurred notwithstanding a recognition of the diculties involved in deningexactly what should be considered public or private. See Esping-Andersen (1990) p. 91;see also Shalev (1996); an exception was van Gunsteren and Rein (1985).
3. The study by the OECD (2005a) simulates pension levels for individuals starting work in2002 in all OECD countries by dierent income levels. Another big dierence to our studyis that only mandatory pension schemes and those with very broad coverage are included,paying less attention to the British second pillar for example. Secondly, the OECD focuseson replacement rates of pensions, and does not compare all pillar entitlements with socialinclusion lines, as we do, this is also true for the Social Protection Committee Report com-missioned by the EU on projected pension replacement rates. Thirdly, the OECD takesinto account gross and net gures, that is it assesses the impact of tax and social insur-ance. We use this study to compare and discuss our results.
4. An earnings-related pension was introduced in 1961 but this was insubstantial anddesigned mainly to provide additional contributions to reduce the Exchequer cost of pen-sions (see Bridgen 2006).
5. However in recent years government regulation has also increased employers costs andsmothered their willingness to engage in occupational pensions (Clark 2003, p. 234;Cutler and Waine 2001, pp. 1089; Quadango and Hardy 1996; Ross and Wills 2002; Sass1997, pp. 2338; Whiteside 2003, pp. 412).
6. These gures include as public expenditure on pensions the contracted-out national insur-ance rebate used in Britain to fund private alternatives to the State Second Pension.
7. To assess the approximate role of funded pensions in each of our countries, we use theOECDs gures on the sum of investments of pension funds and of life insurance com-panies. These gures comprise assets of most occupational and personal pensions savingsin the public and private sectors of the countries we studied (OECD 2005c, 2005d). Thecaveat here is that these gures do not include book reserves, that is, sums entered in thebalance sheet of the plan sponsor as reserves or provisions for occupational pension planbenets (OECD 2005b, p. 41). In our group book reserves are mainly used in Germany,where as a consequence real pension claims of households are higher than suggested bypension fund and life insurance investments (OECD 2005d; Chapter 5).
BIBLIOGRAPHY
Anderson, K.M. (2007), The Netherlands: political competition in propor-tional system, in E.M. Immergut, K.M. Anderson and I. Schulze (eds),Handbook of West European Pension Politics (pp. 71357), Oxford: OxfordUniversity Press.
Aposori, E. and Millar, J. (eds) (2005), The Dynamics of Social Exclusion in Europe:Comparing Austria, Germany, Greece, Portugal and the UK, Cheltenham, UK andNorthampton, MA, USA: Edward Elgar.
30 Introduction
-
Armingeon, K. and Bonoli, G. (eds) (2006), The Politics of Post-Industrial WelfareStates, Oxford: Routledge.
Augusztinovics, M. (2002), Issues in pension system design, International SocialSecurity Review, 55(1), 2135.
Baldwin, P. (1990), The Politics of Social Solidarity: Class Bases of the EuropeanWelfare State 18751975, Cambridge: Cambridge University Press.
Barnes, M., Heady, C., Middleton, S., Millar, J., Room, G., Papadopoulos, F. andTsakloglou, P. (2004), Poverty and Social Exclusion in Europe, Cheltenham, UKand Northampton,