TURUN YLIOPISTON JULKAISUJA – ANNALES UNIVERSITATIS TURKUENSIS
SARJA – SER. B OSA – TOM. 701 | HUMANIORA | TURKU 2024
GENERATIONAL 
ECONOMIC INEQUALITY
Life course approach to income, 
consumption and wealth between cohorts
Esa Karonen
GENERATIONAL ECONOMIC
INEQUALITY
Life course approach to income, consumption and
wealth between cohorts
Esa Karonen
 
 
 
 
TURUN YLIOPISTON JULKAISUJA – ANNALES UNIVERSITATIS TURKUENSIS 
SARJA – SER. B OSA – TOM. 701 | HUMANIORA | TURKU 2024 
University of Turku
Faculty of Social Sciences
Department of Social Research
Sociology
Doctoral programme on Inequalities, Interventions and New Welfare State
Supervised by
Professor, Mikko Niemela¨
University of Turku
Professor, Jani Erola
University of Turku, INVEST Research
Flagship Centre
Reviewed by
Professor, Juho Ha¨rko¨nen
European University Institute, EUI
Professor, Anette Fasang
Humboldt University of Berlin
Opponent
Professor, Juho Ha¨rko¨nen
European University Institute, EUI
The originality of this publication has been checked in accordance with the University
of Turku quality assurance system using the Turnitin OriginalityCheck service.
ISBN 978-951-29-9969-9 (PRINT)
ISBN 978-951-29-9970-5 (PDF)
ISSN 0082-6987 (PRINT)
ISSN 2343-3191 (ONLINE)
Painosalama, Turku, Finland, 2024
To my mother, who always had faith in me even when I didn’t have faith in myself.
UNIVERSITY OF TURKU
Faculty of Social Sciences
Department of Social Research
Sociology
KARONEN, ESA: Generational economic inequality
Doctoral dissertation, 211 pp.
Doctoral programme on Inequalities, Interventions and New Welfare State
November 2024
ABSTRACT
The resources of one generation significantly affect the life chances and eco-
nomic opportunities of the subsequent generation. It isn’t merely about “how the pie
is sliced” but a more complex question concerning how it fluctuates throughout life
at different stages and how these economic life stories compare across generations.
Consequently, one generation’s position in the hierarchy of financial affluence de-
pends on the conditions of a given period. Previous studies have primarily focused
on income distribution, providing a somewhat limited view of inequality. While there
are more multidimensional approaches to economic inequality, they fail to consider
that economic resources do not remain static. The level of income depends on the
life stage or the time the individual was born. The period will dictate the available
opportunities to strive for a better economic outcome.
To build upon and enhance previous analyses, this study offers new information
in two areas. First, the study considers both the life course of each cohort and the
period. Second, a holistic framework is applied to economic measures, accounting
not only for income but also for consumption and wealth. The underlying logic is that
income, consumption, and wealth do not operate in isolation, as one needs income
to acquire goods and services and save the surplus for future use.
This study evaluates how the three types of economic resources are distributed
across cohort differences over the life course, providing a more comprehensive anal-
ysis of economic inequality. The analysis combines high-quality household-level
income, consumption, and wealth data with longitudinal register-based datasets. The
data cover various periods spanning from as early as 1966 to 2021. The data are a
representative sample of the population, and the register datasets are census datasets
which contain the entire Finnish population. The analyses use state-of-the-art age-
period-cohort models and multilevel regression models.
The results demonstrate how income, consumption, and wealth vary between
generations and indicate which background factors contribute to income develop-
ment throughout the life course. This directly supplements previous multidimen-
sional research endeavours on background factors.
Articles I and II focus on income from the age-period-cohort perspective and
especially from the between-cohort perspective. Overall, the results suggest that
from the 1970s cohort onwards, no income development was related to the linear
trend of average income development. Similar trends are found in the second sub-
study on income, as the 1980s cohort did not increase their income in relation to the
iv
1970s cohort. Thus, overall, it can be stated that there is income stagnation among the
younger cohorts. In terms of age, period, and cohorts, the research identified cohort
and period effects as being statistically significant, with no age effects observed after
introducing controls, especially main economic activity as the strongest explanatory
variable. This is also supported by the second sub-study, which was conducted using
registers. Here, education, unemployment, and occupational status explained most
of the income inequality between cohorts, especially at the peak career point during
the middle of the life course. In generational terms, the results support previous
findings where the Baby Boomers are the ‘winners’ and the youngest generation are
the ‘losers.’
Article III analyses necessities and leisure-time expenditures between high- and
low-income earners. Findings suggest that both consumption and income have main-
tained a stagnant gap between both extremes of income groups over cohorts. Overall,
the age-period-cohort analysis reveals that the relative expenditures between the two
income groups have slightly increased in favour of the high-income group. The more
affluent end of the spectrum seems to have been gradually investing in leisure time
over cohorts, while the low-income group remains ‘leisure poor.’ This could imply
that those in more economically adverse situations have more difficulty participating
in certain societal functions and seeking self-improvement in areas such as hobbies
and other downtime activities. This could mean that either the prices of these goods
and services have become more expensive, or there is less opportunity to invest in
these activities when income is low.
Article IV examines wealth inequality between cohorts over the life course. The
results show that overall, inter-cohort wealth inequality in gross and net wealth has
not increased over time. The results show potential periodic differences related to
age, as overall wealth did temporarily decrease during the financial crisis of the
1990s, but they do not have a permanent impact on wealth accumulation. Wealth
accumulation over the life course does not seem to be hindered by economic cir-
cumstances in the long run. Yet, there is an exception. The 1980s cohort displays a
slower rate of wealth accumulation than the preceding cohort. It also bears a higher
debt load coupled with slower loan amortization, which has delayed the purchase of
dwellings and pushed it to a later age. Consequently, the prolonged loan amortiza-
tion could have long-term implications for total wealth accumulation, leaving future
retirees to live on tighter budgets. Financial wealth shows variations among younger
cohorts, where their initial investment ages are lower compared to other cohorts.
These findings support previous results indicating that building personal wealth and
assets is becoming increasingly important.
To summarize, both the period and cohort are pivotal components in all eco-
nomic metrics. While age seems to have a certain degree of influence, the economic
inequality is more significantly associated with the timing at which different cohorts
encounter various periods. The younger generation, in particular, seems to be facing
economic pressure.
KEYWORDS: generational inequality, economic inequality, consumption, in-
come distribution, wealth
v
TURUN YLIOPISTO
Yhteiskuntatieteellinen tiedekunta
Sosiaalitieteiden laitos
Sosiologia
KARONEN, ESA: Generational economic inequality
Va¨ito¨skirja, 211 s.
Doctoral programme on Inequalities, Interventions and New Welfare State
Marraskuu 2024
TIIVISTELMA¨
Yhden sukupolven resurssit vaikuttavat merkitta¨va¨sti seuraavan sukupolven
ela¨ma¨nmuutoksiin ja taloudellisiin mahdollisuuksiin, silla¨ talous vaatii ylla¨pitoa.
Kyse ei ole vain ”kuinka kakku jaetaan”, vaan monimutkaisempi kysymys koskee
sita¨, miten se vaihtelee ela¨ma¨n eri vaiheissa ja miten na¨ma¨ taloudelliset ela¨ma¨ntarinat
vertautuvat sukupolvien va¨lilla¨. Na¨in ollen yhden sukupolven asema taloudellisen
vaurauden hierarkiassa riippuu annetun ajanjakson olosuhteista. Aiemmat tutkimuk-
set ovat pa¨a¨asiassa keskittyneet tulonjakoon, mika¨ antaa melko rajallisen kuvan eri-
arvoisuudesta. Vaikka taloudelliseen eriarvoisuuteen on moniulotteisempia la¨hes-
tymistapoja, ne eiva¨t ota huomioon sita¨, etta¨ taloudelliset resurssit eiva¨t pysy muut-
tumattomina. Tulotason ma¨a¨ra¨a¨ ela¨ma¨nvaihe tai aika, jolloin yksilo¨ on syntynyt.
Ajanjakso ma¨a¨ra¨a¨ saatavilla olevat mahdollisuudet pyrkia¨ parempaan taloudelliseen
lopputulokseen.
Laajentaakseen ja parantaakseen aiempia analyyseja¨, ta¨ma¨ tutkimus tarjoaa uutta
tietoa kahdella alueella. Ensinna¨kin, tutkimus ottaa huomioon seka¨ kunkin kohortin
elinkaaren etta¨ ajanjakson. Toiseksi, taloudellisiin mittareihin sovelletaan holistista
viitekehysta¨, joka huomioi paitsi tulot, myo¨s kulutuksen ja varallisuuden. Taustalla
oleva logiikka on se, etta¨ tulot, kulutus ja varallisuus eiva¨t toimi tyhjio¨ssa¨, silla¨ tuloja
tarvitaan tavaroiden ja palvelujen hankkimiseen ja ylija¨a¨ma¨n sa¨a¨sta¨miseen tulevaa
ka¨ytto¨a¨ varten.
Ta¨ma¨ tutkimus arvioi, miten kolmen tyyppisia¨ taloudellisia resursseja jaetaan
kohorttieroissa elinkaaren aikana, tarjoten kattavampaa taloudellisen eriarvoisuuden
analyysia. Aineisto yhdista¨a¨ korkealaatuista kotitalouksien tulo-, kulutus- ja var-
allisuustietoa pitkitta¨isrekisteripohjaisiin tietoaineistoihin. Aineisto kattaa eri ajan-
jaksot, aina vuodesta 1966 vuoteen 2021 asti. Aineisto on va¨esto¨a¨ edustava otos
ja rekisteriaineistot ovat va¨esto¨nlaskenta-aineistoja, jotka sisa¨lta¨va¨t koko Suomen
va¨esto¨n. Analyysit ka¨ytta¨va¨t huippuluokan ika¨-periodi-kohorttimalleja ja monita-
soregressiomalleja. Tulokset osoittavat, miten tulot, kulutus ja varallisuus vaihtele-
vat sukupolvien va¨lilla¨ ja mitka¨ taustatekija¨t vaikuttavat tulokehitykseen elinkaaren
aikana. Ta¨ma¨ ta¨ydenta¨a¨ aikaisempaa tutkimusta moniulotteisessa viitekehyksessa¨ eri
taustatekijo¨iden osalta.
Artikkelit I ja II keskittyva¨t tulojen tarkasteluun ika¨-periodi-kohorttina¨ko¨kul-
masta ja erityisesti kohorttien va¨lisesta¨ na¨ko¨kulmasta. Yleisesti tulokset osoittivat,
etta¨ 1970-luvun kohortista la¨htien tulojen kehitys ei seurannut keskima¨a¨ra¨isen tu-
lojen trendia¨. Samankaltaisia tulos lo¨ytyy myo¨s toisesta alatutkimuksesta, koska
vi
1980-luvun kohortti ei lisa¨nnyt tulojaan suhteessa 1970-luvun kohorttiin. Na¨in
ollen voidaan todeta, etta¨ nuoremmat kohortit ka¨rsiva¨t tulonkehityksen stagnaa-
tiosta. Ika¨-, periodi- ja kohortti-tekijo¨iden osalta tutkimus tunnisti kohortti- ja
periodiaikutukset tilastollisesti merkitta¨viksi, eika¨ ika¨vaikutuksia havaittu mallien
vakioinnin ja¨lkeen. Erityisesti pa¨a¨asiallisen taloudellisen toiminnan ollessa vahvin
selitta¨va¨ tekija¨ ika¨vaikutusten osalta. Ta¨ta¨ tukee myo¨s toinen alatutkimus, joka
tehtiin rekisterien avulla. Siina¨ koulutus, tyo¨tto¨myys ja ammatillinen asema selit-
tiva¨t suurimman osan tuloeroista kohorttien va¨lilla¨, erityisesti uran huippukohdassa
ela¨ma¨nkulun keskivaiheilla. Sukupolvien osalta tulokset tukevat aiempia lo¨ydo¨ksia¨,
joissa suuret ika¨luokat ovat ”voittajia” ja nuorin sukupolvi on ”ha¨via¨jia¨”.
Artikkeli III analysoi kulutusta niin va¨ltta¨ma¨tto¨myyksien ja vapaa-ajan menojen
na¨ko¨kulmasta korkea- ja matalatuloisten va¨lilla¨. Tulokset osoittavat, etta¨ seka¨ ko-
horttien va¨liset kulutus- etta¨ tuloerot ovat sa¨ilytta¨neet saman va¨limatkaeron tulo-
ryhmien a¨a¨ripa¨iden va¨lilla¨. Ika¨-periodi-kohorttianalyysi paljastaa, etta¨ suhteelliset
menot kahden tuloryhma¨n va¨lilla¨ ovat kasvaneet korkeatuloisen ryhma¨n hyva¨ksi.
Varakkaamman spektrin pa¨a¨ na¨ytta¨a¨ enenevissa¨ ma¨a¨rin kasvattavansa kulutusta
vapaa-aikaan kohorttien yli mitatessa, kun taas matalatuloiset ryhma¨t ovat vapaa-
ajan mahdollisuuksiltaan ”ko¨yhia¨”. Ta¨ma¨ voisi viitata siihen, etta¨ taloudellisesti
hankalammassa tilanteessa olevilla on suurempia vaikeuksia osallistua tiettyihin
yhteiskunnallisiin toimintoihin ja pyrkia¨ itsekehitykseen alueilla kuten harrastuk-
sien ja muut vapaa-ajan toiminnan osalta. Ta¨ma¨ saattaa tarkoittaa, etta¨ joko na¨iden
hyo¨dykkeiden ja palvelujen hinnat ovat nousseet tai etta¨ vapaa-aikaan on va¨hemma¨n
kulutusmahdollisuuksia, koska tulot ovat matalat.
Artikkeli IV tutkii varallisuuseriarvoisuutta kohorttien va¨lilla¨ elinkaaren
na¨ko¨kulmasta. Tulokset osoittavat yleisesti, etta¨ erot kohorttien va¨lisessa¨ brutto-
ja nettovarallisuudessa ei ole kasvanut ajan myo¨ta¨. Ne paljastivat myo¨s mahdol-
lisia periodiin liittyvia¨ eroja, silla¨ kokonaisvarallisuus va¨heni tilapa¨isesti 1990-luvun
talouskriisin aikana, mutta silla¨ ei ole pysyva¨a¨ vaikutusta varallisuuden kertymiseen.
Kohorttien va¨lisen varallisuuden ja ia¨n suhteen varallisuuden kertyminen elinkaaren
aikana ei na¨yta¨ olevan riippuvainen taloudellisista olosuhteista pitka¨lla¨ aikava¨lilla¨,
yhta¨ poikkeusta lukuun ottamatta. 1980-luvun kohortti kerrytta¨a¨ varallisuutta hi-
taammin kuin edelta¨va¨ kohortti. Heilla¨ on myo¨s suurempi velkataakka yhdis-
tettyna¨ hitaampaan lainanlyhennykseen, mika¨ on viiva¨stytta¨nyt asuntojen ostoa ja
siirta¨nyt sen myo¨hempa¨a¨n ika¨a¨n. Na¨in ollen pitka¨aikainen lainanlyhennys saat-
taa vaikuttaa pitka¨lla¨ aikava¨lilla¨ kokonaisvarallisuuden kertymiseen, ja¨tta¨en tule-
vat ela¨kela¨iset ela¨ma¨a¨n tiukemmalla budjetilla. Rahoitusvarallisuus osoittaa vai-
htelua nuorempien kohorttien keskuudessa, joilla alkusijoitusika¨ on alhaisempi ver-
rattuna muihin kohorteihin. Na¨ma¨ tulokset tukevat aiempia tuloksia, joiden mukaan
henkilo¨kohtaisen varallisuuden ja omaisuuden kerrytta¨minen on yha¨ ta¨rkea¨mpa¨a¨.
Lopuksi, ajanjakso etta¨ ika¨luokka ovat keskeisia¨ tekijo¨ita¨ kaikkia taloudellisia
mittareita tarkastellessa. Vaikka ia¨lla¨ na¨ytta¨a¨ olevan tietty yhteys, taloudellinen eri-
arvoisuus liittyy enemma¨n periodiin, jolloin eri sukupolvet kohtaavat erilaisia ajan-
jaksoja. Erityisesti nuorempi sukupolvi na¨ytta¨a¨ kohtaavan taloudellista painetta.
ASIASANAT: sukupolvien va¨linen eriarvoisuus, taloudellinen eriarvoisuus, ku-
lutus, tulonjako, varallisuus
vii
Acknowledgements
I want to express my heartfelt gratitude to the following individuals who have made
this journey possible and have supported me throughout the process of completing
this dissertation.
None of this would have been possible without my supervisors, Mikko
Niemela¨ and Jani Erola. I have the utmost respect for my main supervisor, Mikko,
who initially took a chance on me. Over the years, I have come to admire your tenac-
ity, uncompromising willpower, and your ability to navigate the academic trenches.
Without your guidance, wisdom, and understanding, this dissertation would not ex-
ist. I am equally grateful to my other supervisor, Jani, who has kept this PhD journey
moving forward and provided excellent insights on every aspect of academic work.
You have shown me how far talent, burning passion, and drive can take you, and I
like to think some of that has rubbed off on me. Thank you both for challenging me
to think deeper and always pushing me to strive for excellence.
I had the privilege to work under the wing of a legend, Olli Kangas, who hero-
not clearly left from the corner of eye since the beginning. I also want to thank
ically led our march during the Horizon2020 Beyond4.0 project. The wit and depth
of knowledge left me an everlasting impression, and I appreciate the wink that was
Hannu Ruonavaara for unknowingly guiding me to keep the door open for a PhD
after the master’s thesis.
I thank Juho Ha¨rko¨nen and Anette Fasang for pre-examining this thesis and pro-
viding constructive remarks on the manuscript. In addition, I humbly thank you,
Juho Ha¨rko¨nen, for agreeing to act as the opponent in the public defence of this
dissertation.
During this PhD journey, I had the privilege of working with fantastic people at
the Department of Social Research and the INVEST research flagship centre. First
and foremost, I would like to thank “Vesa” of the Esa and Vesa Co., Hannu Lehti,
who has offered friendship and sharp intellect during the process while ‘keeping it
real’ with the grounded working man energy. I also like to thank Sanni Kotima¨ki for
making me feel at home in academic circles right at the beginning with her humor
and warm personality. In addition, I thank my workroom buddy Antti Ka¨ha¨ri for
being a shield against all pessimistic thoughts with your positive outlook on life and
excellent discussions.
Next, I want to express my gratitude to the TITA-gang, each of whom had their
viii
own area of expertise that encouraged me to think outside the box. I’ll always cher-
ish the excursions we shared. I’ve been fortunate to be part of an fantastic commu-
nity and to have worked with or enjoyed the company of many incredible people:
Laura Heiskala, Laura Salonen, Markus Laaninen, Simo Arhippainen, Janne Salmi-
nen, Aleksi Karhula, Niko Eskelinen, Anna Grahn, Outi Sarpila, Lauri Ma¨kinen,
Pekka Ra¨sa¨nen, Patricia McMullin, Kirsti Sippel, Essi Ahlroth, Hannu Pihlajasalo,
Tero Pajunen, Elina Kilpi-Jakonen, Sampo Varjonen, Heta Po¨ylio¨, Vesa-Matti Pa-
sivaara, Pietari Kujala, Sanna Kailanheimo-Lo¨nnqvist, and Teemu Vauhkonen, just
to name a few. Special thanks to Jaana Ta¨hti for her tremendous assistance with
practical matters and navigating the complexities of administration.
Working on a dissertation needs financial support. I thank the Academy of Fin-
land for funding the TITA consortium project, where I began my research. I’m also
grateful to the INVEST flagship program for helping me finalize this thesis and join
the Horizon2020 MapIneq project. Thanks to Horizon2020 Beyond4.0 for work that
let me continue my thesis. Though I didn’t use them, I appreciate Turun Yliopis-
tosa¨a¨tio¨ for offering grants, as I found better long-term opportunities. Lastly, thanks
to the University of Turku for my university teacher position, which gave me teaching
experience and time for my thesis.
I wholeheartedly thank my friends, which I’ve had the privilege to know from
different walks of life. First and foremost, I’d like to thank the OG, brother-from-an-
another-mother Mika, whose friendship and life-changing conversations, led me to
this path. I also thank Juho for the awesome conversations and for the opportunity to
geek out on various cool things, not forgetting the importance of blowing off some
steam with great and “great” movies. I also thank Tuomas for bringing endless pos-
itive energy and for offering delightful chaos. I thank, ever the street-smart, Tarmo
for his deep insight on life and countless fun stories on your adventures.
Lastly, I would like to thank my family. I’m eternally grateful to my parents,
Leila and Urpo, for their unwavering support and love throughout my life. Their
guidance has enabled me to forge my own path. I also want to thank my big brother
and role model, Pasi, for being an extraordinary person and sharing my enthusiasm
for lifting weights, heavy music, and games. Similarly, I extend thanks to my big
sister, Kaija, and her family for their support. Lastly, I want to thank my mother,
who sadly did not see me graduate but always believed in me with such dedication,
as if my journey was meant to turn one fact into reality. I carry your wisdom in my
heart, which will be my guiding star whenever I stumble.
November 2024, Turku
Esa Karonen
ix
Table of Contents
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
Table of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x
Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii
List of Original Publications . . . . . . . . . . . . . . . . . . . . . . xiii
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2 GENERATIONAL FRAMEWORK . . . . . . . . . . . . . . . . . . 5
2.1 Life course perspective . . . . . . . . . . . . . . . . . . . . . 5
2.2 Generational perspective . . . . . . . . . . . . . . . . . . . . 7
3 GENERATIONAL ECONOMIC INEQUALITY . . . . . . . . . . 13
3.1 Sociological concepts of economic inequality . . . . . . . . . 13
3.2 Economic inequality from the life course paradigm . . . . . 14
3.3 Previous research connected to the holistic perspective . . 19
4 AIMS OF THE STUDY . . . . . . . . . . . . . . . . . . . . . . . . . 25
5 DATA AND METHODS . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.1 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.2 Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.2.1 Dependent variables . . . . . . . . . . . . . . . . . . 30
5.2.2 Independent variables . . . . . . . . . . . . . . . . . 31
5.3 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.3.1 Multilevel approach to between-cohort income in-
equality in Sub-study I . . . . . . . . . . . . . . . . . 33
5.3.2 Age-Period-Cohort methodology to measure in-
come and consumption in Sub-study II and III . . . 35
5.3.3 A linear model to measure wealth differences be-
tween cohorts in Sub-study IV . . . . . . . . . . . . . 40
x
6 RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
6.1 Summary of main findings . . . . . . . . . . . . . . . . . . . 41
6.2 Income inequality between cohorts (Sub-study I & II) . . . . 43
6.3 Consumption differences between top and bottom income
deciles (Sub study III) . . . . . . . . . . . . . . . . . . . . . . 44
6.4 Wealth accumulation between cohorts and the financial
deregulation (Sub study IV) . . . . . . . . . . . . . . . . . . . 45
7 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
7.1 Interpretation of the results . . . . . . . . . . . . . . . . . . . 47
7.2 Policy implications and recommendations . . . . . . . . . . . 51
7.3 Methodological considerations . . . . . . . . . . . . . . . . . 54
7.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
List of References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Original Publications . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
xi
Abbreviations
APC Age-period-cohort
APCD Age-period-cohort Detrended
APCGO Age-period-cohort Blinder-Oaxaca
APCT Age-period-cohort Trended
CGLIM Constrained generalized linear model
EGP Goldthorpe / Erikson / Portocarero class scheme
ESS The European Statistical System
HES Finnish Household Expenditure Surveys
IDS Income Distribution Statistics
IE Intrinsic estimator
ISCED The International Standard Classification of Education
ICC Intraclass correlation coefficient
OSF Official Statistics of Finland
xii
List of Original Publications
This dissertation is based on the following original publications, which are referred
to in the text by their Roman numerals:
I Karonen, Esa; Lehti, Hannu; Erola, Jani; Moisio, Pasi & Kuivalainen, Su-
san. Structural Changes and Generational Income Inequality Over the Life
Course. Submitted manuscript.
II Karonen, Esa & Niemela¨, Mikko: Life course perspective on economic
shocks and income inequality through age-period-cohort analysis: evidence
from Finland. Review of Income and Wealth, 2019, 66(2), 287-310.
III Karonen, Esa & Niemela¨, Mikko: Necessity-Rich, Leisure-Poor: The
Long-Term Relationship be-tween Income Cohorts and Consumption
through Age-Period-Cohort Analysis. Journal of Family and Economic is-
sues, 2022, vol(43), 599-620.
IV Karonen, Esa & Niemela¨, Mikko: Cohort differences in wealth develop-
ment over the life course: the case after of the opening of financial markets
in Finland, 1987–2016. Submitted manuscript.
The original publications have been reproduced with the permission of the copyright
holders.
xiii
1 Introduction
Things are scarce.
Even though one could argue that we are living in an era of abundance, especially
in a high-standard-of-living country like Finland, all societies are fundamentally tied
to this basic premise. Currently, debates about material and social resources primar-
ily focus on the question of ‘how the pie is sliced’ (Rycroft, 2013).
Since we do not live in a world of unlimited resources, one’s position in the so-
cioeconomic hierarchy is influenced by both the overall economic situation and the
opportunities available. This is the starting point of a complex discussion about in-
equality, which is accurately described as multidimensional (Fadda & Tridico, 2016).
Several scholars have proposed more comprehensive and multidimensional ap-
proaches to economic inequality, shifting the focus to various background factors
(see Atkinson & Bourguignon, 1982; Decancq & Lugo, 2012; Piketty, 2014; Sen,
1995; Tsui, 1995, 1999; Weymark, 2006). For example, Aaberge and Brandolini
(2015) recently advocated for a more multidimensional approach, arguing that eco-
nomic well-being is a result of many different attributes of human life and that in-
come, expenditure, or wealth alone is just a rough proxy of living quality (see also
Atkinson & Bourguignon, 1982). Much of the research on economic inequality has
narrowly focused on income distribution.
This dissertation aims to build upon previous research in two significant ways:
by employing a life course approach and by adopting a holistic view of income,
consumption, and wealth, rather than just focusing on income.
The first contribution recognizes that economic resources do not remain static
as individuals transition through various stages of life from childhood to retirement.
This means considering the intersecting factors of age, period, and birth time. Age
differences should be acknowledged because economic reality can change drasti-
cally depending on one’s life stage. Additionally, individuals are born at different
times, which determines the opportunities available for striving for better economic
outcomes. Thus, societal and economic factors unite a generation, as people are
bound together by their birth date and the conditions of their time (Mannheim, 1928;
Mayer, 2009). In addition, the resources of one generation significantly influence the
life changes and economic opportunities of the next generation (Anand & Sen, 2000;
Rawls, 1999).
The second major contribution is a holistic approach to economic resources.
1
Esa Karonen
Rather than focusing primarily on income, this dissertation aims to consider all
economic measures, namely, income, consumption, and wealth. While income and
wealth distribution are rightly seen as key forms of inequality (Piketty, 2014), with
income distribution often viewed as a critical indicator of overall inequality, this dis-
sertation boldly argues that a focus on income alone can create economic blind spots
that prevent deeper understanding. Income, consumption, and wealth do not exist
in isolation; one needs income to acquire goods and services and ideally save the
surplus for future use and financial security.
Moreover, the simple flow of resources cannot be disconnected from the real-
ity that income fluctuates throughout one’s life, affecting the opportunity to pur-
chase goods and accumulate savings. To delve deeper, each of these so-called ‘life
courses,’ from birth to death, is tied to the surrounding economic and social oppor-
tunities. These include, for example, educational opportunities, employment rates,
and the types of jobs available in the labour market. Some of these factors, such as
gender and family background (see Weber, 1922), are neither earned nor chosen but
are assigned to an individual. In other words, the openness of society and opportu-
nities reflects the level of equality of opportunity and the possibilities to accumulate
more resources. This ties the three economic measures to the first contribution, as if
we were to observe changes in age and not by birth cohort or period, we would only
see a snapshot of the larger issue of inequality. To integrate these two contributions,
this dissertation outlines three research objectives. These are divided by each eco-
nomic resource, measuring income, consumption, and wealth, while considering the
dimensions of age, period, and cohort.
First, this dissertation aims to measure income between generations and identify
what background factors contribute to income development throughout life. This
directly supplements previous studies in their multidimensional endeavour on back-
ground factors, as they lack a life course and generational perspective.
Second, in the area of consumption, I examine how expenditures on necessities
and leisure consumption relate to economic inequality between birth cohorts. This
question is connected to the minimum level of decent living and whether the gap
between the rich and poor has narrowed over time. On the other hand, investment
in leisure time can indicate whether those in more economically adverse situations
have increased opportunities for societal participation and self-improvement.
Third, in terms of wealth, I aim to study how not just wealth but all types of
assets and investments affect total wealth accumulation between birth cohorts over
different periods. More often, studies measure just gross or net wealth, but I aim to
expand on the issue of how wealth is constructed over the life course and generations.
This in-depth look is crucial, as it touches on the ability to repay loans and assess
whether financial security has changed.
With the results provided by these three main questions, I answer not only ‘how
the pie is sliced’ but also how and if these differences are found during the life course
2
Introduction
of different generations in Finland. Instead of looking at a single snapshot of income,
I aim to offer a cinematic universe of Finnish generations, which includes a series
of stories about their lives, focusing on the three economic components and how
different background factors influence them. Overall, the goal is to analyse how
lifetime economic outcomes have changed between generations, which are shaped
by opportunities and constraints in different periods.
This dissertation makes multiple contributions. First, the introduction of a life-
course approach allows us to study economic measures ‘on the move’ as the inter-
section of age, period, and cohort offers a more dynamic analysis, while traditional
research offers a static ‘snapshot’ using one of these dimensions. Second, the disser-
tation significantly contributes to the literature on economic inequality by adopting
a multidimensional view of the life course framework and measuring all the key
economic measures together in income, consumption, and wealth. Previously, mul-
tidimensionality has been used in the case of poverty measures, and now I introduce
the same concept to measure economic resources (see Aaberge & Brandolini, 2015;
Alkire et al., 2015; Levitas, 2007). Third, the dissertation uses high-quality register
data and other specialized datasets to provide new knowledge on the development of
economic inequality. Fourth, it contributes to the field of life-course and cohort anal-
yses by applying novel and advanced methods of age-period-cohort and multilevel
modelling techniques to the study of generational economic inequality.
It remains largely unknown how different periodic changes, such as financial pol-
icy or economic cycles, impact the economic prosperity of different generations. For
example, in Finland, there have been several significant changes in financial policy,
such as tax reforms in 1992, which separated the taxation system into labour and
capital tax. Similarly, financial deregulation in 1987 opened up a new world of in-
vestment opportunities for the general public. Additionally, economic shocks, like
the recession in the 1990s, had a major impact on income distribution, which likely
led to a cascading effect on wealth accumulation. As mentioned earlier, these events
can have a multifaceted impact on the long-term economic development of differ-
ent cohorts, as their life stage and social location within society differ from each
other. The extent of these effects is still unknown. Therefore, a holistic viewpoint on
economic inequality could provide a useful framework for planning targeted policy
measures. In this way, the dissertation generates knowledge that is useful for poli-
cymakers who wish to increase economic efficiency and implement taxation reforms
regarding lifetime economic development.
The dissertation is structured as follows: The second chapter discusses the gen-
erational framework and life course theory in sociology, which introduces the mul-
tidimensional aspect of economic thinking. The third chapter focuses on economic
inequality and how it is framed in the life-course paradigm, and how it explains the
three economic outcomes of income, consumption, and wealth. The model in this
chapter draws on the sociological life-course perspective and the literature on eco-
3
Esa Karonen
nomic inequalities. Chapter 2 can be seen as a more general outline of generational
theory and thinking, while the third chapter is more of a literature review for the
dissertation. The fourth chapter outlines the research aims and the design. The fifth
chapter introduces the data and methods used in each sub-study of the dissertation.
The sixth chapter provides an overview of the results obtained in the four research
articles. The last chapter is dedicated to discussing the scientific and practical impli-
cations.
4
2 GENERATIONAL FRAMEWORK
In this chapter, I introduce the life-course approach and the generational framework
in sociology. This lays a foundation for understanding how economic inequality and
social stratification connect with the perspective of age, period, and cohort, as they
contribute to the overall theoretical viewpoint. Last, I present the Finnish context of
generations as depicted in the research literature.
2.1 Life course perspective
Over the last 30 to 40 years, life-course research has blossomed at a high rate as an
interdisciplinary study of human lives from birth to death. Although life course is not
a complex theory in itself, it is utilized in various fields of study such as anthropology,
demography, economics, sociology, and developmental psychology, with sociology
serving as a significant disciplinary anchor. A longitudinal and life-course approach
has become prevalent across the social sciences, and longitudinal data collections
have proliferated, becoming a current objective of quantitative social science. For
instance, the importance of longitudinal studies on social science is highlighted in a
special issue of Science (Butz & Torrey, 2006), where longitudinal studies of the life
course are hailed as the most progressive and innovative research.
The life-course approach examines an individual’s life history, investigating, for
example, how early events influence future decisions and events such as employment
and unemployment, economic shocks, marriage and divorce, or engagement in crime
(Elder et al., 2003; Willson et al., 2007). A life course is defined as “a sequence
of socially defined events and roles that the individual enacts over time” (Giele &
Elder, 1998). More specifically, the approach focuses on the connection between
individuals and the historical and socioeconomic context in which these individuals
live (Willson et al., 2007). The method includes observations in different disciplines.
However, empirical research with a life course approach has not yet resulted in the
development of a formal theory (Mayer, 2009).
To define the field’s criteria and narrow it down to economic matters, it’s essential
to clarify what distinguishes the life course approach from many other overlapping
research areas. The following criteria represent an emerging consensus (Elder et al.,
2003; Fasang & Mayer, 2020; Mayer, 2004; Settersten, 2003). First, the life-course
approach studies changes in human life paths as alterations in personal characteristics
5
Esa Karonen
and transitions between states. These are considered over a long span of a lifetime,
such as from childhood to old age, rather than as specific episodes like the transition
to marriage or first birth, or narrow life phases. The most evident economic example
is an individual’s employment history and lifetime wealth accumulation (Johnson &
Schoeni, 2011; Mayer, 2018). There is also a strong assumption that one’s previous
life history significantly impacts later life outcomes (Bengtson et al., 2012, 1974).
An example of this could be better access to education due to geographical location,
which could provide an individual with a head start compared to their peers. Thus,
changes in human lives are investigated across a larger series of cohorts. This implies
that life course development is often analysed as the outcome of personal character-
istics and individual action, as well as cultural frames and institutional and structural
conditions related to micro-, meso-, and macro-levels of analysis.
Figure 1. Four elements of the life course paradigm.
To generalize, Giele and Elder (1998) proposed the life course theory, which is
grounded in four key principles (refer to Figure 1) which tie into historical time and
geographical place: lifespan development, human agency, timing of decisions, and
linked lives. These events and roles do not necessarily unfold in a predetermined
sequence but instead comprise the aggregate of the individual’s actual experience.
Consequently, the concept of life course suggests age-differentiated social phenom-
ena which are distinct from uniform life-cycle stages and the life span. The term ‘life
span’ refers to the duration of life and characteristics which are closely related to age
but vary minimally across time and place. Ageing and developmental change are
thus continuous processes experienced throughout life. In this way, the life course
mirrors the intersection of social and historical factors with personal biography and
development, within which the study of economic development and social change
can occur (Elder, 1994; Hareven, 1996). A more in-depth discussion of this model
in an economic context can be found in Chapter 3.
To contextualize the life course in analytical and generational terms, refer to
Figure 2, which displays a Lexis diagram illustrating the dynamics of age, period,
6
GENERATIONAL FRAMEWORK
and cohort. They represent the three basic building blocks of the theory mentioned
above: in Mannheim and Mayer’s life course theory, ‘cohort’ represents the genera-
tion. ‘Age’ and ‘period’ follow the causal form of individuals linked to their year of
birth.
Figure 2. Lexis diagram on the relationship of age, period and cohort.
The Lexis diagram graphically represents the age, period and cohort (APC) ef-
fects, which represent three distinct ways in which economic resources can change
over time, and researchers across the social sciences have long been interested in
differentiating and understanding these changes. First, individuals can age, meaning
that they change as they progress through their life course. Second, change can occur
over time due to differences between cohort groups, whereby as new cohorts replace
old cohorts, the social composition (and thus the income, consumption and wealth)
of society as a whole can change. Third, and finally, change can occur as a result of
period effects, whereby passage through time results in a change in economic cycles,
regardless of the actions of the individual (Bell & Jones, 2015.).
2.2 Generational perspective
Now that we have covered the life course approach, we can turn to the generational
perspective, as generations have distinct experiences throughout their lifetimes. The
notion of generation is widely used in everyday language to make sense of differ-
ences between age groupings in society and to place individuals and other people
within a historical timeframe. We commonly distinguish these groups using terms
7
Esa Karonen
like ‘my generation’ and ‘the older generation.’ For example, we refer to those who
grew up in the 1950s as belonging to ‘the fifties generation.’ We make clear distinc-
tions between groups of people and periods of time by using phrases such as ‘a few
generations ago,’ ‘the millennials,’ and ‘the generation gap.’ Despite widespread use
of the word ‘generation,’ contemporary sociologists have only recently begun to pay
attention to its significance (Pilcher, 1994.).
Some scholars, such as Finch (2018), argue that the overlooked aspect of the so-
ciology of generations parallels the lack of attention paid to the social significance of
age. Finch emphasized the necessity of developing a sociology of age by highlight-
ing its value in contributing to understanding of key sociological issues, including
the interplay of the biological and the social, the relationship between personal and
social change, and the intersection of biography and history (Finch, 2018; Pilcher,
1994).
Mannheim’s seminal essay, “The Problem of Generations,” is considered the
most systematic and fully developed theory of generation from a sociological per-
spective (Bengtson et al., 1974). One of the key reasons for its high praise is how
it firmly locates generations within socio-historical contexts and is part of a broader
sociological theory of knowledge. Indeed, Mannheim’s theory of generations must
be acknowledged as a fine-tuned element of his broader scope in the sociology of
knowledge. Mannheim was primarily concerned with examining social location in
terms of class factors (Abercrombie & Longhurst, 1983), although he also conceived
of social location in terms of generational factors.
Before delving into a theoretical account of generations and how it touches upon
several key sociological issues, some terminological clarifications must be addressed.
Many scholars who contribute to the study of generational analysis have noted that
the way in which Mannheim and others have used ‘generation’ is really in the sense
of ‘cohort’ and that this would be a more accurate term to employ (Pilcher, 1994).
This has a direct effect on Mannheim’s generational framework. Glenn (1976), for
example, notes that the technical term ‘generation’ has overlapping roots in kinship
terminology denoting the parent–child relationship. Instead, in Mannheim’s perspec-
tive, a ‘cohort’ is defined as people within a defined population who experience the
same significant event within a given period.
Terminological issues can be particularly important, especially when the sample
used in generational research is composed of members of family groups. In such in-
stances, individuals are generations in the kinship sense, and, on the other hand, they
are also generations in the cohort sense. To be more clear-cut and not to mix up the
two dimensions of generations embedded in research designs, scholars advocate the
use of ‘generation’ when reference is made to kinship relationships and ‘social gen-
eration’ when reference is made to any cohort-related phenomena (Pilcher, 1994.).
This distinction protects the research from liberal use of terminology while main-
taining links with the sociological tradition, which has roots in Mannheim’s use of
8
GENERATIONAL FRAMEWORK
the term ‘generation.’
The aforementioned terminological mismatch surrounding generational analysis
can act as a useful introduction to Mannheim’s generational sociology. The confu-
sion resulting from misapplication of ‘generation’ as a term arises out of, for exam-
ple, the conflation of biological generation with cohort and an insensitivity to the
multiple nature of time and to the complexity of biographical and historical con-
nections. It should be noted that I use ‘generation’ here as an indicator of social
generation, as a synonym for a cohort, and do not use it as a kinship distinction.
Figure 3. Generational framework in a life-course perspective.
Mannheim defined a generation as a group of individuals of similar ages whose
members have experienced a significant historical event within a set period of time
(Mannheim, 1928). According to Mannheim, the social consciousness and perspec-
tive of youth reaching maturity in a particular time and place is significantly influ-
enced by the major historical events of that era. Mannheim termed this as ‘genera-
tional location,’ where a generation experiences unique events, thus becoming a real,
distinct ‘generation.’ A key point, however, is that this major historical event has to
occur (to deviate from the ‘trend’) and has to involve the individuals in their early life
stage, thus shaping their lives, as later experiences will tend to receive meaning from
those early experiences. Mannheim states that mere chronological commonality is
not enough to produce a common generational consciousness. Without unique con-
tent, there is no unique ‘personality’ of a generation. Mannheim even stressed that
9
Esa Karonen
not every generation will develop a distinctive consciousness. Whether a generation
succeeds in developing a distinctive consciousness is significantly dependent on the
pace of social change.
In addition, Mannheim (1928) asserts that social change can occur gradually,
without the necessity for significant historical events. This suggests that such events
are more likely during periods of accelerated social and cultural change. Mannheim
also observed that members of a generation are internally divided by factors such as
location, culture, and class, which can influence their perspectives on various events,
indicating that they are not completely homogeneous. An example of a cultural
effect is the punk subculture in the United Kingdom in the 1970s. Within what
Mannheim termed the ‘generation in actuality,’ there can be varying responses to
specific historical situations, resulting in a stratification into ‘generational units’ or,
more broadly, ‘social generations.’
The key aspect of the social element of a distinct generation is the concept
of a generational ‘scar’ acquired through shared socialization. This concept con-
nects with how different birth cohorts grow up during a similar historical period
(Mannheim, 1928). Economically speaking, this could be seen as inequality, as
some cohorts might have an easier entry into the labour market due to their specific
economic conditions. Therefore, an economic upturn or downturn can significantly
influence how a given generation establishes itself during changing market condi-
tions. For instance, cohorts that reached adulthood during economic booms are more
likely to benefit from that favourable market situation. Conversely, cohorts ‘scarred’
by an economic downturn might be more risk-averse and have a more disadvantaged
economic path, viewed from a life course perspective (Malmendier & Nagel, 2011).
These observations have laid the foundation for the life course hypothesis and paved
the way for new developments in the theory, which mainly focuses on systematic
cycles of advantaged and disadvantaged generations (Myles, 2002, p. 138). This ap-
proach was spearheaded by numerous researchers (e.g. Campbell et al., 1976; Clark
& Oswald, 1994; Frey & Stutzer, 2002; Helliwell, 2003; Helliwell et al., 2012; Kah-
neman et al., 1999; Layard et al., 2014) who explored the impact of other outcomes,
such as income, employment, and educational qualifications, on well-being.
To provide a practical example, similarities between generations have been iden-
tified in the Finnish context. Although Erik Allardt (1981) claims that the compo-
sition of generations has not been a primary concern in Finnish sociology (see also
Toivonen, 2003), later research on lifestyles and biographies elevated generational
research to a new level in the 1980s, adopting the Mannheimian framework (see
Karisto, 2005; Roos, 1980; Virtanen, 2001). While cohort research had modest be-
ginnings, the newer generation of researchers, such as Jani Erola and Terhi-Anna
Wilska (2004), have made significant contributions to the field. A major factor driv-
ing the intensified research of generations is the improved quality and usability of
research data, such as the increased availability of register data. Overall, previ-
10
GENERATIONAL FRAMEWORK
ous research indicates that every framework on Finnish generations recognizes at
least three to four classes (Purhonen, 2007; Roos 1980; Virtanen 1999). A complete
breakdown of cohorts identified by previous research is provided in Table 1.
Table 1. Finnish context and structure of generations in extended form (Purhonen, 2007)
Generation Description
War generation 1900–1929 Generation born during the wars
Deprived generation 1930–1939 Rebuilding period after the war years
Baby–boomers 1940–1954 Largest birth cohort
Oil crisis 1955–1964 Labour market entry at the time of
oil crisis
Precarious generation 1965–1972 Generation first to grow up part of
welfare society
Recession generation 1973–1979 Labour market entry on the 1990’s
great recession
Y–generation 1980–1990 Digital natives; labour market entry
during financial crisis
Z–generation 1991–
Now, to link the generational perspective with the life course approach, it’s im-
portant to understand that individual life courses exist within each cohort, or in sim-
pler terms, each generation. This concept was illustrated in Figure 2 using a Lexis
diagram. The uniqueness of the average life course of a generation forms a dis-
tinct group, which is constituted by similar opportunities and experiences. These
opportunities and experiences are of utmost importance, as they are mechanisms that
shape the economic characteristics of each given cohort. This life course pattern is
generalized in Figure 4.
Figure 4. Distribution of environmental factors and skills over the life course.
11
Esa Karonen
There is a strong assumption that prior life history significantly impacts later life
outcomes. Mayer (2000, 2009) suggests that the focus of studies should be on the
mechanisms associated with inequalities across generations. Different demographic,
social, and economic structures can influence the income development of genera-
tions. These mechanisms have been the enigmatic areas of stratification research,
as the differences in income trajectories among cohorts remain largely unknown
(Mayer, 2009). The primary motivation of this dissertation is not only to answer
“how,” but also “why.” Hence, its main objective is to determine the extent to which
these mechanisms explain economic inequality between generations.
Decisions made during the life course can and will affect the expected income
level, which in turn impacts the accumulation of wealth and opportunities to consume
products and services. Factors such as health, gender, employment, career opportu-
nities, education, labour market competition, and parental social status all have an
impact on an individual’s outcome. Therefore, identifying which mechanism best
explains the extent of generational economic inequality is crucial to understanding
why these differences exist in both the theoretical and empirical realms.
12
3 GENERATIONAL ECONOMIC
INEQUALITY
In this chapter, I link the concept of economic inequality with the life course perspec-
tive. Firstly, I present the sociological viewpoint concerning economic inequality.
Secondly, I apply the life course paradigm model to generational economic inequal-
ity. This explains how social stratification and equality of opportunity are shaped
within the perspective of age, period, and cohort, contributing to the three economic
outcomes.
3.1 Sociological concepts of economic inequality
In the longstanding sociological tradition, five primary theoretical explanations for
economic inequality exist. These revolve around theoretical giants like Karl Marx
(1928), who interpreted economic inequality through the lens of class exploitation,
and Max Weber (1946), who concentrated on social closure between classes. Some
diverge from class theory, like Thorstein Veblen (1899), who focuses on individual
rank competition, and Davis and Moore (1945), who flip the argument around to ad-
vocate for the functional necessity of inequality. Lastly, Polanyi (2001) suggests that
poverty stems frommarkets, which, if left unchecked, can undermine the foundations
of social life and institutional mediation.
To summarize, these approaches to economic inequality are grounded in dif-
ferent sociological perspectives, revolving around three key themes: social classes,
individual competition, and institutions. However, there are no clear, unifying the-
ories on economic inequality within sociology, and these theories alone only offer
an unfocused snapshot of the broader inequality scheme, partly due to their tempo-
ral context. As these main themes are neither mutually exclusive nor exhaustive,
they should not be viewed as complete explanations in themselves. Here, the gen-
erational framework provides an opportunity for a more comprehensive viewpoint.
For instance, all the aforementioned theoretical and empirical foundations are not
static; they change over time. It is therefore logical to assume that the economic en-
vironment varies across generations, as they possess different class and occupational
structures, live under evolving institutions and policies, and face differing levels of
individual competition in the job market. In other words, the available opportunities
and the economic environmentt are the main factors that determine how each cohort
13
Esa Karonen
can develop their economic prowess throughout their lives.
Economic inequality, a term encompassing the distribution of income, consump-
tion, and wealth, arises from asymmetry in the distribution of monetary resources
amongst individuals. From this perspective, the intersection of timing and coinci-
dence can limit available economic opportunities, acting as a constraining factor for
the distribution of the aforementioned resources. To clarify the thesis’s argument,
economic inequality is the unequal distribution of economic resources and oppor-
tunities amongst different societal groups. Here, social stratification refers to a so-
ciety’s categorization of its people into groups based on socioeconomic factors like
income, expenditure, wealth, education, gender, occupation, or social status. Con-
sequently, stratification represents the relative social positions of individuals within
a social group, category, or unit. Importantly, the key social groups here are the
generations/cohorts and their respective age differences, i.e. the life course.
The question, “Why does (in)equality matter?” boils down to contrasting argu-
ments on needs between individuals or earnings based on individual effort (Sen,
1997). Hence, inequality can be viewed not merely as a measure of dispersion but
also as a measure of difference. The underlying idea here, which outlines the distinct
characteristics of generations, is the equality of opportunity. This is the assumption
that the placement of individuals in the social hierarchy is determined by some form
of competitive process, and all members of society are eligible to compete on equal
terms. Following the generational lens, if we consider the primary factors behind
the equality of opportunity (i.e. education, occupation), the measured variation and
their differences are associated with the generations and their life course trajectories,
which should reflect the total ‘inequality’ between the groups.
In summary, the distribution of three economic resources is determined by the
constraint of given opportunities in a specific time for each generation. In the next
section, this theoretical framework will be more formally constructed and linked to
existing life course models.
3.2 Economic inequality from the life course paradigm
Models on the life course can be divided into two primary perspectives. First, Elder
(1994) pinpointed four crucial elements that shape the life course. These include
the historical and geographical location, social relations, development of individ-
ual agency, and the significance of timing (see also Elder et al., 2003). The second
notable contributor is Giele, who, akin to Elder, developed an autonomous four-
function variant of Parsons’ (1966) model of the social system. This model is ap-
plied to the life course and social change. Elder’s model aligns closely with Giele’s
(Giele & Elder, 1998), with both paradigms echoing each other. Elder’s analysis
is interpreted through the individual, while Giele’s dimensions concentrate on the
relationships between the individual and the encompassing social structure. Merg-
14
GENERATIONAL ECONOMIC INEQUALITY
ing these two frameworks is beneficial for tracking the interaction between person
and environment, and the dynamic changes initiated by the individual within the
context of structural transformations, as illustrated in Figure 5 and the subsequent
paragraphs.
Figure 5. Key determinants and elements of the economic life course paradigm.
Figure 5 presents a variant of Giele’s and Elder’s life course models, modified
to accommodate the thesis’s generational perspective. All two-way interactions be-
tween factors are represented by solid black lines with arrow endpoints, while one-
way connections are depicted by grey dashed lines, with a single arrow endpoint
indicating the direction of the connection. The significant modification to the origi-
nal models is that each key factor is linked to the age-period-cohort dynamic, setting
certain limits on how cohorts can influence their economic trajectory throughout
life. It’s also crucial to understand the role of the individual within this dynamic,
as some factors are beyond their control, while others allow for individual agency,
albeit within certain constraints.
Initially, the historical period is the primary causal link since time establishes a
focal point for all cohorts and other elements in the life course model. The period ex-
plicitly determines the range of economic opportunities available to all cohorts, from
educational opportunities to labour market conditions, not to mention institutional re-
alities like the role of social insurance and income redistribution. Essentially, being
part of a cohort means existing within a specific economic environment. However,
15
Esa Karonen
the context of the period can differ from one location to another. For instance, in rural
areas, job opportunities are often limited, pushing individuals towards larger growth
centres where economic drivers like education, career choices, and social networks
are more abundant (Giannakis & Buggeman, 2020). It’s worth noting that only the
historical period can dictate each geographical location’s economic situation, as the
location cannot influence time. Therefore, while the historical period is beyond indi-
vidual control and depends on the time of birth, geographical location is something
an individual can impact by relocating.
From the perspective of cohorts and period, both general and unique aspects
of location influence shared experiences and can be perceived as socially and indi-
vidually patterned in ways that persist over time. For example, children who lived
through the Finnish economic recession in the 1990s experienced a unique histori-
cal phase compared to those born later and encountered varying conditions depend-
ing on the severity of the situation in their hometown (Verho, 2008). From the life
course perspective, birth timing is significant as it correlates with the life stage at
which these events occur: the impact of an economic downturn differs greatly for
an individual transitioning from school to work compared to someone who has al-
ready established their career (von Wachter, 2020). This could result in a delayed
career start, which could subsequently affect the lifelong accumulation of economic
resources and impede major investments, like purchasing a home (Kurz & Blossfeld,
2004; Oreopoulos et al., 2012). In addition to this dynamic, an individual’s location
during an economic downturn could either hinder or shield them depending on the
competitive landscape and opportunities available in the area.
The second component of the life course paradigm is the development of indi-
vidual agency, which can be interpreted as personal goal orientation. As depicted in
Figure 5, the historical period ‘hard codes’ the overall economic constraints, while
the physical location modifies the available choices (Roikonen, 2022). As previously
mentioned, the economic environment takes centre stage in this scenario. If we view
this environment as a combination of the labour market, educational opportunities,
and institutional elements – all of which form a dynamic system that persists and
evolves over time – individuals within this environment are motivated to meet its
requirements in line with their preferences (McQuaid & Lindsay, 2013; Roikonen,
2022). Therefore, individual agency is the freedom to satisfy personal needs through
proactive decision-making aimed at achieving economic security. These actions are
conducted within the constraints of personal efforts, capabilities, and available op-
portunities, which may include (but are not limited to) pursuing education, aiming
for specific careers, or starting a family (Perugini & Martino, 2008). Family or other
social relationships can particularly motivate certain individual choices, such as the
decision to relocate for better economic opportunities or remain in the same location
as the family.
To illustrate, gender plays a significant role in influencing individual agency
16
GENERATIONAL ECONOMIC INEQUALITY
through gender norms and family dynamics at a given time. Women are more likely
to take breaks from the workforce for caregiving responsibilities, such as raising chil-
dren or caring for elderly family members, leading to lower lifetime earnings and
reduced savings. However, government policies, such as parental leave and child-
care support, can impact gender income disparities by enabling men and women to
balance work and family responsibilities more equitably, influencing the individual
agency.
To apply the above in the context of cohorts and demonstrate the interconnect-
edness of these factors, individual choices can alter the characteristics and social
fabric of a location over time. For instance, in Finland, Salo was a major industrial
city during Nokia’s peak, but after the foreclosure from 2008 to 2010, the city’s de-
mography drastically changed within a few years as people opted for employment in
more lucrative cities (Kangas & Karonen, 2023). Those born in the area during the
right period could take advantage of the plethora of jobs and economic prosperity in
the area, but subsequent cohorts are strongly encouraged to seek opportunities else-
where, without the benefit of early employment options. In the case of Salo, a dual
problem emerged as highly specialized and educated individuals found no suitable
jobs, and even manufacturing trades faced a similar issue due to the lack of demand
for manual labour at such a high volume. Simultaneously, the area’s overall prosper-
ity and economic drive declined, triggering a domino effect of labour migration. As
more people choose to relocate, it also impacts social relationships, the composition
of social classes, and networks, creating additional pressure, coupled with dwindling
economic opportunities, to do the same. This exemplifies the push-pull effect in the
life course paradigm, as individual agency and the pursuit of economic security also
influence social relationships and vice versa.
The third element is social relations and linked lives, which broadly encom-
passes the integration of social groups. In this context, social actions at cultural,
institutional, and social levels interact not only as parts of a whole but also as the
result of other individuals who share similar experiences. Expectations, norms, or
social institutions vary between cohorts. For instance, differences between groups
can be found in various aspects such as family backgrounds or experiences in work,
education, and other domains (Webster et al., 2022). To illustrate, during the eco-
nomic recession in the 1990s, certain social classes experienced the economic down-
turn more severely, and both the parents and their children had differing experiences
compared to those with more secure social positions (Haapakorva et al., 2017). This
led to different generational experiences of the same event, which uniquely identified
cohorts, and could have influenced certain economic choices made by the individuals
(Mayer, 2009; Wrzus et al., 2013). Social relations are a complex matter in the life
course paradigm, as individual agency either shapes or is affected by social relations
and geographical location.
In conclusion, all the elements mentioned above are linked and shaped by the
17
Esa Karonen
timing and coincidence of economic events. The timing of birth sets a fixed point
around which other events occur within a historical period. Individuals and cohorts
attempt to achieve their goals, responding to the timing of external events and en-
gaging in behaviour based on available options and resources. Thus, the timing of
life events intersecting with age and cohort can be understood as both passive and
active adaptation for achieving economic goals. Examples include how and when a
person accumulates income, wealth, and education, chooses an occupation, or starts
a family.
To revisit a previous example, the factory closure in Salo is a local event, which
is beyond the individual’s control and can serve as a point where the economic real-
ity changes depending on the life stage (Kangas & Karonen, 2023). Specifically, if
the individual is at the stage of transitioning from school to work, the closure could
mean foreclosure of the possibility of beginning a career for which the person has
educated themselves for over 5 years. This could lead to a later entry into the labour
market and directly affect economic trajectories over the life course. Here, the tim-
ing of the transition, available economic opportunities, and the position in the life
course compared to the event come together, illustrating the multidimensionality and
importance of generational analysis.
Focusing on the outcome of the model, all the elements of individual choice, lo-
cation, historical period, social relations, and the dimension of timing play a part in
economic life course trajectories. Simply put, we can measure how each cohort’s
income, consumption, and wealth have developed over the life course. Each of these
measurements represents a dependent variable, which offers a holistic viewpoint on
cohort economic history: income, consumption, and wealth do not operate in a vac-
uum, as one needs income to acquire goods and services and ideally save the excess
for later use and to reinforce their financial security.
Furthermore, the models can be tailored to accommodate multiple variables
within the life course paradigm. For instance, individual agency can be consid-
ered through years of education, historical periods can be represented by the year
of measurement, and social relations can be depicted by occupational or social class.
Additionally, some variables could represent multiple issues, such as gender, which
can guide individual choices (preferences in the educational- and occupational field)
and time (gender norms, parental policy etc.). When considering timing, a variety of
methods utilizing age-period-cohort modelling can be employed to account for the
individual’s age and year of birth. Once this dynamic across age, period, and cohorts
is captured among all three economic measures, this thesis provides a more precise
answer to the question of generational economic inequality than ever before.
18
GENERATIONAL ECONOMIC INEQUALITY
3.3 Previous research connected to the holistic per-
spective
The holistic economic perspective is built upon the causal relationship between in-
come, consumption, and wealth. One needs a stable and suitable level of income to
accumulate wealth and meet various expenditure needs. It is understood that these
three components vary based on life circumstances, age, environmental demands,
and preferences. In this section, I present a descriptive outlook of the holistic eco-
nomic perspective by connecting major life course events from previous sections to
contextualize the long-term situation of economic inequality in Finland.
To start, let’s begin with the root cause of economic accumulation by focusing
on income. Through the lens of the economic life course paradigm, previous studies
on income inequality between cohorts can be summarized into five main results.
First, studies on European welfare states and the United States reveal discrepancies
in income accumulation between cohorts, indicating that being born in the ‘Baby
Boomer’ generation provides an advantage (Chauvel, 2013; Chauvel & Schroder,
2015; Freedman, 2017). Second, research on income and wealth inequality reports
that younger generations have lower living standards than their parents at similar
ages in Great Britain (Crawford et al., 2016), Europe (Stand & Rising, 2011), the
United States (Chetty et al., 2017), Canada (Kershaw, 2015), and Australia (Daley &
Wood, 2014).
These two points are also reflected in the findings from Finland. The ‘Baby
Boomer’ generation appears to have the highest income cohort compared to other
generations, and studies show discrepancies in income accumulation between co-
horts (Pekkala & Lucas, 2007; Riihela¨ 2006; Suoniemi 2012). Similarly, studies
have found that the younger generation is a victim of unfortunate timing in an eco-
nomic sense, as they faced an economic shock, leading to a lower income level than
expected (Riihela¨, 2006).
Third, focusing on period effects, economic shocks have been identified as a sig-
nificant influence on inter-cohort income inequality. The main hypothesis suggests
that cohorts entering the job market during times of austerity and economic downturn
are – compared to cohorts born during an economic upturn – in a more disadvantaged
position in terms of attaining similar career opportunities. Fourth, connected to this,
the effectiveness of welfare state institutions in buffering external shocks, particu-
larly the generosity of the income transfer system, is known to fluctuate over time.
This impacts how different socioeconomic and demographic factors contribute to in-
dividuals’ income trajectories (McCall & Percheski, 2010). For instance, younger
generations have an 8 percent lower expected income than older generations, a re-
sult of the economic situation and different socio-political reforms (Berloffa & Villa,
2010).
Both the third and fourth points are applicable in the Finnish context. The era of
19
Esa Karonen
welfare state expansion in the 1960s and 1970s decreased income inequality regard-
less of the income concept. Additionally, during the welfare state expansion, Finland
underwent a phase of educational expansion, which led to the increased importance
of growth centres and a shift in the occupational structure. The agricultural sector di-
minished in favour of a more service-based economy (Pekkala & Lucas, 2007). Thus,
overall income did increase until the early 1990s (Blomgren et al., 2014). After this
point, a significant drop in income is observed, revealing the economic recession of
the 1990s, which heavily impacted income. From a comparative perspective, the in-
crease in income inequality was exceptionally rapid and steep in Finland during the
period between 1995 and 2002 (OECD, 2008). The main hypothesis suggests that
cohorts entering the job market during times of austerity and economic downturn are
– compared to cohorts born during an economic upturn – in a more disadvantaged
position in terms of attaining similar career opportunities (Berloffa & Villa, 2010).
Fifth, in addition to economic development, the increased individual agency to
acquire human capital has changed through educational expansion, which is linked
to the returns to income from certain educational fields (Becker, 1962; Mincer, 1958;
Pekkala & Lucas, 2004). Older generations had the opportunity to work towards
higher education but also benefited from expanding job markets and low competition
within the same educational field. As educational expansion progressed, younger
generations had to compete with an ever-growing pool of highly educated individu-
als for the same jobs. Therefore, according to supply and demand, higher degrees
hold less profit-making value for younger generations than for older generations.
The older generations reaped the benefits of starting their careers in an optimal job
market situation while also gaining work experience. In addition, these changes be-
tween cohorts have been especially significant for women who have made gains in
educational attainment, fields of study can influence income. Women may still be
underrepresented in high-paying fields such as science, technology, engineering and
mathematics.
As a result of economic cycles, structural changes, and educational expansion,
prior research has indicated that income typically exhibits an inverted U-shape pat-
tern, increasing with age and then slightly declining as individuals transition into
retirement (Gangl, 2005). However, even taxpayers aged 60 years or older tend to
earn above-average incomes (Alessie & De Ree, 2009; Bo¨nke et al., 2015). This
suggests that as the Baby Boomer generation continues to progress into their peak
earning years, there will be a higher proportion of high-earning individuals compared
to younger, lower-earning individuals, contributing to a perceived rise in inequality.
A similar pattern can be observed in Finland when examining the age distribution
of income; the income trajectory across different age groups also forms an inverted
U-shape (Riihela¨ et al., 2014). Notably, those between the ages of 30 and 40 years
experience a slight delay before their income trajectory increases more rapidly. This
phenomenon can be attributed to gender differences in income, as the 30–40 age
20
GENERATIONAL ECONOMIC INEQUALITY
bracket coincides with the child-rearing stage, which can impact women’s earnings.
The peak of one’s career appears to occur between the ages of 55 and 60 years, half
a decade prior to the typical retirement age.
Expenditures are intrinsically linked to income, thereby establishing a budget
constraint. Without savings or loans, an individual’s spending cannot exceed their
income. As such, income and consumption tend to increase concurrently, with the
ideal situation being that expenditures fall below income, enabling the surplus to be
saved for future use. If consumption surpasses income, it inevitably depletes wealth,
indicating an overspending situation.
Moreover, consumption is a vital measure, as income alone may not accurately
reflect the long-term resources available to a household or individual (Blundell &
Preston, 1998; Uusitalo, 1980). Temporary fluctuations in income can distort the true
economic status of a household, especially when borrowing or saving mechanisms
are utilized to balance out consumption patterns.
Sociologists widely agree that consumption patterns vary throughout an indi-
vidual’s life course, with the desire to purchase certain consumer goods fluctuating
based on life stage, consumer culture, and market conditions (Felson, 1976; Hirsch,
2005; Nicosia & Mayer, 1976). Research has consistently shown that life events
significantly influence consumption preferences (Gourinchas & Parker, 2002; Kol-
srud et al., 2017), including health transitions and the onset of parenthood (Sto¨ver,
2012). In the same vein, numerous variables that change over time, such as family
structure, age, class, and occupation, have a strong correlation with spending patterns
(Bihagen, 1999; Blundell & Preston, 1995; Cohen, 2016; Heslop, 1987; Katz-Gerro,
2003; Katz-Gerro & Talmud, 2005; Koelln et al., 1995; Salcedo & Izquierdo Llanes,
2020).
From the life course perspective, the central concept is that individuals’ con-
sumption propensities vary across cohorts and periods, and different cohorts exhibit
unique life course patterns. For instance, starting a family increases spending needs
beyond the individual, whereas single students can manage with less. Consumption
peaks around the age of 50 and decreases thereafter into retirement (Lim & Zeng,
2016). It is logical that at a young age, income levels are relatively low, while con-
sumption escalates later in life when individuals acquire a home and start a family.
When income decreases post-retirement, expenditures are optimized within this con-
straint.
Similarly, from a cohort perspective, research indicates that as individuals’ in-
come and wealth increase, they diversify their spending; this behaviour is believed to
be driven by resource-optimizing tendencies (Deutsch et al., 2015; Krueger & Perri,
2006; Ra¨sanen, 2003). In low-income groups, individuals typically exhibit con-
centrated spending patterns, but these tend to diversify as income rises (Fernandez-
Villaverde & Krueger, 2011; Krueger & Perri, 2006; Ra¨sanen, 2003). Furthermore,
the level of heterogeneity in expenditure diversity increases alongside income (Chai
21
Esa Karonen
et al., 2015).
In Finland, expenditures have risen over the years, reflecting the overall increase
in income. However, this trend is not always consistent. During the economic re-
cession of the 1990s, consumption rates remained stable, as economic crises affect
incomes and relative prices (Dutt & Padmanabhan, 2011; Hampson & McGoldrick,
2013; Mckenzie et al., 2011; Ra¨sanen, 2003). Nonetheless, the consumption of ne-
cessities remains unchanged despite economic fluctuations, and short-term changes
may be offset by utilizing savings, underscoring the point made by the aforemen-
tioned research (Blundell & Preston, 1995, 1998; Friedman, 1957; Slesnick, 2001).
However, a cohort effect could be at play, as previous research on consumer be-
haviour during an economic crisis has found that younger generations tried to main-
tain their consumption level, whereas older generations aimed to reduce their spend-
ing (Urbonavicˇius & Pikturniene˙, 2010). The divergence in economic behaviours
can be attributed to the older generation’s ability to exhibit more steady and logical
purchasing behaviour in relation to income, as opposed to their younger counterparts.
Finally, wealth accumulation has often been approached from the perspective of
the life cycle hypothesis, which posits that people aim to maintain the same con-
sumption level throughout their lives when their income is low by taking on debt
when they are young and using their savings later in life (Modigliani, 1966). Here,
wealth accumulation is primarily achieved by saving during peak earning years when
income is at an all-time high. The hypothesis assumes that wealth accumulation will
follow a ‘hump-shaped’ pattern: low at the beginning of adulthood and in old age,
and peaking in the middle. However, empirical studies over the years have ques-
tioned this conclusion, as data suggest that retirees do not deplete their wealth as
quickly as the model predicts (Alonso-Garcı´a et al., 2022).
In Finland, two significant historical period changes have major implications for
the wealth trajectories and wealth distribution of cohorts. These events are new mar-
ket opportunities, tax reforms, and economic cycles.
First, in the Finnish context, until the 1980s, the financial market was regulated,
including interest rates, foreign exchange rates, and the import and export of currency
(Pekkarinen & Vartiainen, 1993). During the 1980s, the Bank of Finland began a
slow deregulation process, allowing banks to handle foreign exchange affairs, and by
the mid-1980s, the banks’ lending rates were deregulated. Finally, market operations
opened in 1987, marking the birth of a modern financial market (Jonung et al., 2009).
This opened up practical means for the general public to invest and use financial
instruments. The context of financial markets presents a unique opportunity, as the
dataset for this dissertation covers the period, which could reveal potential effects on
cohort acquisition of financial wealth.
The second major change was the shift in tax reforms. Until the end of 1992,
Finland had a global income tax system where all income was treated the same. The
reform separated the taxation system into labour and capital tax, followed by a shift
22
GENERATIONAL ECONOMIC INEQUALITY
from earnings to capital income. In practical terms, these changes meant a significant
gap between the marginal tax rate on labour and capital income for many taxpayers,
benefiting those with more capital income and financial investments (Pirttila¨ & Selin,
2011.).
Third, economic cycles, such as the recession of the 1990s, are significant peri-
odic influences that can lead to increased unemployment rates among young adults
and hinder wealth accumulation. Within the context of life stages, those who have
established long and well-defined careers in their later years may not experience the
same impact from an economic downturn as those attempting to enter the labour
market for the first time (Frankenberg et al., 2003; Lovenheim & Reynolds, 2013).
Consequently, cohorts affected by economic recessions may tend to be more cautious
in taking risks and might have limited options for wealth accumulation throughout
their lifetime (Attanasio et al., 1999; Cagetti, 2003; Lusardi et al., 2017; Malmendier
& Nagel, 2011).
Similarly, as in the case of income, the second factor in periodical changes is
economic shocks, which can have varying effects across generations. Those who
have long and well-defined careers at a later age may not be similarly affected by
an economic shock compared to those trying to enter the labour market for the first
time (Frankenberg et al., 2003; Lovenheim & Reynolds, 2013). Hence, cohorts af-
fected by an economic downturn may be more risk-averse and have fewer wealth-
building options available during their life course (Attanasio et al., 1999; Cagetti,
2003; Lusardi et al., 2017; Malmendier & Nagel, 2011). In the same vein, since
households need stable income and employment to purchase property, these spe-
cific cohorts might have postponed their initial home acquisition. This is pivotal for
wealth accumulation, as prior findings on the periodic impact on wealth reveal that,
irrespective of time and location, real estate constitutes a significant portion of net
worth (Christelis et al., 2009; Kolb et al., 2013; Sierminska & Takhtamanova, 2007).
From the perspective of a cohort, past findings indicate a link between genera-
tional wealth, featuring strong persistence of wealth across generations, particularly
at the higher end of the distribution (Charles &Hurst, 2003; Conley &Glauber, 2008;
Hansen, 2014; Mulligan, 1997; Pfeffer & Killewald, 2018). This phenomenon, in
the context of Finland, can be traced back to the significant tax reform in 1993. The
resulting gap between the marginal tax rates on labour and capital income was advan-
tageous for those with higher capital income and financial investments. It is plausible
to assume that these changes significantly affected wealth accumulation across gen-
erations, as previous findings suggest that the dual-taxation reform in 1993 boosted
capital gains at the top of the income distribution. Further, higher inflation rates serve
as a broader economic factor that impacts inter-cohort wealth inequality. They tend
to favour the wealth positions of younger, middle-class households at the expense of
older and wealthier households (Doepke & Schneider, 2006).
Age plays a crucial role in the distribution of wealth, with the assumption that
23
Esa Karonen
wealth is gradually built up through savings, inheritances, and other forms of capi-
tal transfers. Studies on age-related variations have consistently shown that wealth
generally escalates throughout one’s life until reaching the age of 60 (Hurst et al.,
1998; Sierminska & Takhtamanova, 2007). In Finland, the pattern of net worth fol-
lows a similar path, with a noticeable decline in older age groups. Interestingly, the
peak net worth tends to lie within the 65–74 year age group, which is slightly older
compared to other countries, although this age range can fluctuate depending on the
time period (Hurst et al., 1998). In contrast, the distribution of assets before debt
reduction appears to be more balanced across different age groups, with the 45–54
and 65–74 year age groups having almost equal average assets (Torma¨lehto, 2018).
It’s important to note that wealth accumulation is influenced by certain factors. For
instance, young adults with lower net worth who have invested in higher education
may not currently have a strong wealth position, but this doesn’t necessarily reflect
their long-term financial potential (Conley, 2001; Conley & Ryvicker, 2005).
24
4 AIMS OF THE STUDY
In this chapter, I address the primary objectives of the research, which involve study-
ing generational economic inequality from a holistic perspective. Practically speak-
ing, it serves as a classifying guideline: the major research questions are categorized
according to the type of resources, split into the domains of income, consumption,
and wealth. With this framework in mind, the dissertation seeks to answer three main
questions:
1. Income
• How has generational income inequality changed, and how have different
structural and socioeconomic mechanisms contributed to income differ-
ences? (Sub-studies I and II)
2. Consumption
• How has generational consumption inequality changed within income
distribution? (Sub-study III)
3. Wealth
• How has generational wealth distribution changed, and how does the
opening up of financial markets relate to changes in wealth between gen-
erations? (Sub-study IV)
The unique aspect of this dissertation is that it takes into account both the life
course of each cohort and the period. Thus, three factors are at play: age, period,
and cohort. Rather than a simple snapshot of a certain time or age, research was
conducted to control for these dimensions. State-of-the-art methods and high-quality
datasets were used to tackle this multidimensional challenge. Furthermore, a holistic
framework is applied in economic measures, as each sub-study is divided to measure
income, consumption, and wealth. As a result, the findings show cohort differences
in all types of economic resources over the life course, providing a more precise
perspective on economic inequality.
The first research aim, income, contains two sub-studies that differ in nature to
cover the scope of the aim. In the first study, a novel methodology is applied to
25
Esa Karonen
measure income equality on the dimensions of age, period, and cohort. The main
objective is to observe if non-linear variation around the linear trend of income has
benefited some birth cohorts. The central idea is that, by taking into account the lin-
ear development of income, one can fairly observe which generations have benefited
the most. Various background factors have been considered to determine if the life
course, period, or timing of birth plays a role in generational income development.
In the second sub-study, high-quality register data are used to delve deeper into the
background factors contributing to income inequality between generations. By using
data that track the entire Finnish population, the dissertation measured how income
inequality has evolved during the life course between generations and how much the
background factors contribute to overall income inequality. This directly contributes
to previous research in a novel way, as it has primarily focused on the background
factors directly and not on how they vary according to period, timing, or life course.
The second aim is consumption. The third sub-study focuses on how expendi-
tures in necessities and leisure consumption between high- and low-income house-
holds connect to economic inequality between birth cohorts. The innovative concept
here is to compare these two subgroups, as the gap between the ‘haves and have-nots’
is related to societal questions of fairness and equal opportunity to participate in so-
ciety’s functions. Thus, if the gap between the rich and poor has narrowed over time,
it should indicate that society has become more inclusive over time. For example,
investment in leisure time can show whether more economically disadvantaged indi-
viduals can participate in society and its functions – in other words, whether society
and generations more well-off and not merely surviving, as they can afford to have
personal goals and interests in life.
In the last research aim, in the fourth sub-study, I examined how not just wealth
but all asset types and investments affect the total wealth accumulation between birth
cohorts. Here, wealth represents the financial security development of generations.
If between-generation investment in financial assets and savings has increased, this
could suggest that the overall economy provides a more financially secure base for
newer generations. In addition, the novel idea is to use rich data to measure beyond
gross or net wealth, touching on different financial asset types and accounts. Fur-
thermore, the Finnish case provides a unique perspective on the matter, as the dataset
starts at the opening of the financial markets, which sets the measurement point at
different life stages for different generations. Thus, it opens up the possibility to ob-
serve how readily different generations at various stages of life react to new financial
opportunities, and if the matter of timing is producing more economic (in)equality.
The empirical framework of the research is demonstrated in Table 2. The re-
search tasks are divided into four main questions.
26
AIMS OF THE STUDY
Table 2. Descriptive table of the dissertation’s sub-studies, research questions, data, and
methods.
Research questions Study Data Methods
RQ1 SUB STUDY I FOLK Generalized
- How much generational Generational differences register data Estimation
income inequality explains in income trajectories in Equations
total income inequality during Nordic welfare state (GEE) &
1970–2018 for men and women Intraclass
in Finland Correlation
Coeffecients
RQ2
- How much diffrent
Income structural factors contribute to
generational income inequality
RQ1 SUB STUDY II Finnish APC
- How income distribution has Life course perspective Income Detrended
developed across periodic on economic shocks and Distribution -model
economic fluctuations in income inequality through Statistics
relation to cohorts and age-period-cohort analysis: (IDS)
age group evidence from Finland
RQ1 SUB STUDY III Finnish APCGO
- How inter-cohort income and Necessity-rich, leisure- Household -model
spending profiles in high- and poor: the long-term Expenditure
low-income deciles have relationship between Surveys
changed income cohorts and (HES)
Consumption over the years 1966–2016
consumption through
RQ2 age-period-cohort
- How much inter-cohort analysis
consumption profiles between
high- and low- income differ in
specific categories of
consumption
RQ1 SUB STUDY IV Finnish OLS
- How between-cohort wealth Inter-cohort wealth Household regression
inequality has developed development in Finland, Assets interaction
Wealth from 1987 to 2016 1987–2016 (HAS models
RQ2
- How wealth trajectories over
age have developed between
cohorts
27
5 DATA AND METHODS
In this chapter, I go through the main research questions and how they have been
addressed. This section contains discussion on how these questions are answered,
including technical details on their data sources, measurement units, and various
methods.
5.1 Data
This thesis utilizes two primary types of data sources to measure income, consump-
tion, and wealth: income, expenditure, and wealth surveys and tax register data. The
first two sub-studies, which focus on income distribution, employ Finnish Income
Distribution Statistics (IDS) and FOLK register data. It should be noted that the IDS
survey data is constructed from the register data used in the second sub-study, render-
ing them analytically comparable. The third sub-study employs Finnish Household
Expenditure Surveys (HESs), while the fourth uses the Household Assets dataset. All
these datasets are administrated and maintained by the Official Statistics of Finland
(OSF).
A common denominator for all the datasets is the register data, which form the
basis for a large portion of the surveys. This implies that all information regarding a
reference person’s taxable monetary sources and personal characteristics are derived
from the registers. Conversely, information that cannot be sourced from registers,
such as wealth, non-taxable information, and household expenditures, is gathered
through surveys. In essence, Sub-studies II to IV are sourced from the first sub-
study, which also serves as the foundation for sample selection. This shared core in
each sub-study enhances the validity of comparisons made between the results when
they are examined comprehensively in the discussion section.
As previously mentioned, the first two sub-studies are utilized to analyse the ini-
tial research question on income. The analysis of the first sub-study uses the FOLK
register data from the OSF, which provides information on the total population. Our
sample from this data includes 4.4 million individuals from 1970 to 2018. Each
person was tracked until the end of the data period, and any new family members
were added to the panel and monitored. The data was amalgamated with informa-
tion from the population register, tax register, and several other official registers (see
OFS, 2024). This data enabled us to follow the income trajectories of cohorts born
28
DATA AND METHODS
between 1920–1989 for four decades of the working-age population, aged 20–65
years.
Sub-study II is based on the Finnish IDS and the Finnish HES provided by Statis-
tics Finland. Both datasets are part of the OSF and the European Statistical System
(ESS) series. HES provides data on incomes and expenditures for the years 1966,
1971, 1976, 1981, and 1985. The data are partly derived from interviews, and since
the 1971 survey, they have also been obtained from official registers. The IDS pro-
vides data on incomes, collected annually since 1987. Thus, the analysis period in
this study spans from 1966 to 2015. In both cases, income data are gathered from
tax and other registers and are generally considered high-quality. The data are har-
monized to provide a representative sample of the Finnish population. The primary
unit of analysis is the household. Data are multiplied to the level of the total popu-
lation by using special weights included in the household surveys. The person with
the highest personal income is chosen as the household’s reference person, serving
as a proxy for demographic and background status. The income variables measure
both the household’s income and the individual income of the household’s reference
person.
For the second research question on consumption, the empirical analyses in Sub-
study III are based on the Finnish household expenditure survey (HES) conducted by
Statistics Finland. This dataset is part of the OSF and the ESS series. The HESs pro-
vide data on households’ annual expenditures for various purposes, such as food and
transportation. They also provide extensive information on households’ structures,
activities, durable goods, housing conditions, income, and social services benefits.
From 1966 to 1990, the survey was conducted regularly at five-year intervals. From
1994 to 1996, the survey was conducted annually. Subsequent HESs were conducted
in 1998, 2001, 2006, 2012, and 2016. Data are partly derived from interviews and,
since the 1971 survey, from official registers. The household budget survey data are
collected via interviews, diaries, and purchase receipts kept by households, as well
as from administrative registers. The data collection process includes interviews
about a household’s background, ownership and purchase of durable goods, residen-
tial costs, and other information. After the main interview, households keep a diary
of their consumption expenditures and retain receipts from their purchases for two
weeks. Demographic and income data are derived from registers.
Finally, the third research question centres on wealth. Sub-study IV utilizes the
OSF Household Assets dataset. This wealth survey delineates household assets,
including their total amount, structure, and distribution among various population
groups. The survey also encompasses other aspects that impact the financial standing
of households, such as income and debts. It also includes data on debt and income.
Data on wealth are procured using different estimation methods, with register data
serving as the data sources. The dataset ranges from 1987 to 2016. It’s noteworthy
that there have been certain changes in data collection within the Household Assets
29
Esa Karonen
dataset. In 1987, 1988, 1994, and 1998, the data were gathered as an interview study
linked with the data collection of IDS. The 2004 survey (the Housing and Wealth
Survey) was a distinct survey. The 2009 survey is rooted in a sample of income
statistics, but with a revised methodology excluding separate data collection. IDS
for the sample households are associated with, for instance, housing, shareholdings,
and vehicle registration information. Not all asset types are available for statistical
use registry information, necessitating the employment of various estimation meth-
ods. Due to the revised methodology, the comparability differs from one asset to
another.
It’s worth noting that since the research involves individual citizens, I have con-
sidered the ethical issues pertaining to data management. Firstly, the datasets employ
pseudonyms as ID codes, ensuring no individual can be identified from them. Sec-
ondly, my co-authors and I have formally signed non-disclosure agreements with
Statistics Finland to refrain from distributing the data or using any information that
could conflict with laws on the use of the statistics. The data are securely stored
behind password protection, while the register data can only be accessed through a
remote network on Statistics Finland’s servers. This research poses no risk to the
well-being of any original survey participants or individuals amongst the register
data. All datasets were collected before the implementation of the General Data
Protection Regulation.
5.2 Indicators
5.2.1 Dependent variables
This dissertation addresses the three primary research questions using indicators fo-
cused on income, consumption, and wealth. Each variable is adjusted for inflation
and equivalized when household measures are applied.
Sub-study I utilizes register data, providing an opportunity to measure individual
income over time. Specifically, the study employs taxable annual income as the de-
pendent variable, which is both inflation-adjusted and log-transformed. As suggested
by previous research, income is adjusted for inflation to eliminate the impact of price
inflation on data. The log-transformation is implemented to convert the skewed in-
come distribution into a normally distributed one, catering to the requirements of
the statistical models (see Canberra Group, 2011). The annual income variable en-
compasses all taxable annual (pre-tax) income: earned income, capital income, and
the majority of social benefits. Despite the data’s comprehensive coverage, they do
not include some non-taxable incomes, such as means-tested, last-resort social assis-
tance benefits. It should be noted that disposable income could not be utilized, as the
registers contain this information only for a few measurement years, making annual
taxable income the second-best option.
30
DATA AND METHODS
In the second sub-study, the dependent variable is the annual household dispos-
able income. This includes monetary income items and benefits in kind related to
employment relationships. However, it does not include imputed income items, such
as imputed rent (see OSF, 2015a). The household income is not top-coded. In the
HES statistical years of 1966–1985, all income variables utilized the older Finnish
national currency, the Markka (FIM). To harmonize the variables, all monetary val-
ues were converted from the old currency, FIM, to EUR. Additionally, the income
has been log-transformed and equivalized with the modified Oxford scale annual
household disposable income, which has been adjusted for inflation.
Sub-study III has multiple dependent variables, which are equivalized and ad-
justed for inflation in euros. The variables used include money income, money
spending, food and grocery expenditures, and cultural and leisure-time expenditures.
The aggregate food and grocery expenditures variable is constructed from 282 sepa-
rate food item variables, which include details on every aspect of food items. Culture
and leisure time are also used as aggregates and comprise the total expenditures of
133 variables.
In Sub-study IV, the dependent variables consist of annual household gross, net,
and financial wealth. The gross wealth variable includes several components: the
value of dwellings and other properties; transportation; forests; the value of arable
land; net worth of business activities; net worth of a group; total deposits; listed
shares; other shares; investment funds; bonds; participation certificates; individual
pension insurance; and savings and investment assets. Household net wealth is cal-
culated by subtracting the household’s liabilities from its gross wealth. Financial
wealth is used as a dependent variable to measure changes in investment rates follow-
ing financial market deregulation. This variable comprises quoted deposits, shares,
other shares, fund shares, pension insurance, and savings. The household wealth
concepts are not top-coded (see OSF, 2015b). As wealth is a highly skewed variable,
the dissertation utilized the recommended inverse hyperbolic sine (I) transformation,
which can incorporate zero and negative values (to account for debt, etc.) (Friedline
et al., 2015; Pence, 2006). The dependent variables are the inverse hyperbolic sine
of all listed wealth concepts, which are equivalized and adjusted for inflation.
For ease of interpretation of the results, the coefficient measures were trans-
formed back into euro values by using the inverse function 12(exp(X) +
exp(X)) (see Pence, 2006).
5.2.2 Independent variables
For independent variables, this dissertation also measured socioeconomic status
(SES) in terms of social class and education, unemployment, family structure, and
income transfers. Different strategies were implemented depending on the accuracy
provided in the dataset.
31
Esa Karonen
In Sub-study I, SES was measured using Erikson–Goldthorpe’s class classifica-
tion (Goldthorpe & McKnight, 2006). While the Goldthorpe / Erikson / Portocarero
class schemes (EGP) traditionally include ten categories, data restrictions in the older
registers from 1970–1980 necessitated harmonization to seven classes: higher ser-
vice (I); lower service (II); higher routine non-manual (IIIa); small proprietors and
employees (IVa); self-employed farmers (IVc); lower technicians (V); and routine
non-manual workers, administration, and sales and services semi-/unskilled manual
workers (IIIb+VIIa+b). In Sub-studies II, III, and IV, SES was measured by the
main type of economic activity, based on Statistics Finland’s Classification of Socio-
economic Groups. This classification includes employees, who are classified into
upper-level and lower-level employees, and manual workers. Self-employed indi-
viduals can be grouped into self-employed without employees, self-employed with
employees, and unpaid family workers.
In Sub-study I, education was measured as a continuous variable, in years. The
indicator referred to the highest acquired degree and the years required to attain that
education level. Finland has a compulsory education system, which sets the min-
imum years of education at 9 years, and the maximum years of education at 21
years. Sub-studies II, III, and IV used a categorical variable, which is the highest
educational level attained by the end of the measurement year. The classification is
based on the The International Standard Classification of Education (ISCED) clas-
sification, and includes basic education, upper secondary education, post-secondary
non-tertiary education, lower tertiary education, higher tertiary education, or higher.
Unemployment in Sub-study I was quantified by the number of months at the end
of the year. Hence, the term ‘unemployed’ encompasses those not employed for var-
ious reasons, such as retirement, studying, or home care. Unfortunately, Sub-study
II did not offer such detailed information; instead, it employed the unemployment
level within each 10-year age group for accuracy, thus providing necessary variation
at the cohort level over time, compared to a mere annual unemployment level.
Family structure was examined in all sub-studies, using the number of children
as an indicator. There were nine categories: no children, one to eight children (each
child reported separately), and a distinct category for families with nine or more
children.
In conclusion, several unique variables were employed in specific sub-studies.
Sub-study I introduced two unique variables not present in Sub-study II. Firstly, in-
come transfers were included to compensate for the absence of disposable income
as an indicator. The income transfer variable comprised social security benefits,
social assistance, and other current transfers received. Secondly, the cohort size
was also controlled, a feat made possible by the register data, to capture structural
changes in the demographic age structure. In addition, analysis was done both for
men and women respectively, while other sub-studies omitted this approach as they
use household-level data based on reference persons without access to the partner’s
32
DATA AND METHODS
information.
Furthermore, Sub-study II included urbanization level to account for varying
consumption preferences and needs in different areas. Three categories were em-
ployed to denote the urbanization level of the household’s municipality: city, densely
populated, and rural.
5.3 Methods
5.3.1 Multilevel approach to between-cohort income inequality
in Sub-study I
The first sub-study of this dissertation employed multilevel modelling to determine
the extent to which between-cohort income inequality accounts for overall income
inequality throughout the lifespan. The choice of methodology was bolstered by the
availability of a high-quality register dataset, tracking all Finnish individuals over
time. Consequently, an age-period-cohort framework was neither required nor ap-
propriate, given the focus was on income variance.
The article utilized three-level random intercept regression models, separately
for men and women. These models clustered observations according to cohorts and
individuals, yielding individual-, time-, and cohort-specific variances. The study
controlled for age in 5-year interval categories across all models, including baseline
models, since birth cohorts encompass individuals with varying age spans. By adjust-
ing models where age groups are accounted for, it made possible to compare income
inequality between individuals who were in the same age groups (5-year intervals)
but belonged to different cohorts. Variance decomposition methods were employed
to identify which specific variables contributed to between-cohort income inequality.
To determine how much between-cohort income inequality explains total income
inequality, the models were designed to estimate variance components in baseline
models that controlled for age in 5-year interval categories. As age is controlled
for, we only observe the average differences in between-cohort income inequality. If
age were not controlled, the model would be biased because age trajectories would
vary by birth cohort. Moreover, the article examined how structural factors con-
tributed to total between-cohort income inequality by decomposing variance from
the baseline models and adding each explanatory variable individually to the mod-
els. Finally, the article estimated variance components for complete models that
included all the explanatory variables. In both instances, variance components and
intraclass correlation coefficients (ICCs) were provided to demonstrate how much
cohort-specific differences in income, both before and after adjustments, account for
the total variance (cohort var. + within individual var. + between individual var.). As
all the models control for age, total income inequality implies differences between
and within individuals of different cohorts, adjusted for age. The ICC serves as a
33
Esa Karonen
measurement for the degree of between-cohort income inequality from total vari-
ance, as between-cohort variance is calculated as a percentage share from the total
variance. The models we estimated can be expressed as follows:
Y = 0 + xcit + ⌧c + µci + ✏cit
Y = 0 + xcit + Zcit + ⌧c + µci + ✏cit
Y denotes income that is the dependent variable, 0 is an overall intercept, and
Xcit is the age of the individuals that are adjusted for in all the models (dummies
for every 5-year interval). Here, cit refers to the levels of the cohort (c), individual (i),
and time (t). Here, time refers within-individual variation, as in the data individuals
are followed through the years. Zcit describes the observed cohort and individual
characteristics for each independent variable; residual ⌧cit reflects differences due
to unobserved cohort-level heterogeneity, which does not vary between individuals
of the same cohort. µci refers to unobserved individual-level heterogeneity, which is
constant for each individual, and ✏cit refers to the residual variance within individuals
due to time (within individual variation).
The random intercept model allowed for comparing cohort- and individual-level
variances with the total variance in the model: 2cohort + 
2
individual + 
2
time. The
amount of variance explained by the cohort with respect to the total variance of in-
come inequality was calculated by ⇢cohort =
⇢2cohort
⇢2cohort+⇢
2
individual+⇢
2
time
, which indicates
the share of between-cohort income inequality. ⇢individual =
⇢2time
⇢2cohort+⇢
2
individual+⇢
2
time
measures how much individual variation in income explains from total variation of
income inequality. When ⇢cohort and ⇢individual are multiplied by 100, the value
can be interpreted as the degree to which cohort and individual variations explain
between-cohort income inequality as percentages of the total variation.
To capture the change over the life course, the study used growth curve multilevel
models in order to investigate how the structural factors explain cohort differences
in income over different stages of life course. To this end, the article estimated the
means and variances of income trajectories by age between cohorts. Similarly, as
in the first part, the baseline model (empty model) was run first, and after this, the
models included each structural factor separately. After this, the study calculated
how much each structural factor changed between-cohort variance in percentages
compared to the empty model. In the growth curve models which are denoted below,
we added an interaction term between cohort and age:
Y = 0 + 1xcit + 2xcit + 1xcit ⇤ 2xcit + ⌧c + ⌧⌧c + µci + ✏cit
Y = 0 + 1xcit + 2xcit + 1xcit ⇤ 2xzcit + zcit + ⌧c + ⌧⌧c + µci + ✏cit
Models used random slope linear regressions with within-subject correlations of
repeated measurements by assuming a correlation structure between them (Twisk,
34
DATA AND METHODS
2004). Y denotes income is the dependent variable, 0 is the intercept (grand mean),
1Xcit is the regression coefficient for the age of the individuals (dummies for ev-
ery 5-year interval), 2Xcit is the regression coefficient for cohort, and Zcit is the
independent variable. In the interaction models, denoted 1Xcit ⇤ 2Zcit, the article
computed age and cohort variables under categorical variables because the results
were presented in easy-to-interpret graphical form. Error terms in the model are de-
noted as follows: ✏cit is for individuals at time t; ⌧c is for cohorts; ⌧⌧c is cohort
within age, and µci is individuals within a cohort.
5.3.2 Age-Period-Cohort methodology to measure income and
consumption in Sub-study II and III
The second and third sub-studies of this dissertation utilized innovative age-period-
cohort modelling techniques. Age-period-cohort models aim to estimate inter-cohort
income trajectories by examining the effects of age, period, and cohort from two
contextual perspectives: the relative and absolute contexts.
However, despite numerous papers suggesting that answers can be found in the
empirical estimations and effects of age (a), period (p), and cohort (c), a significant
problem that arises as the ‘identification problem.’ The variables are collinear, mean-
ing that when two of the variables – age, period, or cohort – are known, the third
is also known. Collinearity between regressors suggests that the statistical model
produces an infinite number of possible solutions for either the least squares or max-
imum likelihood estimators (Yang et al., 2004). Thus, the primary issue is that the
model does not possess a unique solution, and such a solution cannot be identified.
As a result, when the actual underlying process affecting a dependent variable in-
cludes linear effects of some or all of APC, there is a risk of selecting the incorrect
combination, given that we could substitute a term for the combination of the other
two terms without altering the data
Age-period-cohort models are an attempt to grasp the identification problem.
The model aims to explain outcomes through three components: the individual’s
age a (↵a), membership in a cohort c (µc), and statistical period p (⇡p). Thus, the
equation can be stated as follows:
yapc = µ+ ↵a + ⇡p + 
The primary function of the APC model is to determine ‘how’ an outcome is
influenced by the individual’s age (life-cycle stage), time of birth (cohort), and time
(period).
Past research has addressed the identification problem by imposing specific re-
strictions on the model (Hobcraft et al., 1982; Mannheim, 1928; Mason et al., 1973;
O’Brien, 2011; Ryder, 1965; Yang et al., 2004). These restrictions are designed to
35
Esa Karonen
limit the coefficients of certain variables. One such example is the constrained gen-
eralized linear model (CGLIM) estimator, which relies on a theoretical foundation
whereby constraints utilize additional information to limit coefficients based on the-
ory. However, the CGLIM’s dependence on external information is problematic, as
such information often does not exist, and CGLIM is sensitive to the selection of
constraints, as noted by Glenn (1976). This issue is one of the reasons the intrinsic
estimator (IE) model was developed by Yang et al. (2004, 2008).
At its core, the IE employs principal component analysis to reduce the collinear
APC dimensions to a bi-dimensional plane. Despite this, the solution has been criti-
cized by O’Brien (2011) and Luo (2013), who argue that the intrinsic constraint is as
arbitrary as in any other CGLIM. For instance, studies by O’Brien (2011) and Chau-
vel (2013) have demonstrated that the model fails empirical tests such as detecting
educational levels (refer to the discussion in Pelzer et al., 2014).
Age-Period-Cohort Detrended model
The second sub-study employs the age-period-cohort-detrended (APCD) method
(Chauvel, 2011; Chauvel & Schroder, 2015). Simply put, the APCD model calcu-
lates the relative differences between age groups, cohorts, and periods in comparison
to a linear trend. This quality of the model makes it possible to take into account gen-
eral income development over time while controlling for APC components, which
would not be possible in other modeling techniques.
The APCD model acknowledges that due to the identification problem, linear
trends in APC models cannot be firmly ascribed to age, period, and cohort. Instead,
the ‘detrended’ approach concentrates on how the effects of age, period, and cohort
deviate around a linear trend, which this method absorbs. More descriptively, APCD
operates as ‘bump’ detector, demonstrating how various cohorts diverge from the
linear trend. The APCDmodel imposes a set of constraints on zero-sum, zero-slopes,
and the estimation domain of the cohort effects, excluding the first and last cohort.
Instances of omitted observations are the oldest age group from the initial period and
the youngest from the final period that appear once in the model. This technique
enables the model to be identifiable, secure improved estimates, and offer a unique
solution.
In the detrended model, yapc stands for the dependent variable but also for the
independent variables of age (a), period (p) and cohort (c). The APC effects are
the following: period effect ⇡p fits the categorical period, ↵a is the coefficient for
the non-linear age changes, c is the estimates for the cohort effects, 0 denotes the
general intercept, and jxj are coefficients for the control variables. Rescale (a) and
rescale (c) are linear functions that rescale the indexes a and c, which transforms the
coefficients ↵0 and ⇡0 in a standardized form to a scale of -1 to +1. Both rescale (a)
and rescale (b) absorb the linear trend. Finally, if c, as the cohort effect coefficient,
36
DATA AND METHODS
is zero, this means that the cohort does not show any unique cohort-specific behavior
and that it maintains homogenous behavior or effect. The model is stated as follows:
In the detrended model, yapc represents both the dependent and independent vari-
ables, specifically age (a), period (p), and cohort (c). The APC effects are as follows:
the period effect, denoted as ⇡p, fits the categorical period; ↵a serves as the coeffi-
cient for non-linear age changes; c estimates the cohort effects; 0 is the general
intercept, and jxj are coefficients for the control variables. The linear functions
rescale (a) and rescale (c) are used to rescale the indexes a and c, transforming the
coefficients ↵0 and ⇡0 in a standardized form to a scale of 1 to +1. Both rescale (a)
and rescale (b) absorb the linear trend. Finally, if c, the cohort effect coefficient,
is zero, this implies that the cohort does not exhibit any unique, cohort-specific be-
haviour and maintains a homogeneous behaviour or effect. The model is presented
as follows:
8>>>>><>>>>>:
yapc = ↵a + ⇡p + c + ↵0rescale(↵) + ⇡0rescale(p) + 0 +
P
j jxj + ✏i8>><>>:
P
a
↵a =
P
p
⇡p =
P
c
c = 0
Slope↵(↵) = Slopep(⇡p) = Slopec(c) = 0
min(c) < c < max(c)
The APCD model divides trends into two categories: the first category is based
on the linear trend, while the second is based on the non-linear trend on the fluctu-
ations around the linear trend. This model circumvents the traditional identification
dilemma by calculating the constrained trend with zero-sum and zero-slope param-
eters, which are compared against a unique decomposition of a, p, and c. These
contain the estimated fluctuations (Chauvel, 2011). In essence, when at least one
coefficient differs from the zero-slope coefficient, the APCD model will display the
variations. When selecting the appropriate model between AP and APC, it is sensible
to compare (Raftery, 1986) Bayesian information criteria values between the models
to determine which model has superior explanatory power.
When disposable income is used as the dependent variable, there are several
reasons to analyse deviations from the linear trends. APCD distinguishes the rel-
ative share of period variations between cohorts but does not account for the non-
conformity of changing living standards in an absolute sense. Therefore, the age-
period-cohort trended (APCT) model is developed as a tool to measure absolute de-
clines and progressions. The main modification to the APCD model is to eliminate
the zero-slope constraint from the cohort coefficients and thus suppress the 0 rescale
(c)-term.
The period fluctuations are controlled in the same way in both APCD and APCT;
however, rather than absorbing the long-term linear trend, the parameter 0 acts as
a trended cohort effect in the APCT model. This trended effect denotes per-cohort
37
Esa Karonen
change while being controlled. Thus, the APCT will estimate the progression that
controls for period fluctuations at a given age without being affected by the long-
run period trends. Hence, the APCT will illustrate how inflation-adjusted disposable
income changes between cohorts. The APCT formula is defined as follows:8>>>>><>>>>>:
yapc = ↵a + ⇡p + c + ↵0rescale(↵) + 0 +
P
j jxj + ✏i8>><>>:
P
a
↵a =
P
p
⇡p =
P
c
c = 0
Slope↵(↵) = Slopep(⇡p) = 0
min(c) < c < max(c)
Age-Period-Cohort modeling for between-cohort consumption: Blinder-
Oaxaca method for subgroups
The third sub-study focused on expenditure differences between two income
groups, which called for a special variation of the APC model to make this com-
parison possible (for a discussion of this methodology, see Yang et al. 2004). This
is requirement poised by the research question, as the sub study compared high- and
low-income groups between cohorts.
The APCGO model (Chauvel et al., 2017) analyses the birth cohort–based dif-
ferences (‘gaps’) between two groups, decomposing the differences into a part ex-
plained by control variables as well as an unexplained part. It measures a possible
closing gap from social generation to generation.
The process is twofold. First, with the base of the Oaxaca-Blinder models of rel-
evant control variables in each (age by period) cell of the initial Lexis table apc, the
model computes a matrix uapc of ‘unexplained’ differences and the ‘Oaxaca-Lexis
table’ of income and expenditure gaps between the groups. Second, the Oaxaca-
Lexis table is decomposed on the basis of a specific trended APC model to obtain a
measure of the cohort-specific non-explained gap in income (Bar-Haim et al. 2023).
In the first step, the model applies the Blinder-Oaxaca decomposition method
(Blinder, 1973; Jann, 2008; Oaxaca 1973; Oaxaca & Ransom, 1994) to each cell of
the initial Lexis table to obtain the income quintile gaps in household expenditures
(HSE) (un)explained by independent variables. We consider incomes for the first
group G1 and the second group G2, a linear combination of endowments, and a sum
of errors.
log(HSE)
G1
c = X
G1
c b
G1
c + ✏1
log(HSE)
G2
c = X
G2
c b
G2
c + ✏2
38
DATA AND METHODS
In Equation 1, XG1c ) represents the mean of independent variable X at cohort
C for the first quintile; likewise, b in bG1c represents the coefficient for the same
independent variable and quintile cohort groups. Similarly, in the second equation,
the same definitions apply except for the fifth quintile. When we subtract Equations 1
and 2, we express the differences in expenditures to income quintiles for each cohort:
log(HSE)
G1
c  log(HSE)
G2
c = b
G1
c (X
G1
c XG2c ) +XG2c (bG1c  bG2c )
In Equation 3, the subtraction of HSE terms is the overall expenditure gap in
cohort C between income groups, and bG1c (X
G1
c  XG2c ) is the gap explained by
independent variableX in a cohort C. The termXG2c (bG1c  bG2c ) is the unexplained
variation, which contains the effect not observed in the model.
In the twofold decomposition, the mean outcome difference is the difference in
the linear prediction at the group-specific means of the regressors of the difference,
which can, in the case of the two groups, be decomposed. The model applies a
specific trended APC model to the Oaxaca-Lexis table to obtain the trend measure
of the cohort-specific expenditure gap, the APCT-lag coefficient. The new APC-lag
approach uses the ‘linear age effect’ as its baseline (Bar-Haim et al., 2023, 2018).
Once this constraint is given and the period linear trend is constrained to zero, the
cohort effect will absorb the long-term time transformations. This definition means
a new, clear baseline, at which the linear slope of age trend measured by the ↵a
coefficients is designed to equal, the average shift due to age in the Oaxaca-Lexis
table across cohorts Oapc. Consider this average shift ↵:
↵ =
X (O(↵+ 1, p+ 1, c)Oapc)
(A 1)(P  1)
where ↵ represents the average shift for a cohort c when it grows one age group
older in the next period across the window of observation of a age groups and p
periods. Once this is known, APC-lag is identifiable:
oapc = ↵a + ⇡p + c + ✏
whereas the full model is denoted as8><>:
yapc = ↵a + ⇡p + c + 0 +
P
j jxj + ✏i(P
(↵a) =
P
(⇡p) = 0
Trend(⇡p) = 0;Trend(↵a) = 0
The formula of operator trend for age coefficients, when A is the number of age
coefficients, is
39
Esa Karonen
Trend(↵a) = 12
P
(↵a(2↵A 1))
(A 1)A(A+ 1)
In APC-lag, c absorbs the constant (larger when the gap is high), its trend shows
the variation in the intensity of the gap by cohort for age and period controlled, and
the fluctuations show possible non-linear accelerations or decelerations in the cohort
trend.
It should be noted that the complete APCGO method cannot provide direct es-
timations of confidence intervals due to the complexity of the succession of the
Blinder-Oaxaca and APC methods. Therefore, we bootstrap the entire process con-
sidered, including the Oaxaca-Blinder decomposition of each cell of the initial Lexis
table of apc to obtain the non-explained Oapc Oaxaca-Lexis table. For a more com-
prehensive methodological discussion, see the relevant research literature (Bar-Haim
et al., 2023; Chauvel et al. 2017).
5.3.3 A linear model to measure wealth differences between co-
horts in Sub-study IV
In the fourth sub-study, the dissertation utilized linear regression models. The wealth
data weren’t as harmonized as the other datasets, potentially leading to reliability
issues with APC modelling. Therefore, more traditional models were selected to
capture cohort differences over the life course.
The analysis is confined to comparing differences in wealth association between
age groups during the observation period. For this purpose, regressions with inter-
action terms of the cohort dummies and wealth components are performed for the
whole sample.
The sub-study use the marginal effects from the ordinal least squared. The depen-
dent variable in the model is inverse hyperbolic sine transformed household wealth of
gross wealth, net wealth, financial wealth, and all the decomposed variables of finan-
cial wealth (stock assets, exchange-traded funds and other funds, other stocks, and
bonds). In the first set of models, we use the interaction term of cohorts (Xcohort),
and in the second set of models cohort and age groups (Xage). The simplified models
for each wealth component are denoted as follows:
ashin(wealth)i = 0 + 1XcohortXage + ✏i
In addition, as a robustness check, the models were ran as as multilevel models
with statistical years as a level, but the observed change on intraclass correlation
was extremely low. This test result validated the use of marginal effects instead of a
multilevel model. Results of the multilevel models did not deviate from the marginal
effect results.
40
6 RESULTS
6.1 Summary of main findings
Table 3 shows the main results of the dissertation, which are ordered by the main
research questions posed in Table 2. The logic is the holistic context of economic
resources, as income is the prerequisite for consumption and wealth.
Table 3. Summary of main results for main research aims.
How has generational income inequality has changed, and how have the different structural
and socioeconomic mechanisms contributed to income differences?
• Period and cohort changes were identified as the most important driver of income, while age
was not after adjusting for the main economic activity
• Income inequality was explained by cohort differences by 6% for men and 26% for women
• For both men and women, cohort size and education were the most important factors associated
with income inequality. For men, especially unemployment was a major factor
• Income has increased each subsequent cohort: Baby Boomers as ‘winners’; youngest generation
as ‘losers’
• 1980s cohort’s income development has stagnated: average income has not increased from
the 1970s cohort’s income, probably affected by the financial crisis
How has generational consumption inequality changed within income distribution?
• Incomes have maintained a stagnant gap between high- and low-income groups over cohorts;
however, expenditures have risen slightly more in favour of the high-income group
• Expenditures in food and groceries have become more similar between high- and low-income groups,
with low-income group becoming more ‘necessity rich’
• The high-income group have been gradually investing in leisure time over cohorts,
while the low-income group remains ‘leisure poor’
How has generational wealth distribution changed, and how does the opening up of financial
markets relate to changes in wealth between generations?
• Overall, wealth has increased each subsequent cohort, while wealth accumulation starts to
accelerate at increasingly younger age
• Regardless of financial deregulation, deposits and traditional saving methods were still ‘king’
• The 1980s cohort displays a slower rate of wealth accumulation than the preceding cohort, while
having similar wealth profile and level as the 1970s cohort
• The 1980s cohort bears a higher debt level coupled with slower loan amortization and has delayed
the purchase of dwellings
41
Esa Karonen
Overall, the most crucial findings indicate that the cohort from the 1980s is in
a vulnerable position, seeming to face more unfavourable circumstances concerning
income, consumption, and wealth. While the war generation also faced adverse sit-
uations regarding income and wealth, these outcomes were anticipated due to the
crisis, underdeveloped social institutions, and the unfortunate reality that the time to
alleviate these issues had already passed from a life course perspective.
Considering the overall progression of income development, the relative income
for the 1980s cohort is lower compared to that of other cohorts. This observation is
corroborated by a more detailed analysis of individual income, which revealed that
the younger cohort is grappling with income stagnation. Here, the income trajectory
over the life course mirrors that of the 1970s cohort.
From a consumption standpoint, younger generations exhibit the largest expen-
diture gap in leisure time between low- and high-income groups. This discrepancy
significantly impacts societal participation beyond merely ‘making ends meet.’
Additionally, the 1980s cohort displays a slower rate of wealth accumulation
than the preceding cohort. They also bear a higher debt load coupled with slower
loan amortization, which has delayed the purchase of dwellings and pushed it to
a later age. Consequently, the prolonged loan amortization could have long-term
implications for total wealth accumulation, leaving future retirees to live on a tighter
budget.
In summary, the younger generation often find themselves playing ‘catch-up’
games with older generations. From the holistic perspective, the dissertation offers
a more comprehensive picture of how income, consumption, and wealth connect to
each other. The case from the 1980s illustrates the causal relationship of the three
economic resources, demonstrating how the stagnant growth of income contributed
to the deceleration of wealth accumulation. Concurrently, it altered the budget con-
straints for consumption. Moreover, the findings of the thesis highlight the impor-
tance of the economic life course approach presented in Chapter 3 (Elder, 1994;
Giele et al., 1998). It revealed that individual agency, manifested through decisions
related to education and professional fields, plays a significant role. Alongside this,
historical periods were characterized by both increased opportunities in human cap-
ital and economic recession, the latter resulting in unemployment spells or delayed
careers (see also Kurz & Blossfeld, 2004; Lovenheim & Reynolds, 2013; Oreopou-
los et al., 2012). Significant contributors to inequality were identified as adverse
events, as they lie outside the control of individuals, and, consequently, their respec-
tive cohorts. This observation is emphasized by the experiences of the cohort from
the 1980s, reinforcing earlier discoveries (Chetty et al., 2017; Riihela, 2006).
The sections that follow present the main findings in greater detail, along with
the research questions posed by the sub-studies.
42
RESULTS
6.2 Income inequality between cohorts (Sub-study I &
II)
The dissertation’s first research question was on how generational income inequality
has changed, and how the different structural and socioeconomic mechanisms con-
tributed to income differences. The first and second articles connect, as they both
analysed income distribution between cohorts. The first study focused on determin-
ing both how much generational income inequality explains total income inequality
during 1970–2018 for men and women in Finland and how much different structural
factors contribute to generational income inequality. The second study aimed to an-
swer how income distribution has developed across periodic economic fluctuations
in relation to cohort and age group.
The first sub-study scrutinizes the variations in income inequality over the life
course across Finnish birth cohorts born between 1920 and 1989. It also explores
how structural changes related to education and the labour market contribute to these
differences. In simpler terms, the central question of the article is to determine the
magnitude of cohort differences and identify the ‘winning’ and ‘losing’ birth cohorts.
Our findings reveal that cohort differences account for 26% and 6% of total in-
come inequality in women and men, respectively. On the whole, for both genders,
each successive cohort until the 1980s exhibits a higher average income trajectory.
However, for women, the income trajectory differences between cohorts are signifi-
cantly larger compared to men.
Our results corroborate previous findings that suggest the younger generation is
being left behind (see Chetty, 2017; Manduca et al., 2020). The study discovers
that for both genders, each successive cohort until the 1980s has a higher average
income trajectory. However, the income development of the 1980s generation is
stagnating, as their average income has not increased compared to the 1970s cohort.
Regarding the second research question, the sub-study addresses which structural
factors account for cohort income inequality. Education, social class, and cohort size
emerge as the most significant factors for both men and women. Education accounts
for most of the variation between cohorts, suggesting that the accumulation of human
capital is crucial for income development among cohorts. Cohort size determines
the number of applicants in the labour queue, and the findings show that cohort size
explains a large portion of the varying income trajectories between cohorts. From
an individual perspective, a large birth cohort size may be disadvantageous, as it
intensifies labour market competition, thereby limiting opportunities to secure higher
labour market positions (Mugiyama& Toyonaga, 2022). The economic situation also
plays a significant role: unemployment is a crucial factor for both genders; however,
43
Esa Karonen
it explains almost twice as much for men as it does for women.
The second sub-study evaluates how economic shocks influence income trends
and inter-cohort income dynamics by using APC modelling. The findings suggest
that period and cohort effects primarily impact relative income within the context
of APC methodology, while age effects have no substantial influence when control
variables are considered. This result underscores the relationship between the ef-
fects of economic shocks and cohort placement at labour market entry, whereas the
occupational position achieved explains for all age variations.
The results indicate that those born in 1970, who entered the labour market on the
cusp of the economic shock of 1993 (at the age of 23 earliest), exhibit no income de-
velopment in relation to the income trend. Every cohort born after 1975 experiences
a negative relative income effect. A similar effect is discovered in the first sub-study,
where the 1980s cohort demonstrates a stagnated income trend, mirroring the in-
come trajectory of the 1970s cohort. Moreover, the analysis reveals that economic
shocks create stagnation points in income development. In the context of cohort in-
comes, these stagnation points are especially detrimental to cohorts transitioning into
the labour markets. However, while the older cohorts didn’t experience such effects
during the oil crisis, younger generations faced birth-positional disadvantages during
both the economic recession of the 1990s and 2007/2008.
Overall, the APC framework and the analysis conducted using register data com-
plement each other well. To reiterate, APCmodelling reveals that both cohort and pe-
riod effects are significant in income, while the multilevel approach with the registers
provides an opportunity to delve deeper into identifying the ‘winners’ and ‘losers’
within generational income development.
6.3 Consumption differences between top and bottom
income deciles (Sub study III)
The dissertation’s second research question asks how generational consumption in-
equality has changed within income distribution. The third sub-study investigates the
differences in consumption between various cohorts. First, the article explores the
way inter-cohort income and spending profiles in the high- and low-income deciles
have evolved over the years 1966–2016. Subsequently, it probes the degree to which
inter-cohort consumption profiles between high- and low-income households vary in
certain categories of consumption. The analysis uncovers an inequality between the
two income groups. Although income and consumption remain static across different
cohorts, expenditures on necessities (such as food and groceries) and income-elastic
goods (such as culture and leisure activities) indicate shifts in consumption inequal-
ity.
First, the relative expenditure on food and groceries between high-income and
low-income groups appears to exhibit more uniform spending patterns over time, as
44
RESULTS
indicated by a distinct reduction in the spending disparity between these two groups
on necessities. The findings indicate that long-term income increases are equaliz-
ing between-cohort expenditures on necessities, irrespective of their socioeconomic
status.
However, expenditure on culture and leisure activities between high- and low-
income groups reveals that the gap is expanding over cohorts to the advantage of
the high-income group. This suggests that the high-income group possesses more
resources for consumption, and that this group tends to allocate surplus resources
towards more income-elastic goods, such as cultural activities and leisure time. It
appears that a surplus of resources affords more freedom to invest in leisure time,
while the low-income group remains ‘leisure poor.’ Consequently, this dissertation
finds that expenditures on goods related to leisure time are increasing – but predom-
inantly in favour of households with superior economic resources. This implies that
households from older to younger generations are investing more in leisure time, but
increasingly, only those that are economically affluent do so.
6.4 Wealth accumulation between cohorts and the fi-
nancial deregulation (Sub study IV)
The dissertation’s final research question is on how generational wealth distribution
has changed and how the opening up of financial markets relates to changes in wealth
between generations. Thus, the fourth sub-study examines wealth disparities among
different cohorts. The primary objective is to evaluate how wealth distribution across
generations is shifting and how the deregulation of financial markets correlates with
these changes. Additionally, the study explores the evolution of wealth dynamics
among cohorts throughout their lives, particularly following significant institutional
changes such as the advent of open financial markets. To provide a comprehensive
view of wealth fluctuations, the study measures gross, net, and financial wealth.
The sub-study assesses the scale of changes in absolute terms and the composi-
tion and significance of financial wealth in relative terms as a portion of total wealth
throughout the life course. The results indicate a relatively stable pattern of wealth
accumulation across cohorts, except for the war generation. The findings also reveal
that each cohort’s wealth is not only higher overall but also begins its upward trend
at an increasingly younger age with each successive cohort. Consequently, the accu-
mulation of wealth starts at an earlier stage in life. However, the wealth trends of the
1930s and 1980s cohorts contradict this pattern, with the oldest generation showing
a stagnant trend and the youngest exhibiting a less steep slope than the preceding
cohort.
For the younger cohort, the less steep wealth trajectory could be linked to higher
debt levels (and more costly homes) and slower loan repayment. The study also
discovers that early life debt related to housing significantly impacts wealth, with the
45
Esa Karonen
debt decreasing as retirement approaches. Younger cohorts carry more housing debt,
and they tend to purchase homes later in life compared to older cohorts. Moreover,
loan repayment is slower for each subsequent cohort, indicating that the stage of life
free from debt is delayed compared to earlier cohorts.
Considering the historical context, the deregulation of financial markets, and the
structure of financial wealth, previous study has found that younger cohorts may
be more receptive to financial instruments, despite their limited resources (Lim &
Zeng, 2016). In contrast, some research suggests that only older generations show
a greater interest in financial assets (Greene, 2014). The dissertation’s study yields
mixed results. It shows variations in financial wealth among younger cohorts, with
them starting to invest at younger ages compared to other cohorts. However, the
significance of financial wealth in relation to total net wealth remains virtually the
same across all cohorts.
46
7 DISCUSSION
This dissertation studied generational economic inequality throughout the life course,
considering factors such as age, period, and cohort and using a unified and holistic
approach. This approach incorporates income, consumption, and wealth. Theoret-
ically, the dissertation enhances the research field by acknowledging that economic
resources fluctuate as individuals progress through their life course. Likewise, indi-
viduals are born at different times, which determines the opportunities available for
improving economic outcomes. Additionally, a significant contribution is the holis-
tic consideration of economic resources, which incorporates all economic measures:
income, consumption, and wealth.
Previous research has expressed concerns about the income development of
younger generations, both internationally and within the Finnish context (see Chetty,
2017; Hammer, 2022; Riihela, 2006). This dissertation’s findings resonate with these
concerns, as the 1980s generation exhibits clear income stagnation, an increasing
consumption gap between income groups, and lower wealth accumulation coupled
with higher debt levels. However, previous research has not examined the details of
this phenomenon.
This dissertation provides new insights into the vulnerability of younger cohorts
and unveils how different cohort dynamics influence the life course. The multidimen-
sional age-period-cohort approach successfully unravels many previously unknown
issues by utilizing innovative methods and considering a holistic approach to eco-
nomic measures. It emphasizes the importance of understanding the interplay of
income, consumption, and wealth, as changes in one component could affect the
others. In this chapter, I first review the substantive and theoretical contributions of
the dissertation’s research articles then discuss potential policy implications, assess
methodological considerations, and inally conclusions, which also suggest recom-
mendations for future research.
7.1 Interpretation of the results
This dissertation reflects the theoretical framework of the life course paradigm, as
introduced and discussed in the third chapter. It illustrates how income, consump-
tion, and wealth provide a complex picture when examining the long-term economic
development of various cohorts over the life course. A key aspect of the theory is
47
Esa Karonen
the intersection of cohort and time, influencing the timing of opportunities and the
economic environment availabIe. The results indicate that the timing of the cohort is
significant, as several historical periods were relevant for the cohorts.
From the perspective of Finnish economic history, the period 1945–1980 was a
time of substantial economic growth. The cohorts of 1930–1950 entered the labour
market during this period. High employment rates and increased educational oppor-
tunities created an optimal platform for social mobility, positively influencing the
ability to secure lucrative positions in the labour market. These findings support pre-
vious results indicating that the so-called ‘Baby Boomer’ generation had a relative
economic advantage in terms of income returns, while also providing more insight
into the impact of this change on overall income distribution.
In comparison to previous research, this dissertation found that cohort differences
can explain income inequality. For women, the figure stands at 26.1%, and for men,
it is 9.5%. This highlights the change in educational opportunities for women, pro-
viding a solid foundation for income and social mobility. The increased educational
level of women also led to significant changes in occupational structures, altering
women’s participation in the labour market and resulting in higher between-cohort
variation (Breen et al., 2009, 2010). These findings support previous results indicat-
ing that the so-called ‘Baby Boomer’ generation had a relative economic advantage
in terms of income returns, while also providing more insight into the impact of this
change on overall income distribution.
In contrast, younger generations have faced the economic recession of the early
1990s and the financial crisis of 2007/08. The results from both sub-studies on in-
come indicate that the 1980s cohort, in particular, does not increase their income
level compared to the previous 1970s cohort. Detailed analysis for both men and
women suggests that the 1980s cohort has a more unstable income trajectory and
even falls below the income trajectory of the previous cohort. Furthermore, more de-
tailed cohort results mirror the impact of educational expansion for women, leading
to greater income growth. However, certain life events from ages 20 to 40 are associ-
ated to women’s income. Occupational segregation plays a role, as men and women
work in different fields, with traditionally male-dominated industries offering higher
wages. In addition, women are more likely to take breaks from the workforce for
caregiving, leading to lower lifetime earnings and reduced retirement savings.
However, unlike previous studies (Dutt & Padmanbhan, 2011; Mckenzie et al.,
2011; Urbonavicˇius & Pikturniene˙, 2010), this thesis does not find an association
between economic crises and the expenditures of cohorts, as both income and con-
sumption gap trends remain unaffected. A similar result was found regarding wealth,
as economic recessions did not seem to have a long-term direct impact on wealth
development. However, the result is dependent on the measurement period, as a
slight change in 1-year wealth estimations is observable during the 1990s recession,
yet during the 5-year measurement period it is not, indicating that it is not a long-
48
DISCUSSION
term change. In addition to period effects, the question of whether the dual-taxation
reform would impact wealth accumulation for those already possessing substantial
financial investments, or those demographic groups with a greater amount of dis-
posable wealth for investments, remains unanswered. In both cases, the disserta-
tion reveals no direct evidence, as the financial wealth shares appear to be similar
or marginally higher than the total wealth of preceding cohorts. In the case of con-
sumption, it is to be noted that the APCmethodology period was arranged into 5-year
groups, which could increase insensitivity to more rapid changes, for example during
the financial crisis of 2007/2008.
To contextualize these historical periods from the life course perspective, it may
very well be that the 1970s cohort who were 14 to 23 years old during the 1990s
recession were still mostly in education, where they could fend off the crisis and
transition to the labour market later. On the other side, the 1980s cohort were at the
youngest 18 to 28 years old during the financial crisis. The average age for school-
to-work transition for tertiary education is 27 years in Finland. It is probable that
the cause for the income stagnation is the belated beginning of a working career, as
findings support previous results showing that credential inflation contributes to in-
come inequalities between cohorts (Bol, 2015; Van deWerfhorst & Andersen, 2005).
It is to be noted that the results indicated lower income trajectories among younger
cohorts compared to older ones when education is adjusted for. As educational ex-
pansion has focused mostly on higher education levels, it is reasonable to assume that
returns on higher degrees have diminished for younger generations. Here, increased
individual agency does not necessarily include a better social position, as the internal
competition within cohorts also matters.
To continue the story of the 1980s cohort, the dissertation found that although
there has been a general increase in wealth over time, the disparity between differ-
ent cohorts has also grown. Specifically, evidence shows that each cohort not only
possesses more wealth overall but also starts accumulating wealth at an increasingly
younger age compared to the previous one. This implies that the process of wealth
accumulation commences at an earlier stage of life with each subsequent cohort.
However, an exception was observed in the wealth progression of the cohort from
the 1980s, which displayed a contrasting trend.
The 1980s cohort accumulated less wealth than their predecessors. For the
younger cohort, the smaller slope of the wealth trajectory could be linked to higher
debt, more expensive housing, and slower loan amortization. Although younger co-
horts might have a greater potential to amass more financial wealth over their lifes-
pan, our study indicates that they have a slightly higher or similar propensity to
acquire financial assets. Thus, these younger cohorts also possess the potential for
substantial financial wealth accumulation over their lifetime. This is contrary to pre-
vious research findings, which have suggested that individuals with fewer resources,
like the younger cohorts, would be more risk-averse (Attanasio et al., 1999; Cagetti,
49
Esa Karonen
2003; Lusardi et al., 2017; Malmendier & Nagel, 2011).
The narrative becomes more complex when examining differences in consump-
tion. This thesis has uncovered several new insights into consumer behaviour.
The dissertation revealed that nondurable goods, especially those related to
leisure time, accounted for a significant part of consumption, while the proportion
of spending on essentials like food was decreasing. More specifically, the spend-
ing gap between older and younger cohorts has significantly widened, demonstrating
an increased preference for cultural and recreational activities across cohorts. This
trend is linked to overall income growth, where income is being directed more to-
wards income-elastic goods rather than necessities.
However, in the APC analysis, this research was able to distinguish between two
income groups, uncovering new evidence not previously identified (see e.g. Segall,
2013) that while spending on leisure-time goods has increased, it mostly benefits
households with superior economic resources. Thus, the findings affirm previous
results suggesting that households from older to younger cohorts are increasingly
investing more in leisure, but as new evidence reveals, this trend is increasingly
exclusive to those who are financially well-off. This observation adds to previous
research (Fan & Lewis, 1999; Semyonov et al., 1996), suggesting a disparity in con-
sumption opportunities among different socio-demographic groups. In conclusion,
from a social inclusion perspective, Finnish society appears to be equal in terms
of access to necessary resources, but leisure options seem to divide high- and low-
income earners. In other words, this dissertation identified more precisely the role of
consumption in the stratification process, which theoretical literature has suggested
(Douglas & Isherwood, 1996; Katz-Gerro & Talmud, 2005; Slater, 1999)
To summarize these findings under the economic life course paradigm introduced
in Chapter 3, the main focus seems to be on the economic vulnerability of the 1980s
cohort. A clear narrative emerges when taking a holistic approach to economic re-
sources. The dissertation observed that the mean income and wealth trajectories of
each successive cohort up until the 1980s had a higher average income and wealth
trajectory. This is partly due to timing and coincidence, with the 1980s cohort transi-
tioning from school-to-work during the 2007/2008 financial crisis. Coupled with the
increased market competition due to credential inflation, the results support previous
findings that income development for the younger generation has stagnated (Chetty,
2017; Manduca et al., 2020).
This has impacted both consumption behaviour and wealth accumulation. The
dissertation found that debt on dwellings is higher among younger cohorts, especially
the 1980s group, while the acquisition of dwellings occurs at a later age compared
to older cohorts. This is likely due to stagnant income and rising housing market
prices. In addition, loan amortization is slower in each successive cohort, which
means that the debt-free life stage has been delayed compared to other cohorts. To
add insult to injury, the 1980s cohort’s loan amortization has stagnated, resulting in
50
DISCUSSION
lower net wealth and hindered net wealth accumulation compared to older cohorts.
Furthermore, income stagnation has also led to decreased spending on leisure time,
which could potentially be interpreted as lower societal inclusion and attachment
for this generation. The results overall highlight the importance of a holistic view of
economic resources, as income is the ‘prime mover’ of both consumption and wealth
accumulation. However, all components contain significant information that could
be overlooked if not considered collectively.
In conclusion, all the above results support the research literature’s concept of
‘a scarring effect’ from the life course perspective, as younger cohorts have lower
income attainment than previous ones (Chauvel & Schroder, 2015; Mannheim, 1928;
Mayer, 2009). In this case, the 1980s cohort has the ‘privilege’ of occupying an
awkward economic position. However, there is still time to consider potential policy
solutions or at the very least ensure that the following generation does not inherit
this outcome. Here, the multidimensional approach is crucial; without it, this thesis
would only present a partial picture of economic inequality, much like many previous
studies.
7.2 Policy implications and recommendations
The concept of fairness often arises when discussing policy issues across genera-
tions (Dworkin, 1981; Rawls, 1971). This fairness, or equity, refers to the necessary
conditions for individuals to lead full, healthy lives. Equality, on the other hand, in-
volves providing individuals with the same resources they need to achieve such lives
(Bronfenbrenner, 1973). These notions are interconnected from a generational pol-
icy perspective, as different periods present varying needs and desired outcomes for
policy.
In terms of social sciences and economics, fairness, or justice, is primarily asso-
ciated with taxation and welfare policy (Summers & Smith, 2014). It also applies to
individuals or generations at different life stages (Cannon & Kendig, 2018; Luscher
et al., 2016). Thus, in a specific time frame, a person or a generation could ben-
efit from distribution, while over a life course, they might contribute to it, or vice
versa. Implementing fairness at a practical level, and reflecting the findings of this
dissertation, policy recommendations ideally consider equity in four distinct norms
suggested by Rydell (2005). These norms ensure just treatment and allocation of so-
cial expenditures across successive cohorts, equitable sharing of the welfare state’s
costs, fair returns on contributions made throughout the life cycle, and the right bal-
ance of transfers to maximize opportunities and prospects, which Piachaud et al.
(2009) identify as the primary focus of generational equity.
In the following sections, I will explore the issue of generational policy from two
angles. Firstly, I will delve into broader aspects, such as labour markets and pen-
sion systems. Secondly, I will focus on the significance of timing and birth cohort
51
Esa Karonen
coincidence. From a generational view, both labour markets and pension systems
are heavily influenced by an ageing population, which results from increased life
expectancy and declining fertility rates. This demographic shift has dramatically al-
tered economic relations between generations, thereby driving generational fairness
or unfairness.
This significant demographic shift has raised genuine concerns about its fiscal
implications and its impact on labour markets and the overall economy (Engbom,
2019). Although greater labour force participation by older individuals correlates
with increased youth employment and reduced youth unemployment (van Dalen &
Henkens, 2002; Gruber & Wise, 2010), the immediate effect of retirement is a po-
tential labour shortage. For instance, as Baby Boomers, who make up a significant
portion of the workforce, retire, they could leave many positions vacant. This could
increase competition for skilled workers, potentially inflating wages and benefits
(Roberts, 2012). One argument is that younger generations lack the experience of
the retiring generation, leading to a loss of professional knowledge unless efficient
skill transfer processes are in place. However, past research suggests this concern
may be exaggerated (Freeman, 2006; Neumark et al., 2013). On the bright side,
younger workers, particularly those born in the 1980s and beyond, may advance
more quickly due to changes in workforce dynamics. However, as Baby Boomers
retire and transition from being net contributors to net recipients of benefits, there
could be increased pressure on public finances.
Pensions play a vital role in the fairness discussion, as they represent a transfer
from the working generation to retirees (Go´ra, 2008). Moreover, the ageing and
declining population in Finland poses an additional challenge from a life course
perspective. Regardless of the pension system model, pension benefits are always
financed through the division of current GDP (Chybalski & Marcinkiewicz, 2018).
Future GDP is crucial in a pension system, as the consumption of future pensioners
will depend on future production, primarily by their children’s generation (Barr &
Diamond, 2006). In an era of an ageing population, ensuring pension benefits is a
significant challenge. The question of restructuring the intergenerational contract as
part of the welfare state may become politically relevant.
Timing and coincidence are extremely important. To illustrate, the cohort from
the 1980s suffered from the financial crisis in a different way than, for example,
the cohort born in the early 1970s, whose transition to the labour market coincided
with the recession of the 1990s, which was much worse in Finland than the financial
crisis. What distinguishes these last two recessions is that the late 1990s and the
early 2000s were a time of strong economic growth, so for those born in the early
1970s, the ICT boom and economic growth driven by Nokia acted as an upswing,
i.e. a return to the mean. On the other hand, for those born in the 1980s, the same
economic growth was not found after the financial crisis; instead, they were offered
a state of permanent austerity and economic stagnation. For comparison, a similar
52
DISCUSSION
upswing was also experienced by the Baby Boomers, which lasted several decades
and was not even slowed by the oil crisis during 1973.
The unfortunate timing of the 1980s cohort with the economic recession earned
them the label of ‘the most vulnerable generation.’ They were hit by the economic
recession at a critical age for the school-to-work transition, resulting in a stagnated
income trajectory, which in turn led to lower wealth accumulation. The connection
between the timing of school-to-work transitions and economic shocks is signifi-
cant, leading to persistent earnings declines lasting 8 to 10 years (Oreopoulos et al.,
2012, 2006). Previous research has also found that the effects of recession shocks
are strongest for young workers, while those with a few years of labour market ex-
perience are less affected (Oreopoulos et al., 2012; von Wachter, 2020). Recessions
initially lead workers to start working at less attractive employers, which can hin-
der career development, leaving some young people playing catch-up with earnings,
especially during the first years after the shock. However, there is evidence that if
social institutions implement correct transition systems to support young people dur-
ing the school-to-work transition, it is associated with lower national levels of youth
unemployment (Schoon & Bynner, 2019).
The 1980s cohort’s stagnated income and slower career start could also lead to
lower wealth accumulation, creating several potential issues. The dissertation found
these to include a higher debt load combined with slower loan amortization, which
has delayed property purchases and pushed them to a later age. As acquiring prop-
erty is, on average, the most significant economic investment that people make during
their life course (Bernstein & Koudijs, 2020; Davies et al., 2008; Wolff, 2017), there
could be considerations for optimizing the system to lessen the impact of loan amor-
tization. This is seen as crucial, as studies have found that within-cohort education,
class, and income differentials grew when young adults from lower socioeconomic
backgrounds became relatively less likely to enter homeownership (Kurz & Bloss-
feld, 2004). The dissertation showed that the youngest generation carries a higher
debt burden due to the rising housing market prices and the fact that individuals ac-
quire property at a later age than before (Arundel & Ronald, 2021). Thus, it would
be logical to either consider options to incentivize property purchases at an earlier
age or to reconsider the loan system as a whole. However, during later life, studies
have shown that the impact of later loan amortization is not significant in general,
but it could pose problems for those with irregular work histories and lower pensions
(Mastrogiacomo, 2016).
Several options could have the desired policy effect. These include removing
the transfer tax to lessen the impact on property acquisition, which could also sup-
port workforce mobility. Another solution could be to change the loan payment
period. Finland has a ‘loan payment culture’ where individuals are expected to re-
pay their loans within a set period, typically 15 to 25 years. In contrast, Sweden has
a loan payment period of over 100 years, which has partly resulted in the average
53
Esa Karonen
Swedish individual having twice the wealth of their Finnish counterparts (Credit Su-
isse, 2023). Therefore, it could be valuable to replace our loan payment system with
a more dynamic model to relieve the increased loans required to acquire property,
especially in growth centres. Here, decreasing the pressure of monthly loan amor-
tization would adjust the risk of property acquisition at an earlier age and stimulate
wealth-building in the long run.
A more traditional approach is changes in taxes. From the purchasing power
viewpoint, the most general change would be to adjust the value-added tax (VAT). It
is known that prices respond twice as much to VAT increases as to VAT decreases,
resulting in higher markups (Benzarti et al., 2020). Furthermore, a one-percentage-
point increase in the VAT rate leads to about a one per cent reduction in the level
of per capita aggregate consumption (Alm & El-Ganainy, 2013). As Finland aims
to increase the VAT rate to 25.5 per cent (a 1.5 percentage unit increase), this could
have a long-term impact on overall purchasing power. The dilemma here is the form
of taxation, as taxing consumption rather than income generates more savings for the
individual; however, it also decreases the level of consumption (Alm & El-Ganainy,
2013).
To summarize the above policy discussion, if policy measures are to be imple-
mented, they should be directed optimally to the sensitive part of the life course
during the school-to-work transition to aid individuals’ long-term economic growth.
This would also account for the stunted pension accumulation due to income stag-
nation. Any policy measure should aim to maximize impact over the life course
while considering the principles of equivalent treatment of cohorts. As previous co-
horts were more prosperous, it would make sense to focus policies on the group that
would balance out the adverse period effects.
It’s worth noting that the foregoing policy discussion is based on descriptive
results. A more comprehensive analysis using a causal study setting is needed to
reveal possible real-world effects. It is challenging to test the causal relationship
of policy effects, as this would require tracking individuals’ income, wealth, and
consumption habits, which is not possible as both wealth and consumption data are
not tracked over the years at the individual level.
7.3 Methodological considerations
This dissertation utilized notably long-running time-series datasets from Statistics
Finland, both in survey and register form. These datasets spanned periods from 31
to 52 years depending on the sub-study, thereby enabling long-term cohort analysis
with sociologically intriguing characteristics. Without such comprehensive and high-
quality data, the age-period-cohort approach would be unfeasible. Furthermore, the
register data on the entire Finnish population allowed tracking of individuals over
time, capturing intricate details of the life course.
54
DISCUSSION
The integration of register analysis with the age-period-cohort methodology us-
ing APCD, APCT, and ACPGOmodels yielded insightful results, extending previous
findings in a significant way. The age-period-cohort analysis facilitated examination
of age patterns across the adult life course and identified which of the three dimen-
sions of APC were relevant for economic measures. However, conclusions about
birth cohort effects should be interpreted as descriptive suggestions, as even the most
sophisticated age-period-cohort methods cannot entirely untangle the relationship
between age, period, and cohort. Replication of the results would dispel any doubts,
and striving for more accurate datasets, especially concerning wealth and consump-
tion, would be ideal.
It is worth noting that some researchers have debated whether solving the ‘identi-
fication problem’ around APC methodology is possible (Bell & Jones, 2013, 2015):
distinguishing age, period, and birth cohort effects, including linear effects if they
exist. The creators of the model and others have countered that simulations are an un-
suitable method for assessing the significance of APC methods (Reither et al., 2015).
This debate influenced my choice of models, as APCD, APCT, and APCGO models
have been employed in empirical research and have yielded reliable estimates (see
Chancel, 2014; Chauvel, 2011, 2013; Chauvel & Schroder, 2015; Freedman, 2017).
This dissertation underwent several robustness checks, and these methods did not
encounter the same issues as some of the other APC models (see discussion on IE
model, Luo, 2013; O’Brien, 2011). Even manipulating data to undermine the model
does not significantly alter the results, provided the data have a high number of cases.
While methodological questions remain unanswered, these methods have offered
valuable insights into several questions that would otherwise remain unknown. Thus,
considering modelling limitations, this methodological approach provided a signifi-
cant advantage in terms of resolving the age-period-cohort conundrum with respect
to economic inequality, whereas traditional models only utilize two of the three di-
mensions.
7.4 Conclusion
This dissertation has made a substantial contribution to the literature by adopting a
holistic perspective to economic measures and employing a life course approach. To
reiterate, this multidimensional research framework has provided new insights into
the development of income, consumption, and wealth, while considering the move-
ment of individuals through different periods and respective cohorts which make
these groups unique, not forgetting their life courses within. The use of the life course
approach as a theoretical and methodological amalgamation has provided new and
novel information.
To specify, this thesis contributed to two main theoretical domains mentioned in
the introduction.
55
Esa Karonen
First, the dissertation contributed to the life course perspective in the theoretical
level. The strength of the life course perspective is its focus on the interconnectivity
of factors that change around the individuals over time, yet the framework itself does
not make assumptions about which of the moving parts holds more weight. The
dissertation did disentangle this issue by finding that both cohort and period play a
more important role in economic resources than age. This has major implications
in Elder’s and Giele’s framework on individual agency, as period dictates the ‘stage’
which individual agency can take place. Thus, differences are found between cohorts
because the periodical constraints dictate how an individual can act at given time,
which in turn makes age conditional on the two dimensions. Thus, from theoretical
perspective, Giele’s emphasis on structural change over individual’s like Elder seems
to be more dominant in empirical sense, yet individual choices remain valid (yet
constrained) calling for using both viewpoints to form a robust theoretical life course
perspective.
Second, the multidimensional approach contributed to the literature on economic
inequality by finding a linkage between income, consumption, and wealth. As previ-
ous research has mainly focused on individual economic resources, this dissertation
provided insight into how income loss impacts both consumption and wealth. The
decreased income profiles had lead to belated dwelling acquisition and decreased
spending profiles on non-necessities among 1980’s cohort, identifying how income
serves as the main driver for consumption and wealth while periodical shocks’ im-
pact on earnings can have far-reaching consequences on all three components, not
just one.
Furthermore, the dissertation contributed in the methodological field, as the use
of new age-period-cohort modelling techniques has maximized the utility of sur-
vey data. Without this approach, it would have not identified theoretical insights on
the life course perspective mentioned above. The high-quality registers with multi-
level models enabled tracking of the entire population over time. Finally, studying
the social determinants of income, consumption, and wealth between cohorts has
led to a more profound understanding of population-level changes and could pro-
vide possible avenues for targeting policies to ensure economic prosperity for future
generations. Thus, while most of the studies on economic resources fall within the
domain of economics, the sociological approach has its merits. The flexibility and
novelty of the concepts highlight the possibilities of a multidimensional approach
over generations. Such an approach is not just warranted, it is complementary.
In conclusion, this dissertation has found that income, consumption, and wealth
have been on the rise. However, not all cohorts have been so fortunate. It was dis-
covered that the 1980s cohort, in particular, is facing economic challenges, as they
had to confront a financial crisis during their school-to-work transition phase. Their
income development has stagnated, leading to a similar stumble in wealth. This has
partly affected home ownership, as the age of acquisition has been delayed com-
56
DISCUSSION
pared to other cohorts, while simultaneously, younger generations have had to take
on larger loans than before. As a result, the time required for loan amortization is
lengthy, which will, in turn, affects the total wealth available at the end of the life
course. Even free-time consumption has shifted increasingly towards high-income
earners, which could leave the 1980s generation excluded from self-improvement
spaces. However, the ‘generational game’ winner was the Baby Boomer generation,
which had the highest relative income and total wealth. Nevertheless, younger gen-
erations have a higher overall income level, and wealth begins to accumulate at an
earlier age. Furthermore, returns to education have dwindled due to credential in-
flation, as younger generations are far more educated than their older counterparts,
intensifying inter-cohort competition.
The widening cohort differences in income, consumption, and wealth should
command our attention. The younger generation is taking on the role of torchbearers,
responsible for funding the previous generation’s pensions, healthcare, and social
security system. These generations bear the burden of the economy while striving to
build their own lives and future generations. Hence, this motivates a continuation of
this research to expand on the findings of this dissertation. Future research should
further illuminate under which types of recessions are particularly consequential for
labor market entrants’ future earnings and wealth trajectories. The novel findings of
the dissertation call for a comparative extension to fully elaborate their implications
for theories of social change and how economic cycles impact economic attainment
across generations in different institutional environments.
I began this journey by stating that resources are scarce, with the underlying
question being “how is the pie sliced?” I have discovered that the so-called ‘mil-
lennials’ are on a diet, as they have seen more modest development in economic
resources than their older counterparts. While the situation is not as dramatic as it
may sound, for the overall development of the Finnish economy and future genera-
tions, it is hoped that this trend is either a temporary hiccup or can be reversed.
57
58 
List of References 
 
Aaberge, R., & Brandolini, A. (2015). Multidimensional poverty and inequality. In Handbook of 
income distribution (Vol. 2, pp. 141-216). Elsevier. 
Abercrombie, N., & Longhurst, B. (1983). Interpreting Mannheim. Theory, Culture & Society, 2(1), 
5-15.  
Alessie, R., & De Ree, J. (2009). Explaining the hump in life cycle consumption profiles. De 
Economist, 157, 107-120. 
Alkire, S., Roche, J. M., Ballon, P., Foster, J., Santos, M. E., & Seth, S. (2015). Multidimensional 
poverty measurement and analysis. Oxford University Press, USA. 
Allardt, E. (1981). Experiences from the comparative Scandinavian welfare study, with a 
bibliography of the project. European Journal of Political Research, 9(1), 101-111. 
Alm, J., & El-Ganainy, A. (2013). Value-added taxation and consumption. International Tax and 
Public Finance, 20, 105-128.  
Alonso-García, J., Bateman, H., Bonekamp, J., van Soest, A., & Stevens, R. (2022). Saving 
preferences after retirement. Journal of Economic Behavior & Organization, 198, 409-433. 
Arundel, R., & Ronald, R. (2021). The false promise of homeownership: Homeowner societies in 
an era of declining access and rising inequality. Urban Studies, 58(6), 1120-1140. 
Anand, S., & Sen, A. (2000). Human development and economic sustainability. World development, 
28(12), 2029-2049. 
Atkinson, A. B., & Bourguignon, F. (1982). The comparison of multi-dimensioned distributions of 
economic status. The Review of Economic Studies, 49(2), 183-201. 
Attanasio, O. P., Banks, J., Meghir, C., & Weber, G. (1999). Humps and bumps in lifetime 
consumption. Journal of Business & Economic Statistics, 17(1), 22-35. 
Bar-Haim, E., Chauvel, L., Gornick, J. C., & Hartung, A. (2023). The persistence of the gender 
earnings gap: Cohort trends and the role of education in twelve countries. Social Indicators 
Research, 165(3), 821-841. 
Barr, N., & Diamond, P. (2006). The economics of pensions. Oxford review of economic policy, 
22(1), 15-39. 
Becker, G. S. (1962). Investment in human capital: A theoretical analysis. Journal of political 
economy, 70(5, Part 2), 9-49. 
Bell, A., & Jones, K. (2013): The impossibility of separating age, period and cohort effects. Social 
Science Medicine, 93, 163–165. 
Bell, A., & Jones, K. (2015). Age, period and cohort processes in longitudinal and life course 
analysis: A multilevel perspective. A life course perspective on health trajectories and 
transitions, 197-213. 
Bengtson, V. L., Elder Jr, G. H., & Putney, N. M. (2012). The life course perspective on ageing: 
Linked lives, timing, and history. Adult lives: A life course perspective, 9-17. 
Bengtson, V. L., Furlong, M. J., & Laufer, R. S. (1974). Time, aging, and the continuity of social 
structure: Themes and Issues in Generational Analysis 1. Journal of Social Issues, 30(2), 1-30. 
List of References 
 59 
Benzarti, Y., Carloni, D., Harju, J., & Kosonen, T. (2020). What goes up may not come down: 
asymmetric incidence of value-added taxes. Journal of Political Economy, 128(12), 4438-4474. 
Berloffa, G., & Villa, P. (2010). Differences in equivalent income across cohorts of households: 
evidence from Italy. Review of Income and Wealth, 56(4), 693-714. 
Bernstein, A., & Koudijs, P. (2020). Mortgage amortization and wealth accumulation. Stanford 
Graduate School of Business. 
Bihagen, E. (1999). How do classes make use of their incomes?. Social Indicators Research, 47, 
119-151. 
Blinder, A. S. (1973). Wage discrimination: Reduced form and structural estimates. The Journal of 
Human Resources, 8(4), 436–455. 
Blomgren, J., Hiilamo, H., Kangas, O., & Niemelä, M. (2014). chapter 10 Finland: Growing 
Inequality With Contested Consequences. Changing Inequalities and Societal Impacts in Rich 
Countries: Thirty Countries' Experiences, 222. 
Blundell, R., & Preston, I. (1995). Income, expenditure and the living standards of UK households. 
Fiscal Studies, 16(3), 40-54. 
Blundell, R., & Preston, I. (1998). Consumption inequality and income uncertainty. The Quarterly 
Journal of Economics, 113(2), 603-640. 
Bol, T. (2015). Has education become more positional? Educational expansion and labour market 
outcomes, 1985–2007. Acta sociologica, 58(2), 105-120. 
Breen, R., Luijkx, R., Müller, W., & Pollak, R. (2009). Nonpersistent inequality in educational 
attainment: Evidence from eight European countries. American journal of sociology, 114(5), 
1475-1521. 
Breen, R., Luijkx, R., Müller, W., & Pollak, R. (2010). Long-term trends in educational inequality 
in Europe: Class inequalities and gender differences. European sociological review, 26(1), 31-
48. 
Bronfenbrenner, M. (1973). Equality and equity. The ANNALS of the American Academy of Political 
and Social Science, 409(1), 9-23. 
Butz, W. P., & Torrey, B. B. (2006). Some frontiers in social science. Science, 312(5782), 1898-
1900.  
Bönke, T., Corneo, G., & Lüthen, H. (2015). Lifetime earnings inequality in Germany. Journal of 
Labor Economics, 33(1), 171-208. 
Cagetti, M. (2003). Wealth accumulation over the life cycle and precautionary savings. Journal of 
Business & Economic Statistics, 21(3), 339-353.  
Campbell, A., Converse, P. E., & Rodgers, W. L. (1976). The quality of American life: Perceptions, 
evaluations, and satisfactions. Russell Sage Foundation. 
Canberra Group (2011). Handbook on Household Income Statistics. Second Edition. United 
Nations, Geneva. 
Cannon, L., & Kendig, H. (2018). ‘Millennials’: Perceived generational opportunities and 
intergenerational conflict in Australia. Australasian Journal on Ageing, 37(4), E127-E132. 
Chai, A., Rohde, N., & Silber, J. (2015). Measuring the diversity of household spending patterns. 
Journal of Economic Surveys, 29(3), 423-440. 
Chancel, L. (2014). Are younger generations higher carbon emitters than their elders?: 
Inequalities, generations and CO2 emissions in France and in the USA. Ecological 
Economics, 100, 195-207. 
Charles, K. K., & Hurst, E. (2003). The correlation of wealth across generations. Journal of political 
Economy, 111(6), 1155-1182. 
Chauvel, L. (2011). Age-period-cohort detrended APC-D model. 
Chauvel, L. (2013). Welfare regimes, cohorts and the middle classes. Gornick, J. C., & Jäntti, M. 
(Eds.). Income inequality: Economic disparities and the middle class in affluent countries. 
Stanford University Press, 115-141. 
 60 
Chauvel, L., & Bar-Haim, E. (2017). ISOGRAPH: Stata module to compute inequality over logit 
ranks of social hierarchy. Boston College Department of Economics. 
Chauvel, L., & Schröder, M. (2015). The impact of cohort membership on disposable incomes in 
West Germany, France, and the United States. European Sociological Review, 31(3), 298-311. 
Chauvel, L., Hartung A., & Bar-Haim, E. (2017). APCGO: Stata module to calculate age-period-
cohort effects for the gap between two groups (based on a Blinder-Oaxaca decomposition), 
including trends for each parameter. 
Chetty, R., Grusky, D., Hell, M., Hendren, N., Manduca, R., & Narang, J. (2017). The fading 
American dream: Trends in absolute income mobility since 1940. Science, 356(6336), 398-
406. 
Christelis, D., Jappelli, T., Paccagnella, O., & Weber, G. (2009). Income, wealth and financial 
fragility in Europe. Journal of European Social Policy, 19(4), 359-376.  
Chybalski, F., & Marcinkiewicz, E. (2018). Myopic behaviour and state involvement in a pension 
system: A cross-section study for OECD countries. Ekonomista, 1, 68-88. 
Clark, A. E., & Oswald, A. J. (1994). Unhappiness and unemployment. The Economic Journal, 
104(424), 648-659. 
Cohen, S. B. (2016). The concentration of health care expenditures in the US and predictions of 
future spending. Journal of Economic and Social Measurement, 41(2), 167-189. 
Conley, D. (2001). Decomposing the black‐white wealth gap: The role of parental resources, 
inheritance, and investment dynamics. Sociological Inquiry, 71(1), 39-66. 
Conley, D., & Glauber, R. (2008). All in the family?: Family composition, resources, and sibling 
similarity in socioeconomic status. Research in Social Stratification and Mobility, 26(4), 297-
306. 
Conley, D., & Ryvicker, M. (2005). The Price of Female Headship: Gender, Inheritance and Wealth 
Accumulation in the United States. Journal of Income Distribution, 13(3-4), 3-3. 
Crawford, R., Innes, D., & O'Dea, C. (2016). Household wealth in Great Britain: distribution, 
composition and changes 2006–12. Fiscal Studies, 37(1), 35-54. 
Credit Suisse (2023). Global Wealth Report 2023. UBS group. 
Daley, J., & Wood, D. (2014). The wealth of generations. 
Davies, J. B., Sandström, S., Shorrocks, A., & Wolff, E. (2009). The global pattern of household 
wealth. Journal of International Development: The Journal of the Development Studies 
Association, 21(8), 1111-1124. 
Davis, K., & Moore, W. E. (1945). Some Principles of Stratification. American Sociological Review, 
10(2), 242. 
Decancq, K., & Lugo, M. A. (2012). Inequality of wellbeing: A multidimensional approach. 
Economica, 79(316), 721-746. 
Deutsch, J., Guio, A. C., Pomati, M., & Silber, J. (2015). Material deprivation in Europe: which 
expenditures are curtailed first?. Social Indicators Research, 120, 723-740. 
Doepke, M., & Schneider, M. (2006). Inflation and the redistribution of nominal wealth. Journal of 
Political Economy, 114(6), 1069-1097. 
Douglas, M., & Isherwood, B. C. (1996). The world of goods: Towards an anthropology of 
consumption. Psychology Press. 
Dutt, P., & Padmanabhan, V. (2011). Crisis and consumption smoothing. Marketing Science, 30(3), 
491-512.  
Dworkin, R. (1981). What is equality? Part 2: Equality of resources, philosophy and public affairs. 
Sovereign Virtue: The Theory and Practice of Equality. 
Elder Jr, G. H. (1994). Time, human agency, and social change: Perspectives on the life course. 
Social psychology quarterly, 4-15. 
Elder, G. H., Johnson, M. K., & Crosnoe, R. (2003). The emergence and development of life course 
theory. In Handbook of the life course (pp. 3-19). Springer, Boston, MA. 
List of References 
 61 
Engbom, N. (2019). Firm and worker dynamics in an aging labor market (No. 756, pp. 1-58). 
Minneapolis, MN: Federal Reserve Bank of Minneapolis. 
Erola, J., & Wilska, T. A. (2004). Yhteiskunnan moottori vai kivireki?: suuret ikäluokat ja 1960-
lukulaisuus (No. 92). Jyväskylän yliopisto. 
Fadda, S., & Tridico, P. (2016). Varieties of Economic Inequality. Routledge. 
Fan, J. X., & Lewis, J. K. (1999). Budget allocation patterns of African Americans. Journal of 
Consumer Affairs, 33(1), 134-164. 
Fasang, A. E., & Mayer, K. U. (2020). Lifecourse and social inequality. In Handbook on 
demographic change and the lifecourse (pp. 22-39). Edward Elgar Publishing. 
Felson, M. (1976). The differentiation of material life styles: 1925 to 1966. Social Indicators 
Research, 3, 397-421. 
Fernandez-Villaverde, J., & Krueger, D. (2011). Consumption and saving over the life cycle: How 
important are consumer durables?. Macroeconomic dynamics, 15(5), 725-770. 
Finch, J. (2018) Age. In Key variables in social investigation (Vol. 4). Routledge. 
Frankenberg, E., Smith, J. P., & Thomas, D. (2003). Economic shocks, wealth, and welfare. Journal 
of Human Resources, 38(2), 280-321. 
Freedman, M. (2017). Are Recent Generations Catching Up or Falling Behind? Trends in Inter-
Generational Inequality (No. 689). LIS Working Paper Series.  
Freeman, R. B. (2006). Is a great labor shortage coming? Replacement demand in the global 
economy. 
Frey, B., & Stutzer, A. (2002). The economics of happiness. World Economics-Henley On Thames-
, 3(1), 25-42. 
Friedline, T., Masa, R. D., & Chowa, G. A. (2015). Transforming wealth: Using the inverse 
hyperbolic sine (IHS) and splines to predict youth’s math achievement. Social science research, 
49, 264-287. 
Friedman, M. (1957). The permanent income hypothesis. In A theory of the consumption function 
(pp. 20-37). Princeton University Press.  
Gangl, M. (2005). Income inequality, permanent incomes, and income dynamics: Comparing 
Europe to the United States. Work and Occupations, 32(2), 140-162. 
Giannakis, E., & Bruggeman, A. (2020). Regional disparities in economic resilience in the European 
Union across the urban–rural divide. Regional Studies, 54(9), 1200-1213. 
Giele, J. Z., & Elder, G. H. (Eds.). (1998). Methods of life course research: Qualitative and 
quantitative approaches. Sage. 
Glenn, N. D. (1976). Cohort analysts' futile quest: Statistical attempts to separate age, period and 
cohort effects. American sociological review, 41(5), 900-904. 
Goldthorpe, J. H., & McKnight, A. (2006). The economic basis of social class. Mobility and 
inequality: Frontiers of research in sociology and economics, 109-136. 
Góra, M. (2008). Retirement decisions, benefits and the neutrality of pension systems (No. 51). 
CEPS. 
Gourinchas, P. O., & Parker, J. A. (2002). Consumption over the life cycle. Econometrica, 70(1), 
47-89. 
Greene, R. H. (2014). How financial literacy influences long-and short-term financial behaviors in 
different age cohorts. University of Georgia. 
Gruber, J., & Wise, D. A. (Eds.). (2010). Social security programs and retirement around the world: 
The relationship to youth employment. University of Chicago Press. 
Haapakorva, P., Ristikari, T., & Gissler, M. (2017). The impact of parental employment trajectories 
on the children’s early adulthood education and employment trajectories in the Finnish birth 
cohort 1987. Longitudinal and Life Course Studies, 8(4), 342-364. 
Hammer, B., Spitzer, S., & Prskawetz, A. (2022). Age-specific income trends in Europe: The role 
of employment, wages, and social transfers. Social indicators research, 162(2), 525-547. 
 62 
Hampson, D. P., & McGoldrick, P. J. (2013). A typology of adaptive shopping patterns in recession. 
Journal of business research, 66(7), 831-838. 
Hansen, M. N. (2014). Self-made wealth or family wealth? Changes in intergenerational wealth 
mobility. Social Forces, 93(2), 457-481. 
Hareven, T. K. (1996). What difference does it make?. Social Science History, 20(3), 317-344. 
Helliwell, J. F. (2003). How's life? Combining individual and national variables to explain 
subjective well-being. Economic modelling, 20(2), 331-360.  
Helliwell, J., Layard, R., & Sachs, J. (2012). World happiness report. 
Heslop, L. A. (1987). Cohort analysis of the expenditure patterns of the elderly. ACR North 
American Advances, NA-14. 
Hirsch, F. (2005). Social limits to growth. Routledge. 
Hobcraft, J., Menken, J., & Preston, S. (1985). Age, period, and cohort effects in demography: a 
review (pp. 89-135). Springer New York. 
Hurst, E., Luoh, M. C., Stafford, F. P., & Gale, W. G. (1998). The wealth dynamics of American 
families, 1984-94. Brookings papers on Economic activity, 1998(1), 267-337. 
Jann, B. (2008). The Blinder-Oaxaca decomposition for linear regression models. The Stata Journal, 
8(4), 453–479. 
Johnson, R. C., & Schoeni, R. F. (2011). The influence of early-life events on human capital, health 
status, and labor market outcomes over the life course. The BE journal of economic analysis & 
policy, 11(3).  
Jonung, L., Kiander, J., & Vartia, P. (Eds.). (2009). The great financial crisis in Finland and Sweden: 
the Nordic experience of financial liberalization. Edward Elgar Publishing. 
Kahneman, D., Diener, E., & Schwarz, N. (Eds.). (1999). Well-being: Foundations of hedonic 
psychology. Russell Sage Foundation. 
Kangas, O., & Karonen, E. (2023). Creative destruction: Consequences of the collapse of Nokia 
Phones in Oulu. The Practical Side Of Digital Transformation: A Tool Book For Practitioners, 
115. 
Karisto, A. (2005). Suuret ikäluokat kuvastimessa. In Suuret ikäluokat (pp. 17-58). Vastapaino. 
Katz-Gerro, T. (2003). Consumption-based inequality: Household expenditures and possession of 
goods in Israel, 1986–1998. Transitions in domestic consumption and family life in the modern 
middle east, 167. 
Katz-Gerro, T., & Talmud, I. (2005). Structural analysis of a consumption-based stratification 
indicator: Relational proximity of household expenditures. Social Indicators Research, 73(1), 
109–132. 
Kershaw, P. (2015). Population Aging, Generational Equity & The Middle Class. Vancouver, BC: 
Generation Squeeze. 
Kertzer, D. I. (1983). Generation as a sociological problem. Annual review of sociology, 9(1), 125-
149. 
Koelln, K., Rubin, R. M., & Picard, M. S. (1995). Vulnerable elderly households: Expenditures on 
necessities by older Americans. Social Science Quarterly, 619-633. 
Kolb, K., Skopek, N., & Blossfeld, H. P. (2013). The two dimensions of housing inequality in 
Europe: Are high home ownership rates an indicator of low housing values?. Comparative 
Population Studies, 38(4). 
Kolsrud, J., Landais, C., & Spinnewijn, J. (2017). Studying consumption patterns using registry 
data: Lessons from Swedish administrative data. 
Krueger, D., & Perri, F. (2006). Does income inequality lead to consumption inequality? Evidence 
and theory. The Review of Economic Studies, 73(1), 163-193. 
Kurz, K. Blossfeld, H.-P. (2004). Home ownership and social inequality in comparative perspective. 
Stanford: Stanford University Press. 
List of References 
 63 
Kurz, K., & Blossfeld, H. P. (2004). Social Stratification and Access to Home Ownership. Home 
ownership and social inequality in comparative perspective. 
Layard, Richard, et al. (2014). What predicts a successful life? A life-course model of well-being. 
The Economic Journal 124(580), F720-F738. 
Levitas, R., Pantazis, C., Fahmy, E., Gordon, D., Lloyd-Reichling, E., & Patsios, D. (2007). The 
multi-dimensional analysis of social exclusion. 
Lim, G. C., & Zeng, Q. (2016). Consumption, income, and wealth: Evidence from age, cohort, and 
period elasticities. Review of Income and Wealth, 62(3), 489-508. 
Lovenheim, M. F., & Reynolds, C. L. (2013). The effect of housing wealth on college choice: 
Evidence from the housing boom. Journal of Human Resources, 48(1), 1-35. 
Luo, L. (2013). Assessing validity and application scope of the intrinsic estimator approach to the 
age-period-cohort problem. Demography, 50(6), 1945-1967. 
Lusardi, A., Michaud, P. C., & Mitchell, O. S. (2017). Optimal financial knowledge and wealth 
inequality. Journal of political Economy, 125(2), 431-477. 
Lüscher, K., Hoff, A., Lamura, G., Renzi, M., Sánchez, M., Viry, G., Widmer, E., Klimczuk, A., de 
Salles Oliveira, P., Neményi, Á. & Veress, E., (2016). Generations, intergenerational 
relationships, generational policy: A multilingual compendium. Andrzej Klimczuk.  
Mckenzie, D., Schargrodsky, E., & Cruces, G. (2011). Buying less but shopping more: The use of 
nonmarket labor during a crisis. Economía, 11(2), 1–43 
Mastrogiacomo, M. (2015). Will we repay our debts before retirement? Or did we already, but 
nobody noticed. The legacy of Interest-Only Mortgages, Voluntary Repayments and Savings 
Deposits in the Netherlands. Netspar paper. 
Malmendier, U., & Nagel, S. (2011). Depression babies: Do macroeconomic experiences affect risk 
taking?. The quarterly journal of economics, 126(1), 373-416. 
Manduca, R., Hell, M., Adermon, A., Blanden, J., Bratberg, E., Gielen, A. C., van Kippersluis, H., 
Lee, K. B., Machin, S. J., Munk, M., Nybom, M., Ostrovsky, Y., Rahman, S. & Sirniö, O. 
(2020). Trends in Absolute Income Mobility in North America and Europe. IZA Discussion 
Paper No. 13456. 
Mannheim, K. (1928). Das Problem der Generationen (1928/29). Jugend in der modernen 
Gesellschaft, 23-48. 
Marx, K. (1873). Capital: A critical analysis of capitalist production. Humboldt. 
Mason, K. O., Mason, W. M., Winsborough, H. H., & Poole, W. K. (1973). Some methodological 
issues in cohort analysis of archival data. American sociological review, 242-258. 
Mastrogiacomo, M. (2016). Will we repay our debts before retirement? Or did we already, but 
nobody noticed. The legacy of Interest-Only Mortgages, Voluntary Repayments and Savings 
Deposits in the Netherlands. Netspar paper. 
Mayer, K. U. (2000). Promises fulfilled? A review of 20 years of life course research. European 
Journal of Sociology/Archives Européennes de Sociologie, 41(2), 259-282. 
Mayer, K. U. (2004). Life courses and life chances in a comparative perspective (pp. 24-25). 
Mannheimer Zentrum für Europäische Sozialforschung. 
Mayer, K. U. (2009). New directions in life course research. Annual review of sociology, 35, 413-
433. 
Mayer, K. U. (2018). Whose lives? How history, societies, and institutions define and shape life 
courses. In Contextual Influences on Life Span/life Course (pp. 161-187). Routledge. 
McCall, L., & Percheski, C. (2010). Income inequality: New trends and research directions. Annual 
review of sociology, 36, 329-347. 
Mckenzie, D., Schargrodsky, E., & Cruces, G. (2011). Buying less but shopping more: The use of 
nonmarket labor during a crisis. Economía, 11(2), 1–43. 
McQuaid, R. W., & Lindsay, C. (2013). The concept of employability. In Employability and Local 
Labour Markets (pp. 6-28). Routledge. 
 64 
Mincer, J. (1958). Investment in human capital and personal income distribution. Journal of 
political economy, 66(4), 281-302.  
Modigliani, F. (1966). The life cycle hypothesis of saving, the demand for wealth and the supply of 
capital. Social research, 160-217. 
Mugiyama, R., & Toyonaga, K. (2022). Role of cohort size in trends in class and occupational 
returns to education at first job: the case of Japan. European Sociological Review, 38(2), 269-
285. 
Mulligan, C. B. (1997). Parental priorities and economic inequality. University of Chicago Press. 
Myles, J. (2002). A new social contract for the elderly. Why we need a new welfare state, 1, 130-
173. 
Neumark, D., Johnson, H., & Mejia, M. C. (2013). Future skill shortages in the US economy?. 
Economics of Education Review, 32, 151-167. 
Nicosia, F. M., & Mayer, R. N. (1976). Toward a sociology of consumption. Journal of consumer 
research, 3(2), 65-75. 
O’brien, R. M. (2011). Constrained estimators and age-period-cohort models. Sociological Methods 
& Research, 40(3), 419-452. 
Oaxaca, R. L. (1973). Male-female wage differentials in urban labor markets. International 
Economic Review, 14(3), 693–709. 
Oaxaca, R. L., & Ransom, M. R. (1994). On discrimination and the decomposition of wage 
differentials. Journal of Econometrics, 61(1), 5–21. 
OECD (2008), Growing Unequal?: Income Distribution and Poverty in OECD Countries, OECD 
Publishing, Paris. 
Official Statistics of Finland (OSF) (2015a): Income distribution statistics [online publication]. 
ISSN=1799-1331. Helsinki: Statistics Finland [Referenced: 7.6.2024]. Access method: 
https://stat.fi/en/statistics/tjt 
Official Statistics of Finland (OSF) (2015b): Households' assets [online publication].ISSN=2242-
3230. Helsinki: Statistics Finland [Referenced: 7.6.2024]. Access method: 
https://stat.fi/en/statistics/vtutk 
Official Statistics of Finland (OSF) (2024). Taika - research data catalogue. 
https://taika.stat.fi/en/index 
Oreopoulos, P., Von Wachter, T., & Heisz, A. (2012). The short-and long-term career effects of 
graduating in a recession. American Economic Journal: Applied Economics, 4(1), 1-29. 
Parsons, T. (1966). Societies: Evolutionary and comparative perspectives. 
Pekkala, S., & Lucas, R. E. (2007). Differences across cohorts in Finnish intergenerational income 
mobility. Industrial Relations: A Journal of Economy and Society, 46(1), 81-111.  
Pekkarinen, J., & Vartiainen, J. (1993). Suomen talouspolitiikan pitkä linja. Jyväskylän yliopisto. 
Pelzer, B., Te Grotenhuis, M., Eisinga, R., & Schmidt-Catran, A. W. (2015). The non-uniqueness 
property of the intrinsic estimator in APC models. Demography, 52(1), 315-327. 
Pence, K. M. (2006). The role of wealth transformations: An application to estimating the effect of 
tax incentives on saving. Contributions in Economic Analysis & Policy, 5(1). 
Perugini, C., & Martino, G. (2008). Income inequality within European regions: Determinants and 
effects on growth. Review of Income and Wealth, 54(3), 373-406. 
Pfeffer, F. T., & Killewald, A. (2018). Generations of advantage. Multigenerational correlations in 
family wealth. Social Forces, 96(4), 1411-1442. 
Piachaud, D., J. Macnicol, and J. Lewis (2009), A Think-Piece on Intergenerational Equity (London: 
Equality and Human Rights Commission). 
Piketty, T. (2014). Capital in the twenty-first century. Harvard University Press.  
Pilcher, J. (1994). Mannheim's sociology of generations: an undervalued legacy. British Journal of 
sociology, 481-495.  
List of References 
 65 
Pirttilä, J., & Selin, H. (2011). Income shifting within a dual income tax system: Evidence from the 
Finnish tax reform of 1993. Scandinavian Journal of Economics, 113(1), 120-144. 
Polanyi, K. (2001). The great transformation: The political and economic origins of our time. 
Beacon press. 
Purhonen, S. (2007). Sukupolvien ongelma: tutkielmia sukupolven käsitteestä, 
sukupolvitietoisuudesta ja suurista ikäluokista. 
Rawls, A. (1971). Theories of social justice. 
Rawls, J. (1999). A theory of justice: Revised edition. 
Reither, E.N., Land, K.C., Jeon, S.Y., Powers, D.A., Masters, R.K., Zheng, H., Hardy, M.A., Keyes, 
K.M., Fu, Q., Hanson, H.A. & Smith, K.R., (2015). Clarifying hierarchical age–period–cohort 
models: A rejoinder to Bell and Jones. Social Science & Medicine, 145, pp.125-128. 
Riihelä, M. (2006). Kotitalouksien kulutus ja säästäminen: Ikäprofiilien ja kohorttien kuvaus. VATT 
discussion papers, 386.  
Riihelä, M., Sullström, R., & Tuomala, M. (2014). Ikä, sukupuoli ja tuloerot. Yhteiskuntapolitiikka, 
49(6), 669-678. 
Roberts, K. (2012). The end of the long baby-boomer generation. Journal of youth studies, 15(4), 
479-497. 
Roikonen, P. (2022). Income inequality in Finland, 1865–2019. Scandinavian Economic History 
Review, 70(3), 234-251.  
Roos, J.P., (1980). Elämisen laatu ja elämäntapa 1970-luvulla. Sosiaalinen aikakauskirja 6. 
Rycroft, R. S. (Ed.). (2013). The Economics of Inequality, Poverty, and Discrimination in the 21st 
Century. Bloomsbury Publishing USA. 
Rydell, I. (2005). Equity, justice, interdependence: intergenerational transfers and the ageing 
population (No. 2005: 5). Institute for Futures Studies. 
Ryder, N. B. (1965). The cohort as a concept in the study of social change. American sociological 
review, 30(6), 843-861. 
Räsänen, P. (2003). In the twilight of social structures: A mechanism-based study of contemporary 
consumer behavior. Turun yliopisto. 
Salcedo, A. M., & Izquierdo Llanes, G. (2020). Refining the monetary poverty indicators under a 
join income-consumption statistical approach: An application to Spain based on empirical data. 
Social Indicators Research, 147(2), 501-516. 
Schoon, I., & Bynner, J. (2019). Young people and the Great Recession: Variations in the school-
to-work transition in Europe and the United States. Longitudinal and Life Course Studies, 
10(2), 153-173. 
Segall, S. (2013). Equality and opportunity. OUP Oxford. 
Semyonov, M., Levin-Epstein, N., & Spilerman, S. (1996). The material possessions of Israeli 
ethnic groups. European Sociological Review, 12(3), 289-301.  
Sen, A. (1995). Inequality reexamined. Harvard university press. 
Sen, A. K. (1997). From income inequality to economic inequality. Southern Economic Journal, 
64(2), 384-401. 
Settersten Jr, R. A. (2003). Age structuring and the rhythm of the life course. In Handbook of the 
life course (pp. 81-98). Boston, MA: Springer US. 
Sierminska, E., & Takhtamanova, Y. (2007). Disentangling the wealth effect: some international 
evidence. Federal Reserve Bank of San Francisco. 
Slater, D. (1999). Consumer culture and modernity. Wiley. 
Slesnick, D. T. (2001). Consumption and social welfare: Living standards and their distribution in 
the United States. Cambridge University Press.  
Stand, D. W., & Rising, W. I. K. (2011). An overview of growing income inequalities in OECD 
countries: main findings. Divided we stand: Why inequality keeps rising, 21-45. 
Stöver, B. (2012). The influence of age on consumption (No. 3808). EcoMod. 
 66 
Summers, J. K., & Smith, L. M. (2014). The role of social and intergenerational equity in making 
changes in human well-being sustainable. Ambio, 43, 718-728. 
Suoniemi, I. (2012). Income mobility, income risk and age: Finnish experiences in 1995–2008. 
Toivonen, T. (2003). Sukupolvi yhteiskunnallisen muutoksen selittäjänä. Teoksessa Melin, Harri & 
Nikula, Jouko (toim.): Yhteiskunnallinen muutos. Tampere: Vastapaino, 109-120. 
Törmälehto V-M. (2018). Tilastokeskus - Kotitalouksien varallisuus 2016. Statistics Finland. 
Tsui, K. Y. (1995). Multidimensional generalizations of the relative and absolute inequality indices: 
the Atkinson-Kolm-Sen approach. Journal of Economic Theory, 67(1), 251-265. 
Tsui, K. Y. (1999). Multidimensional inequality and multidimensional generalized entropy 
measures: An axiomatic derivation. Social Choice and Welfare, 16(1), 145-157. 
Twisk J. W. R. (2004). Longitudinal Data Analysis. A Comparison Between Generalized Estimating 
Equations and Random Coefficient Analysis. European Journal of Epidemiology, 19(8), 769–
776. 
Urbonavicius, S., & Pikturniene, I. (2010). Consumers In The Face Of Economic Crisis: Evidence 
From Two Generations In Lithuania. Economics & Management. 
Uusitalo, L. (1980). Identification of consumption style segments on the basis of household budget 
allocation. Advances in consumer research, 7(1). 
Van Dalen, H. P., & Henkens, K. (2002). Early-retirement reform: can it and will it work?. Ageing 
& Society, 22(2), 209-231. 
Van de Werfhorst, H. G., & Andersen, R. (2005). Social background, credential inflation and 
educational strategies. Acta sociologica, 48(4), 321-340. 
Veblen, T. (1899).  The Theory of the Leisure Class: An Economic Study of Institutions. New York: 
Macmillan 
Verho, J. (2008). Scars of recession: the long-term costs of the Finnish economic crisis (No. 2008: 
9). Working Paper. 
Virtanen, M. (2001). Fennomanian perilliset. Poliittiset traditiot ja sukupolvien dynamiikka. 
Hämeenlinna: SKS, 2001. 
Virtanen, M. (1999). Sukupolven tasot, fraktiot ja elämänkaari. Sosiologia 36 (2), 81–94. 
Wachter, T. V. (2020). The persistent effects of initial labor market conditions for young adults and 
their sources. Journal of Economic Perspectives, 34(4), 168-194. 
Weber, M. (1922) Economy and Society. University of California Press, Berkeley. 
Weber, M. (1946). From Max Weber. Translated by Gerth, Hans; Wright Mills, C. New York: Free 
Press.  
Webster, N. J., Antonucci, T. C., & Ajrouch, K. J. (2022). Linked lives and convoys of social 
relations. Advances in life course research, 54, 100502. 
Weymark, J. A. (2006). The normative approach to the measurement of multidimensional 
inequality. Inequality and economic integration, 303-328.  
Willson, A. E., Shuey, K. M., & Elder, Jr, G. H. (2007). Cumulative advantage processes as 
mechanisms of inequality in life course health. American Journal of Sociology, 112(6), 1886-
1924. 
Wolff, E. N. (2017). Household wealth trends in the United States, 1962 to 2016: has middle class 
wealth recovered? (No. w24085). National Bureau of Economic Research. 
Wrzus, C., Hänel, M., Wagner, J., & Neyer, F. J. (2013). Social network changes and life events 
across the life span: a meta-analysis. Psychological bulletin, 139(1), 53. 
Yang, Y., Fu, W. J., & Land, K. C. (2004). A methodological comparison of age‐period‐cohort 
models: the intrinsic estimator and conventional generalized linear models. Sociological 
methodology, 34(1), 75-110. 
Yang, Y., Schulhofer-Wohl, S., Fu, W. J., & Land, K. C. (2008). The intrinsic estimator for age-
period-cohort analysis: what it is and how to use it. American Journal of Sociology, 113(6), 
1697-1736. 
ISBN 978-951-29-9969-9 (PRINT)
ISBN 978-951-29-9970-5 (PDF)
ISSN 0082-6987 (PRINT)
ISSN 2343-3191 (ONLINE)
Pa
in
os
al
am
a,
 T
ur
ku
, F
in
la
nd
 2
02
4